Cs 7.2 dql

EMC®
Documentum®
Content Server
Version 7.2
DQL Reference
EMC Corporation
Corporate Headquarters:
Hopkinton, MA 01748-9103
1-508-435-1000
www.EMC.com

Legal Notice
Copyright © 1994–2015 EMC Corporation. All Rights Reserved.
EMC believes the information in this publication is accurate as of its publication date. The information is subject to change
without notice.
THE INFORMATION IN THIS PUBLICATION IS PROVIDED "AS IS." EMC CORPORATION MAKES NO REPRESENTATIONS
OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY
DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.
For the most up-to-date listing of EMC product names, see EMC Corporation Trademarks on EMC.com. Adobe and Adobe PDF
Library are trademarks or registered trademarks of Adobe Systems Inc. in the U.S. and other countries. All other trademarks
used herein are the property of their respective owners.
Documentation Feedback
Your opinion matters. We want to hear from you regarding our product documentation. If you have feedback about how we can
make our documentation better or easier to use, please send us your feedback directly at IIGDocumentationFeedback@emc.com.

Table of Contents
Preface ................................................................................................................................ 11
Chapter 1 DQL Language Elements ............................................................................. 13
Literals ............................................................................................................ 13
Integer literals .............................................................................................. 13
Floating point literals.................................................................................... 14
Character string literals................................................................................. 14
ID literals..................................................................................................... 15
Date literals.................................................................................................. 15
Default formats ........................................................................................ 15
Short date formats ................................................................................ 16
ANSI format ........................................................................................ 16
Other character string formats............................................................... 16
Date literal keywords................................................................................ 17
Date output formats.................................................................................. 18
Date storage and handling ........................................................................ 18
Special keywords.............................................................................................. 18
Functions......................................................................................................... 19
Scalar functions ............................................................................................ 20
ASCII....................................................................................................... 20
BITAND, BITCLR, BITSET ........................................................................ 20
UPPER..................................................................................................... 21
LOWER .................................................................................................. 21
SUBSTR ................................................................................................... 22
Aggregate functions ..................................................................................... 22
COUNT .................................................................................................. 23
MIN ........................................................................................................ 23
MAX........................................................................................................ 23
AVG ....................................................................................................... 24
SUM ....................................................................................................... 24
Date functions.............................................................................................. 24
DATEDIFF .............................................................................................. 25
DATEADD............................................................................................... 26
DATEFLOOR ........................................................................................... 26
DATETOSTRING ..................................................................................... 26
DATEFLOOR_LOCAL.............................................................................. 27
DATETOSTRING_LOCAL ........................................................................ 28
The CHECK_ACL function ........................................................................... 28
Examples ................................................................................................. 29
The ID function ............................................................................................ 30
The MFILE_URL function ............................................................................. 31
Examples ................................................................................................. 31
Predicates ........................................................................................................ 32
Arithmetic operators..................................................................................... 32
Comparison operators .................................................................................. 32
Column and property predicates ................................................................... 33
Predicates for columns and single-valued properties................................... 33
EMC Documentum Content Server 7.2 DQL Reference 3

Table of Contents
Predicates for repeating properties ............................................................ 34
Pattern matching with LIKE ...................................................................... 36
The percent sign ................................................................................... 36
The underbar ....................................................................................... 37
Matching cases ..................................................................................... 37
The ESCAPE character.......................................................................... 37
SysObject predicates ..................................................................................... 38
The TYPE predicate .................................................................................. 38
The FOLDER predicate ............................................................................. 39
The CABINET predicate............................................................................ 39
Logical operators.............................................................................................. 40
AND and OR ............................................................................................... 40
NOT ............................................................................................................ 40
Order of precedence ..................................................................................... 41
DQL reserved words......................................................................................... 42
Chapter 2 DQL Statements .......................................................................................... 43
Chapter 3 Administration Methods ............................................................................ 165
Invoking administration methods.................................................................... 165
Scope of the administration methods ........................................................... 166
Administration method operations .................................................................. 166
Chapter 4 Using DQL ................................................................................................. 333
Introducing DQL............................................................................................ 333
Quoting object type and property names.......................................................... 335
NULLs, default values, and DQL..................................................................... 335
Default values returned without SPACEOPTIMIZE ...................................... 336
Default values returned with SPACEOPTIMIZE ........................................... 336
Testing for default and NULL values ........................................................... 338
Default values and aggregate functions........................................................ 339
Sorting and nulls ........................................................................................ 339
Repeating properties in queries ....................................................................... 339
Modifying repeating attributes .................................................................... 340
Adding new values................................................................................. 340
Inserting values.................................................................................. 340
Appending values .............................................................................. 341
Updating values ..................................................................................... 341
Deleting values....................................................................................... 342
Forcing index correspondence in query results ............................................. 342
Performance note for MS SQL Server users................................................... 343
Querying virtual documents ........................................................................... 344
Full-text searching and virtual documents........................................................ 344
Querying registered tables .............................................................................. 345
Referencing registered tables in queries........................................................ 345
Security controls......................................................................................... 345
Default object-level permissions and table permits.................................... 346
Caching queries.............................................................................................. 346
Privileges, permissions, and queries................................................................. 346
Appendix A Using DQL Hints ........................................................................................ 349
General guidelines for all ................................................................................ 350
CONVERT_FOLDER_LIST_TO_OR................................................................. 350
4 EMC Documentum Content Server 7.2 DQL Reference

Table of Contents
FETCH_ALL_RESULTS N............................................................................... 350
SQL Server, the hint, and subqueries ........................................................... 351
FORCE_ORDER ............................................................................................. 351
FTDQL and NOFTDQL................................................................................... 351
FT_CONTAIN_FRAGMENT........................................................................... 352
GENERATE_SQL_ONLY ................................................................................ 352
GROUP_LIST_LIMIT N .................................................................................. 352
HIDE_SHARED_PARENT DQL Hint............................................................... 353
IN and EXISTS ............................................................................................... 353
OPTIMIZE_ON_BASE_TABLE........................................................................ 354
OPTIMIZATION_LEVEL level_1 level_2.......................................................... 355
OPTIMIZE_TOP N ......................................................................................... 355
RETURN_RANGE.......................................................................................... 356
RETURN_TOP N............................................................................................ 357
More details about RETURN_TOP N ........................................................... 357
Database-specific implementations .............................................................. 357
SQL Server............................................................................................. 358
Subqueries and the hint ...................................................................... 358
DB2 ................................................................................................... 358
Oracle.................................................................................................... 359
Effects of a SEARCH clause......................................................................... 359
Recommended use ..................................................................................... 359
ROW_BASED................................................................................................. 359
Effects on returned results........................................................................... 359
Effects on query syntax rules....................................................................... 360
SQL_DEF_RESULT_SET N.............................................................................. 361
TRY_FTDQL_FIRST........................................................................................ 362
UNCOMMITTED_READ ............................................................................... 362
DM_LEFT_OUTER_JOIN_FOR_ACL............................................................... 362
Including multiple hints limiting rows returned ............................................... 362
Passthrough hints........................................................................................... 363
Syntax ....................................................................................................... 363
Error handling and debugging .................................................................... 364
Appendix B Database footprint reduction of dmr_content objects ............................... 365
Appendix C IDQL and IAPI ............................................................................................ 367
Using IDQL.................................................................................................... 367
Starting IDQL............................................................................................. 367
The IDQL commands.................................................................................. 369
Entering queries ......................................................................................... 370
Clearing the buffer...................................................................................... 371
Entering comments..................................................................................... 371
Stopping IDQL........................................................................................... 371
Using IAPI ..................................................................................................... 371
Starting IAPI .............................................................................................. 371
IAPI commands.......................................................................................... 374
Entering method calls ................................................................................. 375
Entering comments..................................................................................... 375
Shutting down and restarting Content Server............................................... 375
Quitting an IAPI session ............................................................................. 376

Table of Contents
Appendix D Implementing Java Evaluation of Docbasic Expressions .......................... 377
Docbasic expression handling by Content Server .............................................. 377
How DFC Version 6 and later handles the expressions ...................................... 378
Migrating the expressions for pre-6 clients ....................................................... 378
Repository storage of migrated expressions.................................................. 378
Migrating Docbasic expressions to Java........................................................ 379
Disabling or re-enabling Java evaluation .......................................................... 380
Docbasic expression components support ........................................................ 381
Operators................................................................................................... 381
Supported functions for Java evaluation....................................................... 382
Unsupported functions for Java evaluation................................................... 386
Supported constants ................................................................................... 387
Unsupported constants .............................................................................. 388
Implicit objects ........................................................................................... 388
Appendix E DQL Quick Reference ................................................................................ 389
The DQL statements ....................................................................................... 389
Execute...................................................................................................... 392
DQL reserved words ...................................................................................... 396
Appendix F Document Query Language Examples ...................................................... 401
Basic examples ............................................................................................... 401
The simplest format.................................................................................... 401
Using the WHERE clause ............................................................................ 402
Searching repeating properties in a WHERE Clause.................................. 402
Using aggregate functions........................................................................... 402
Using the GROUP BY clause ................................................................... 403
Using the HAVING clause....................................................................... 403
The ORDER BY clause ................................................................................ 403
Using the asterisk (*) in queries ................................................................... 404
Searching cabinets and folders ........................................................................ 404
Querying registered tables .............................................................................. 405
Querying virtual documents ........................................................................... 405
Determining the components ...................................................................... 406

Table of Contents
List of Figures
Figure 1. Sample virtual document ................................................................................... 129
Figure 2. Repository storage of Docbasic expressions for object types................................... 377
Figure 3. Repository storage of manually migrated Docbasic expressions............................. 378
Figure 4. Virtual document model ..................................................................................... 406

Table of Contents
List of Tables
Table 1. Valid ranges for floating point literals .................................................................... 14
Table 2. Default character string formats for dates............................................................... 15
Table 3. registered_table_1................................................................................................. 29
Table 4. Arithmetic operators............................................................................................. 32
Table 5. Valid comparison operators for DQL statements..................................................... 33
Table 6. Predicates for columns and single-valued properties............................................... 33
Table 7. Predicates for repeating properties ........................................................................ 34
Table 8. ALTER GROUP argument descriptions.................................................................. 45
Table 9. ALTER ASPECT argument descriptions................................................................. 47
Table 10. Syntax variations for full-text-indexing of aspect properties .................................... 49
Table 11. ALTER TYPE argument descriptions ..................................................................... 51
Table 12. Alter Type operations ........................................................................................... 54
Table 13. CHANGE...OBJECT argument descriptions............................................................ 68
Table 14. CREATE GROUP argument descriptions ............................................................... 72
Table 15. CREATE...OBJECT argument descriptions.............................................................. 74
Table 16. CREATE TYPE argument descriptions ................................................................... 77
Table 17. DB2 tablespace page sizes and associated maximum row lengths ............................ 81
Table 18. DELETE argument descriptions............................................................................. 91
Table 19. DELETE...OBJECT argument descriptions.............................................................. 93
Table 20. DROP GROUP syntax description ......................................................................... 97
Table 21. DROP TYPE argument descriptions....................................................................... 98
Table 22. EXECUTE argument descriptions .......................................................................... 99
Table 23. Administration methods by category for the EXECUTE statement ......................... 100
Table 24. GRANT argument descriptions ........................................................................... 105
Table 25. INSERT argument descriptions ........................................................................... 107
Table 26. REGISTER argument descriptions ....................................................................... 110
Table 27. REVOKE argument descriptions.......................................................................... 114
Table 28. SELECT argument descriptions ........................................................................... 116
Table 29. DQL standard hints ............................................................................................ 149
Table 30. UNREGISTER argument descriptions .................................................................. 153
Table 31. UPDATE argument descriptions.......................................................................... 155
Table 32. UPDATE...OBJECT argument descriptions ........................................................... 158
Table 33. Administration methods by category ................................................................... 166
Table 34. BATCH_PROMOTE arguments........................................................................... 171
Table 35. CHECK_CACHE_CONFIG arguments ................................................................ 174
Table 36. Properties in the CHECK_CACHE_CONFIG result object..................................... 175
Table 37. CHECK_RETENTION_EXPIRED arguments........................................................ 177

Table of Contents
Table 38. CHECK_SECURITY arguments........................................................................... 181
Table 39. CLEAN_DELETED_OBJECTS arguments ............................................................ 183
Table 40. DB_STATS arguments......................................................................................... 186
Table 41. Query result object properties for DB_STATS administration method..................... 186
Table 42. DELETE_REPLICA arguments ............................................................................ 188
Table 43. DO_METHOD arguments................................................................................... 192
Table 44. Properties of the query result object returned by DO_METHOD............................ 194
Table 45. DROP_INDEX arguments................................................................................... 199
Table 46. ESTIMATE_SEARCH arguments......................................................................... 201
Table 47. EXEC_SQL arguments ........................................................................................ 203
Table 48. EXPORT_TICKET_KEY arguments ..................................................................... 205
Table 49. GENERATE_PARTITION_SCHEME_SQL arguments........................................... 209
Table 50. GET_FILE_URL arguments................................................................................. 214
Table 51. GET_INBOX arguments...................................................................................... 216
Table 52. SysObject-related property names for GET_INBOX query result objects................. 217
Table 53. GET_LAST_SQL arguments ................................................................................ 219
Table 54. GET_PATH arguments........................................................................................ 221
Table 55. HTTP_POST arguments...................................................................................... 224
Table 56. Query result object properties for HTTP_POST..................................................... 225
Table 57. IMPORT_REPLICA arguments............................................................................ 228
Table 58. IMPORT_TICKET_KEY arguments...................................................................... 230
Table 59. LIST_RESOURCES arguments............................................................................. 233
Table 60. Collection properties for LIST_RESOURCES ........................................................ 233
Table 61. Collection properties for LIST_RESOURCES ........................................................ 234
Table 62. LIST_SESSIONS arguments................................................................................. 237
Table 63. Complete information returned by LIST_SESSIONS ............................................. 237
Table 64. Query result object properties for LIST_TARGETS................................................ 240
Table 65. LOG_OFF arguments.......................................................................................... 243
Table 66. MAKE_INDEX arguments .................................................................................. 245
Table 67. MARK_FOR_RETRY arguments.......................................................................... 249
Table 68. MIGRATE_CONTENT arguments ....................................................................... 251
Table 69. MIGRATE_TO_LITE arguments .......................................................................... 265
Table 70. MODIFY_TRACE arguments .............................................................................. 273
Table 71. MOVE_INDEX arguments .................................................................................. 275
Table 72. PARTITION_OPERATION arguments ................................................................. 278
Table 73. PING arguments ................................................................................................ 286
Table 74. PURGE_AUDIT arguments................................................................................. 287
Table 75. PUSH_CONTENT_ATTRS arguments ................................................................. 294
Table 76. RECOVER_AUTO_TASKS arguments ................................................................. 296
Table 77. REGISTER_ASSET arguments ............................................................................. 298
Table 78. Queue item property values set by REGISTER_ASSET ......................................... 299
Table 79. REGISTER_ASSET arguments ............................................................................. 300
Table 80. REORGANZE_TABLE arguments ....................................................................... 301

Table of Contents
Table 81. REPLICATE arguments....................................................................................... 303
Table 82. RESTORE_CONTENT arguments........................................................................ 306
Table 83. ROLES_FOR_USER arguments............................................................................ 308
Table 84. SET_APIDEADLOCK arguments ........................................................................ 310
Table 85. Valid operation names for SET_APIDEADLOCK.................................................. 311
Table 86. SET_CONTENT_ATTRS arguments..................................................................... 313
Table 87. Example settings for content metadata properties in content objects....................... 315
Table 88. SET_OPTIONS arguments .................................................................................. 317
Table 89. Trace options for SET_OPTIONS ......................................................................... 318
Table 90. SET_STORAGE_STATE arguments...................................................................... 320
Table 91. TRANSCODE_CONTENT arguments.................................................................. 324
Table 92. Queue item properties set by TRANSCODE_CONTENT ...................................... 325
Table 93. UPDATE_STATISTICS arguments ....................................................................... 327
Table 94. WEBCACHE_PUBLISH arguments ..................................................................... 329
Table 95. Valid arguments for ARGUMENTS ..................................................................... 329
Table 96. DQL basic query statements................................................................................ 334
Table 97. Default property values by datatype ................................................................... 336
Table 98. Repeating values without SPACEOPTIMIZE ........................................................ 337
Table 99. Repeating values with SPACEOPTIMIZE............................................................. 337
Table 100. Added repeating values with SPACEOPTIMIZE................................................... 337
Table 101. Predicates that test for NULL and default values .................................................. 338
Table 102. RETURN_RANGE argument descriptions............................................................ 356
Table 103. sorting_clause argument descriptions.................................................................... 357
Table 104. Example of object-based query results.................................................................. 360
Table 105. Example of row-based query results .................................................................... 360
Table 106. Comparison of logical reads with and without default result sets........................... 361
Table 107. Dmr_content object space savings........................................................................ 365
Table 108. IDQL command line arguments........................................................................... 368
Table 109. IDQL commands ................................................................................................ 369
Table 110. IAPI command line arguments ............................................................................ 372
Table 111. IAPI commands.................................................................................................. 374
Table 112. Arguments for the Docbasic expression migration methods .................................. 379
Table 113. dmc_SetJavaExprEnabled arguments................................................................... 381
Table 114. Docbasic operators supported by Java evaluation ................................................. 381
Table 115. Docbasic functions supported for Java evaluation ................................................. 382
Table 116. Docbasic functions not supported for Java evaluation ........................................... 386
Table 117. Docbasic constants supported for Java evaluation................................................. 387
Table 118. Constants not supported for Java evaluation ........................................................ 388
Table 119. Summary of FTDQL query rules.......................................................................... 393
Table 120. DQL reserved words........................................................................................... 397

Preface
This manual is the reference manual for Documentum’s Document Query Language, supported by
Content Server. It is a companion to EMC Documentum Content Server System Object Reference, EMC
Documentum Content Server Fundamentals Guide, EMC Documentum Administrator User Guide, and EMC
Documentum Content Server Distributed Configuration Guide.
Intended audience
This manual is written for application developers and system administrators and any others who
want to build a content or workgroup management application that uses DQL. It assumes that
you are familiar with the concepts of document processing, object-oriented programming, and
client-server applications. It also assumes working knowledge of SQL.
Conventions
This manual uses the following conventions in the syntax descriptions and examples.
Syntax conventions
Convention Identifies
italics A variable for which you must provide a value.
[ ] square brackets An optional argument that may be included only once
{ } curly braces An optional argument that may be included multiple times
| vertical line A choice between two or more options
Revision history
Revision date Description
April 2015 Updated the section Date storage and handling,
page 18.

Preface
Revision date Description
March 2015 Updated the table, Arguments , page 77.
February 2015 Initial publication.

Chapter 1
DQL Language Elements
This chapter describes the building blocks of a DQL statement, including:
• Literals, page 13, which describes the literal formats for the Documentum datatypes
• Special keywords, page 18, which describes the special keywords that you can use in DQL queries
• Functions, page 19, which describes the functions that you can use in DQL queries
• Predicates, page 32 , which describes the predicates that you can use in expressions in queries
• Logical operators, page 40 , which describes the logical operators supported by DQL
• DQL reserved words, page 42 , which cross-references you to a list of the words reserved in DQL
Literals
Literals are values that are interpreted by the server exactly as they are entered. Content Server
recognizes five types of literals:
• Integer literals, page 13
• Floating point literals, page 14
• Character string literals, page 14
• ID literals, page 15
• Date literals, page 15
Integer literals
An integer literal specifies any whole number and is expressed in the following format:
[+ | -] n
where n is any number between -2147483647 and +2147483647.
DQL does not support the negative integer value -2147483648 because this number is not supported
in a number of relational databases. If you enter this number, your results are unpredictable.

Floating point literals
A floating point literal specifies any number that contains a decimal point and is expressed in the
following format:
5.347
21.
0.45
.66
-4.12
A floating point literal can also be expressed in scientific notation. For example:
10.4e-6
or
-3.6E7
or
12e-3
DQL accepts either uppercase or lowercase in the notation.
If you assign an integer literal to a property that has a floating point datatype, the system
automatically converts the integer to a floating point number.
The underlying RDBMS determines the maximum and minimum values that you can assign as a
floating point literal. Table 1, page 14 lists the ranges for supported databases.
Note: Do not reset the decimal symbol at the operating system level to a comma. Doing so results in
incorrect execution of some Documentum Administration jobs.
Table 1. Valid ranges for floating point literals
RDBMS Range Significant digits
Oracle 1.0 x 10-129 to 9.99 x 10-129 15
DB2 1.0 x 10-307 to 1.798 x 10+308 15
MS SQL Server 1.7e-308 to 1.7e+308 15
Character string literals
Character string literals are strings of printable characters and are enclosed in single quotes. You
cannot place non-printable characters, such as line feeds or carriage returns, in a character string
literal. To include a single quote as part of the literal, include it twice. For example:
'The company's third quarter results were very good.'
The maximum length of a character string literal is determined by the maximum allowed by the
underlying RDBMS, but in no case will the maximum length exceed 1999 bytes.
If a property is defined as a string datatype, the maximum length of the character string literal
you can place in the property is defined by the property’s defined length. If you attempt to place a
longer value in the property, DFC will throw an exception. You can change the behavior and allow

DFC to truncate the character string value to fit the property by setting the dfc preference called
dfc.compatibility.truncate_long_values in dfc.properties.
ID literals
An ID literal specifies an object ID as a 16-character string enclosed in single quotes. ID literals are
generally used in DQL queries. For example:
SELECT "object_name" FROM "dm_document"
WHERE "r_object_id" = '099af3ce800001ff'
Note that you cannot use ID literals to assign object IDs to objects. Object IDs are assigned
automatically by the server when the object is created.
Date literals
A date literal specifies a date using the following syntax:
DATE('date_value[utc]' [,'pattern'])
date_value can be defined using any of the valid character string formats representing a date, or it can
be one of the keywords that represent dates.
If utc is included, Content Server assumes that the specified date_value is UTC time. The specification
of utc is not case sensitive.
There are some date formats that the server can interpret as more than one date. For example,
03/04/96 can be interpreted as March 4, 1996 or as April 3, 1996. To resolve this, some formats require
you to specify what pattern the server uses to interpret the date.
Valid dates range from 01/01/1753 (January 1, 1753) to 12/31/4712 (December 31, 4712).
The following paragraphs describe the character string formats and keywords you can use to specify
a date. When you use a character string format, you must enclose date_value in single quotes.
Default formats
Table 2, page 15 lists the character string formats for date_value that are accepted by default:
Table 2. Default character string formats for dates
Format Examples
mm/dd/[yy]yy DATE(’03/24/1989’) DATE(’4/7/1992’)
dd-mon-[yy]yy DATE(’4-Apr-1975’)
month dd[,] [yy]yy DATE(’January 1, 1993’)
mon dd [yy]yy DATE(’March 23 1990’)

Format Examples
the client’s localized short date format DATE(’30-11-1990’) (Assumes short date format
dd-mm-yyyy is defined on the client machine)
ANSI format (dow mon dd hh:mm:ss yyyy) No example
Although it is not required, you can specify a pattern for any of these formats except the short date
format and the ANSI format. You cannot specify a pattern when you enter a date using either
of those two formats.
When using the formats listed in Table 2, page 15, the following rules apply:
• It is not necessary to include leading zeros for months or days that are represented as a single digit.
• You can abbreviate a month’s name to three letters.
• If you enter only the year and not the century, the server uses the century defined by the current
date in the server machine. For example, 03/23/95 is interpreted as March 23, 1995 before the year
2000 and is interpreted as March 23, 2095 after the year 2000.
Short date formats
The server accepts a client’s localized short date format as a valid character string format as long as
the format contains only numeric characters. For example, dd-mmm-yy is not an accepted short
date format.
Windows, Solaris, and AIX client platforms provide a default short date format. The Windows
platform also lets you define your own default short date format. (Refer to the EMC Documentum
Content Server Administration and Configuration Guide for information about defining short date
formats.) For Linux clients, which do not have a default short date format, the server assumes the
default format is mm/dd/yy hh:mi:ss.
If the locally defined short date format conflicts with one of the default input formats, the locally
defined format takes precedence. For example, assume that the locally defined format is dd/mm/yyyy.
If a user wants to enter March 14, 1994 and enters 03/14/1994 (mm/dd/yyyy), the server interprets this
as the 3rd day of the 14th month of the year 1994 and returns an error because there is no 14th month.
If you use the pattern argument to specify a format for a date, that pattern takes precedence over
the short date format.
ANSI format
You can also specify the date using the ANSI format (dow mon dd hh:mm:ss yyyy). However, DQL
ignores the time fields. (To enter a date and time using the ANSI format, use DFC.)
Other character string formats
The following character string formats for date_value are also acceptable but require a pattern
argument:
[dd/]mm/[yy]yy[hh:mi:ss]
[yy]yy/mm[/dd] [hh:mi:ss]

[mon-][yy]yy [hh:mi:ss]
month[,] [yy]yy [hh:mi:ss]
If you do not enter the time (hh:mi:ss), the server assumes the time is 00:00:00. If you do not enter the
day or month, the server assumes the first day of the month or the first month of the year, respectively.
For example, suppose the pattern argument is [dd/]mm/[yy]yy [hh:mi:ss]. The following date literal is
interpreted as April 1, 1996 00:00:00 (assuming the current century is 1900).
DATE('04/96','mm/yy')
When you specify date_value, you can use any character except a space or an alphanumeric character
as the delimiter. You must use the same delimiter to separate all elements of the date or time. For
example, if you use a forward slash to separate dd and mm, you must use a forward slash to separate
mm and yy. Similarly, if you use a period to separate hh and mi, you must use a period to separate
mi and ss.
The delimiter that you use in the date can be different from the delimiter that you use in the time
portion. For example, the following is valid:
'23-04-96 12.32.04'
It is not necessary to use the same delimiter for the date_value and pattern arguments. For example,
the following function call is valid:
DATE('23-04-96 12.32.04','dd/mm/yy hh:mi:ss')
Date literal keywords
Four keywords represent dates. They are:
• TODAY, which returns the current date in UTC. The time defaults to 00:00:00. For example,
SELECT "supervisor_name", "object_name" FROM "dm_workflow"
WHERE "r_start_date" <= DATE(TODAY)
AND "r_runtime_state"=1
ORDER BY 1
• NOW, which returns the current date and time. For example,
SELECT "supervisor_name", "object_name" FROM "dm_workflow"
WHERE "r_start_date" <= DATE(NOW) AND "r_runtime_state"=1
ORDER BY 1
• YESTERDAY, which returns the current date minus one day in UTC. The time defaults to 00:00:00.
For example,
SELECT * FROM dm_document
WHERE "r_creation_date" >= DATE(YESTERDAY)
AND
"r_creation_date" <= DATE(TODAY)
AND
ANY "authors" IN (john,henry,jane)
• TOMORROW, which returns the current date plus one day in UTC. The time defaults to 00:00:00.
For example,
SELECT "r_act_name", "object_name", "process_id"
FROM "dm_workflow"
WHERE ANY "r_pre_timer" >= DATE(TOMORROW)
ORDER BY 2

Date output formats
By default, the server uses the client’s localized short date format to output dates to the client. If the
client’s format represents years using two digits, dates in the current century are displayed using
two digits. However, to avoid ambiguity, years in other centuries are displayed using all four digits
(1834 or 1792).
The default short date formats, by platform, are:
• Windows: The default short date format is the format specified in the regional settings accessed
through the Control Panel dialog box.
• UNIX/Solaris and UNIX/AIX: The default short date format is the format defined by the machine’s
locale. (Locale is set by the UNIX setlocale command. The EMC Documentum Content Server
Administration and Configuration Guide for information about setting this value.
• UNIX and Linux: The format is assumed to be mm/dd/yy hh:mi:ss.
Note: The session config property r_date_format contains the date format that the server returns
to a client for a given session.
If the date is entered as NULLDATE, then the output is the string NULLDATE.
You can override the default format. If you retrieve a property has a Date datatype, you can include a
pattern argument that defines the format in which the date is returned.
Dates that are included in error messages or a dump file are displayed in ANSI date format.
Date storage and handling
How dates are stored in the repository and handled during transmission between Content Server and
a client application is not affected by the input or output format chosen for use.
By default, when communicating with a Documentum Version 6 client, a Content Server connecting
to a new repository assumes that all date values in a query are expressed as server local time and
stores all dates in the repository normalized to UTC time based on that assumption. This is the
recommended format for date storage. Content Server and DFC will make the necessary time zone
adjustments when dates are sent and received. This behavior can be changed so that Content Server
stores the dates in server local time, rather than in UTC time. However, this is not recommended.
When communicating with a client version earlier than Documentum Version 6, Content Server
assumes that the date values in queries are server local time.
For complete information about date storage and handling, refer to the EMC Documentum Content
Server Administration and Configuration Guide.
Special keywords
DQL includes some special keywords for use in queries. These keywords have special meanings for
Content Server when used in a DQL query. (They cannot be used in DFC methods.) The keywords are:
• USER, which identifies the current user.

You can use USER in comparison operations in queries. For example,
SELECT "process_id", "object_name" FROM dm_workflow
WHERE supervisor_name=USER
• TRUE and FALSE, which represent the Boolean true and false.
You can use TRUE and FALSE in comparison operations involving any property having a
BOOLEAN datatype. For example,
SELECT * FROM "dm_user"
WHERE "r_is_group" = TRUE
• DM_SESSION_DD_LOCALE, which represents the data dictionary locale most appropriate for
the client’s session locale.
This keyword is particularly useful if the client session locale has no exact match recognized by
the server. If you use this keyword, the server will return the information from the best matching
locale. For example, suppose the server recognizes three locales, English (en), French (fr), and
Spanish (es) and the client session locale is French Canadian (fr_cn). Suppose the client issues
the following query:
SELECT type_name, label_text from dmi_dd_type_info where
nls_key='fr_cn'
The query returns nothing because the locale fr_cn is not recognized by the server. However, the
following query returns the information from the French locale:
SELECT type_name, label_text from dmi_dd_type_info where
nls_key=DM_SESSION_DD_LOCALE
The server checks whether the locale identified in the client’s session locale, fr_cn, exists in the
repository. Since fr_cn isn’t found, the server attempts to find a good match among the existing
locales. It determines that fr (French) is a good match and uses that locale to execute the query.
If a good match isn’t found, the server uses the default locale to execute the query.
Functions
Functions are operations on values. DQL recognizes three groups of functions and three unique
functions:
• Scalar functions, page 20
Scalar functions operate on one value and return one value.
• Aggregate functions, page 22
Aggregate functions operate on a set of values and return one value.
• Date functions, page 24
Date functions operate on date values.
• The CHECK_ACL function, page 28
The CHECK_ACL function enforces security for queries to registered tables.
• The ID function, page 30

The ID function, a unique function recognized by DQL, is used in the FOLDER and CABINET
predicates and in the IN DOCUMENT and IN ASSEMBLY clauses.
• The MFILE_URL function, page 31
The MFILE_URL function returns URLs to content files and renditions in particular format.
Scalar functions
The scalar functions are:
• ASCII
• BITAND
• BITCLR
• BITSET
• UPPER
• LOWER
• SUBSTR
ASCII
The ASCII function takes one argument and returns the ASCII code value of the first character of
the argument. The argument can be a character string literal, a property that has a character string
datatype, or the SUBSTR function. The underlying database provides the ASCII function, so there
may be slight variations in behavior, depending on the database.
The following example returns the number of documents whose owner name begins with a lower
case letter b (ASCII value=98):
select count(*) from dm_document where ascii(owner_name)=98
BITAND, BITCLR, BITSET
The three functions BITAND, BITCLR, and BITSET can appear in the WHERE clause of the DELETE,
SELECT, and UPDATE DQL statements, and as a SELECT list item in the SELECT DQL statement.
These are bitwise functions that take two integer expressions as arguments and return an integer.
They are typically used to test bit values stored in integer properties.
To understand how the return values of the three functions are computed, imagine each input
integer as its binary representation (for example 5 is 101 binary and 9 is 1001 binary). Examine the
corresponding bits, then AND, clear, or OR each pair (use leading zeros if necessary), and use that
value for the corresponding bit of the integer return value.

The BITAND function ANDs the bit values of the two input arguments and returns that result.
For example, BITAND(5,9) returns 1, since 101 binary and 1001 binary ANDed together bitwise is
0001 binary.
The BITCLR function sets the bit in the result to zero (clears it), if the corresponding bit in the second
parameter is one, otherwise it uses the value of that corresponding bit in the first parameter. For
example, BITCLR(5,9) returns 4, since 101 binary cleared bitwise by 1001 binary is 0100 (the bit in
the lowest order position of 9 clears the bit in the lowest order position of 5). As another example,
BITCLR(9,5) returns 8 (1001 binary cleared bitwise by 101 binary is 1000 binary). Notice that the
order of the arguments affects the return result.
The BITSET function ORs the bit values of the two input arguments and returns the ORed value. For
example, BITSET(5,9) returns 13, since 101 binary ORed with 1001 binary is 1101 binary.
For example, the following statement returns the user names of all users that have the lowest order
bit of the user_privileges property set to one. These users have the privilege of creating types (this
query will not return all the users who can create types, since sysadmins and superusers can also
create types).
SELECT "user_name" FROM "dm_user"
WHERE BITAND("user_privileges",1) = 1
The following statement returns the user names of all users that have the bit set of the user_privileges
property that represents 16 in binary. These should be all the superusers.
WHERE BITAND("user_privileges",16) = 16
UPPER
The UPPER function takes one argument and returns the uppercase of that value. The value supplied
as the argument must be a character string or a property that has a character string datatype.
For example, the following statement returns the object names of all documents that have the word
government in their title. Because LIKE returns only exact matches, the UPPER function is used to
ensure that all instances are found, regardless of the case (upper, lower, or mixed) in which the
word appears in the title.
WHERE UPPER("title") LIKE '%GOVERMENT%'
LOWER
The LOWER function takes one argument and returns the lowercase of that value. The value supplied
as the argument must be a character string or a property that has a character string datatype.
For example, the following statement returns the subjects in lowercase of all documents owned
by regina:
SELECT LOWER("subject") FROM "dm_document"
WHERE "owner_name" = 'regina'

SUBSTR
The SUBSTR function returns some or all of a particular string. Its syntax is:
substr(string_value,start[,length])
The value of string_value can be a literal string or a column or property name. If you specify a column
or property, the server uses the value in that column or property as the string value. The property you
specify can be either a single-valued property or a repeating property.
The start argument identifies the starting position, in the specified string, of the substring you want
returned. Positions are numbered from the left, beginning at 1. (If you specify 0 as the start, the server
automatically begins with 1.) The argument must be an integer.
The length argument defines how many characters should be returned in the substring. Length is
an optional argument. If you do not specify a length, the default behavior of the function differs
depending on the RDMBS you are using. For Oracle and DB2, the default is the entire string value
or the full width of the column if a property or column is specified. For all other databases, the
function returns 1 character.
For example, suppose you have a document subtype called purchase_order, which has a property
called order_no. The values in this property are 10-digit numbers, with the last three digits
representing a specific sales region. The following statement returns all documents for which the last
three digits in the order_no property are 003:
SELECT "r_object_id","order_no" FROM "purchase_order"
WHERE SUBSTR("order_no",8,3) = '003'
You can also use the SUBSTR function with the SELECT statement. For example:
SELECT SUBSTR("emp_name",1,4) AS short_name FROM "employee"
You must use the AS clause option to rename the query result property that holds the return value of
the SUBSTR function. If you do not, the server cannot return the result of the function.
If you specify a repeating property, the function returns one value for each value in the property.
You cannot use the SUBSTR function in assignment statements.
Additionally, you cannot specify SUBSTR in a LIKE predicate. Refer to Pattern matching with LIKE,
page 36, for more information about using LIKE.
Aggregate functions
The five aggregate functions are:
• COUNT
• MIN
• MAX
• AVG
• SUM

COUNT
The COUNT function counts values. The syntax is:
COUNT ([DISTINCT] name | *)
The name argument must identify a property or column. You can count all values for the property or
column or you can include the DISTINCT keyword to count the number of unique values.
Using an asterisk directs the server to count all items that match specified criteria. For example, the
following statement counts all documents that belong to the user named grace:
SELECT COUNT(*) FROM "dm_document"
WHERE "owner_name" = 'grace'
MIN
The MIN function returns the minimum value in a given set of values. The syntax is:
MIN(DISTINCT name | [ALL] value_expresssion)
The DISTINCT name argument directs the server to first select all distinct values from the set (no
duplicates) and then return the minimum value. name can be a property or column name. Note that
for this function, the DISTINCT name option has little meaning.
[ALL] value_expression directs the server to return the minimum value found in the set of values
specified by the value_expression argument. value_expression can be any valid numeric expression or a
property or column name. The keyword ALL is optional; whether it is specified or omitted, all of
the values in the set are evaluated.
For example, assuming that rental_charges is a user-defined object type, the following statement
returns the minimum rent charged for a two-bedroom apartment:
SELECT MIN("charge") FROM "rental_charges"
WHERE "style" = 'apt' AND "bedroom" = 2
MAX
The MAX function returns the maximum value in a given set of values. The syntax is:
MAX(DISTINCT name | [ALL] value_expresssion)
The DISTINCT name argument directs the server to first select all distinct values from the set (no
duplicates) and then return the maximum value. name can be a property or column name. Note that
for the MAX function, this option has little meaning.
[ALL] value_expression directs the server to return the maximum value found in the set of values
specified by the value_expression argument. value_expression can be any valid numeric expression or a
property or column name. The keyword ALL is optional; whether it is specified or omitted, all of
the values in the set are evaluated.
returns the maximum rent charged for a two-bedroom apartment:
SELECT MAX("charge") FROM "rental_charges"

AVG
The AVG function returns an average. The syntax is:
AVG(DISTINCT name | [ALL] value_expression)
The DISTINCT name option directs the server to select all distinct values from the set (no duplicates)
and return the average of those distinct values. name can be a property or column name.
[ALL] value_expression directs the server to return the average value found in the set of values
specified by the value_expression argument. The value of value_expression can be any valid numeric
expression or a property or column name. Use the optional keyword ALL to include all of the values
in the set in the operation.
returns the average rent charged on a two-bedroom apartment:
SELECT AVG("charge") FROM "rental_charges"
SUM
The SUM function returns a total. The syntax is:
SUM(DISTINCT name | [ALL] value_expression)
The DISTINCT name option directs the server to select all distinct values from the set (no duplicates)
and return the sum of those distinct values. name can be a property or column name.
[ALL] value_expression directs the server to return the sum of the values in the set of values specified
by the value_expression argument. The value of value_expression can be any valid numeric expression
or a property or column name. Use the optional keyword ALL to include all of the values in the
set in the operation.
For example, assuming that rent_records is a user-defined object type that contains payment records
of tenants, the following statement provides a total of the rents paid in May:
SELECT SUM("rent_amt") FROM "rent_records"
WHERE "mon" = 'may' AND UPPER("paid") = 'Y'
Date functions
The six date functions are:
• DATEDIFF (refer to DATEDIFF , page 25)
• DATEADD (refer to DATEADD, page 26)
• DATEFLOOR (refer to DATEFLOOR, page 26)
• DATETOSTRING (refer to DATETOSTRING, page 26)

• DATEFLOOR_LOCAL (refer to DATEFLOOR_LOCAL, page 27
• DATETOSTRING_LOCAL (refer to DATETOSTRING_LOCAL, page 28
For information about DATE(), refer to Date literals, page 15.
Note: When you include these date functions in the selected values list of a SELECT statement, you
must use the AS clause to assign a name to the column representing the returned values. For example:
SELECT DATEFLOOR(month,"r_creation_date") AS Created...
SELCT DATETOSTRING("r_creation_date",'dd/mm/yyyy hh:mi:ss') AS date_val...
DATEDIFF
The DATEDIFF function subtracts two dates and returns a number that represents the difference
between the two dates. The syntax is:
DATEDIFF(date_part, date1, date2)
The date_part argument defines the units in which the return value is expressed. Valid values are
year, month, week, and day.
The date1 and date2 arguments specify the dates to be subtracted. date1 is subtracted from date2. You
can specify the dates as date literals or the names of single-valued properties with a date datatype.
For example, the following statement uses the DATEDIFF function to return all tasks that were
started a month or more late:
SELECT "task_number", "supervisor_name"
FROM dm_tasks_queued
WHERE
DATEDIFF(month,"plan_start_date", "actual_start_date")>=1
This next example returns all tasks that were started more than one week ago:
SELECT "task_number", "r_task_user"
FROM dm_tasks_queued
WHERE DATEDIFF(week, "actual_start_date", DATE(TODAY))>=1
If the repository is using Oracle or DB2, the return value is a floating point number.
If the repository is using DB2, the server assumes 365 days per year and 30.42 days per month (the
mean number of days in a month). These assumptions can cause return values that differ from the
expected value. To illustrate, the following example, which asks for the number of days between
March 1, 1996 and Feb 1, 1996, returns 30.42 instead of 28:
DATEDIFF(day, date('03/01/1996 0:0:0'),
date('02/01/1996 0:0:0'))
If the repository is using MS SQL Server, the return value is an integer for all units except day. If you
specify day in the function, the return value is a floating point.
The MS SQL Server implementation round up if the difference is one half or greater of the specified
unit. The implementations round down if the difference is less than one half of the specified unit. To
illustrate, the following example, which asks for the difference between March 1, 1996 and July 1,
1996 expressed as years, returns 0 because the difference is only 4 months.
DATEDIFF(year,date('03/01/1996'),date('07/01/1996'))

DATEADD
The DATEADD function adds a number of years, months, weeks, or days to a date and returns
the new date. The syntax is:
DATEADD(date_part, number, date)
Thedate_part argument defines the units that are being added to the specified date. Valid date parts
are year, month, week, and day.
The number argument defines how date_part values are being added to the date. For example, the
following statement uses the DATEADD function to obtain all tasks that started more than a week
ago but are not yet finished:
SELECT "task_number", "supervisor_name" FROM dm_tasks_queued
WHERE DATEADD(week, 1, "actual_start_date") < DATE(TODAY)
AND "actual_completion_date" IS NULLDATE
AND NOT("actual_start_date" IS NULLDATE)
DATEFLOOR
The DATEFLOOR function rounds a given date down to the beginning of the year, month, or day
in UTC. The syntax is:
DATEFLOOR(date_part,date)
The date argument is the name of a date property. The function rounds the value of the property
down to the beginning of the unit specified as the value of date_part. Valid date_part values are year,
month, and day.
For example, suppose that a document’s r_creation_date property has a value of March 23, 1996 at
9:30:15 am. Using the DATEFLOOR function on this property returns these values:
Specifying Returns
DATEFLOOR(year,"r_creation_date") January 1, 1996 at 00:00:00
DATEFLOOR(month,"r_creation_date") March 1, 1996 at 00:00:00
DATEFLOOR(day,"r_creation_date") March 23, 1996 at 00:00:00
DATETOSTRING
The DATETOSTRING function returns a date as a character string in UTC in a particular format. The
syntax is:
DATETOSTRING(date,'format')

The date argument is the name of a date property. The format argument defines how you want the
character string formatted. The format can be any of the valid character string input formats for date
literals, subject to the following RDBMS restrictions and qualifications:
• MS SQL Server does not support month. If you specify a format that contains month, such as
month dd yyyy, the date is returned in the default short date format.
• Oracle and DB2 always return the month name as a fixed-length 9-character string. This means
that if the name of the month is shorter than 9 characters, the string value is padded.
Suppose that a document’s r_creation_date property has a value of May 14, 1995. Here are some
examples of the values returned by the DATETOSTRING function using this date and a variety
of formats:
Specifying Returns
DATETOSTRING("r_creation_date",'dd-mon-yy') 14-May-95
DATETOSTRING("r_creation_date",'mm/dd/yy')
05/14/95
DATETOSTRING("r_creation_date",'month yy') May 95
DATEFLOOR_LOCAL
The DATEFLOOR_LOCAL function rounds a date down to the beginning of an indicated granularity
(day, month, or year) in the server local time. The syntax is:
DATEFLOOR_LOCAL(date_part,date)
The date argument is the name of a date property. It indicates a date literal, name of a date attribute,
or a date keyword. The date_part argument indicates the level of granularity. The valid date_part
values are year, month, and day.
For example, suppose that the r_creation_date property of a document has a value of October
26, 2014 at 9:30:15 AM and the time zone of the Content Server machine is GMT+5:30. Using the
DATEFLOOR_LOCAL function on this property returns these values:
Specifying Returns
DATEFLOOR_LOCAL(year,"r_creation_
date")
1/1/2014 12:00:00 AM (January 1, 2014
00:00:00)
DATEFLOOR_LOCAL(month,"r_creation_
date")
10/1/2014 12:00:00 AM (October 1,
2014 00:00:00)
DATEFLOOR_LOCAL(day,"r_creation_
date")
10/26/2014 12:00:00 AM (October 26,
2014 00:00:00)

DATETOSTRING_LOCAL
The DATETOSTRING_LOCAL function formats a date value to a character string in the server local
time. This function always expresses the result in the server local time. The syntax is:
DATETOSTRING_LOCAL(date,'format')
The date argument is the name of a date property. The format can be any of the valid character string
input formats in which you want to see the date. If the format is not supported, result is displayed in
the default format along with a warning message.
For example, suppose that the r_creation_date property of a document has the value of October 26,
2014 at 9:30:15 AM. Here are some examples of the values returned by the DATETOSTRING_LOCAL
function using this date and a variety of formats:
Specifying Returns
DATETOSTRING_LOCAL("r_creation_date",'dd-mon-yy') 26-Oct-14
DATETOSTRING_LOCAL("r_creation_date",'mm/dd/yyyy
hh:mi:ss') 10/26/2014 09:30:15
DATETOSTRING_LOCAL("r_creation_date",'month, yy') October, 14
The CHECK_ACL function
The CHECK_ACL function provides similar security enforcement for queries against registered
tables as DQL does for queries against sysobjects. When a non-superuser issues a DQL statement
against dm_sysobject or its subtypes, an ACL security-checking clause is appended at the end of the
SQL statement issued against the database. The CHECK_ACL function is used similarly to enforce
security with registered tables. Since there is no security-checking clause added to DQL statements
run against registered tables, the CHECK_ACL function must be ANDed with the other qualifications
in the statement. The registered table must contain a column that has the ACL domain and the ACL
name for each table row to use for the security check. The syntax of the function is:
CHECK_ACL('acl_domain_column','acl_name_column'
[,'user_name'[,'active_dynamic_groups_list']]
)
The parameters of the function:
• acl_domain_column—the name of the column that contains the ACL domains in the table. This
column can be qualified with a table name or its alias. If not qualified, then it must be unique
among the registered tables in the FROM clause.
• acl_name_column—the name of the column that contains the ACL names in the table. This column
can be qualified with a table name or its alias. If not qualified, then it must be unique among the
registered tables in the FROM clause.

• user_name—an optional parameter that specifies the user against whose security is checked. If not
provided, the session user is used.
• active_dynamic_groups_list—an optional parameter that specifies the list of active dynamic groups
which the specified user is a member of.
Examples
For example, if we have a registered table, registered_table_1, like the following:
Table 3. registered_table_1
loan_r_object_id loan_acl_domain loan_acl_name
0b00000000000001 domainA car_loan_acl
0b00000000000002 domainA car_loan_acl
0b00000000000003 domainB home_loan_acl
0b00000000000004 domainB home_loan_acl
We can issue the following DQL statement to enforce security for the user, user_1:
SELECT <any_attributes_in_the_table>
FROM registered_table_1
WHERE loan_r_object_id = '<some_id>' AND
CHECK_ACL('loan_acl_domain', 'loan_acl_name', 'user_1')
The ACL domain and ACL name columns are not qualified because there is only one registered table
in the FROM clause. If you have two registered tables to join and use an ACL, then you must provide
the table name if the column names are not unique.
The following example assume that registered_table_2 also has columns for loan_acl_domain and
loan_acl_name. In this example, you are not required to qualify the ACL domains and names if
they are unique among the two registered tables.
SELECT <any_attributes_in_either_table>
FROM registered_table_1, registered_table_2
WHERE (<predicate_for_table_1_or_2>) AND
CHECK_ACL('registered_table_1.loan_acl_domain',
'registered_table_1.loan_acl_name', '<user_name>')
AND
CHECK_ACL('registered_table_2.loan_acl_domain',
'registered_table_2.loan_acl_name', '<user_name>')
If aliases are used in the FROM clause, then you can use the alias as shown in the following example:
FROM registered_table_1 r1, registered_table_2 r2
WHERE (<predicate_for_table_1_or_2>) AND
CHECK_ACL('r1.loan_acl_domain',
'r1.loan_acl_name', '<user_name>')
AND
CHECK_ACL('r2.loan_acl_domain',
'r2.loan_acl_name', '<user_name>')
If registered_table_1 has another set of ACL domain and name (for example, customer_acl_domain,
customer_acl_name), you can issue the following DQL statement to enforce security by the two ACLs:

WHERE (<predicate_for_table_1>) AND
CHECK_ACL('loan_acl_domain',
'loan_acl_name', '<user_name>')
AND
CHECK_ACL('customer_acl_domain',
'customer_acl_name', '<user_name>')
In the preceding example, the two ACL predicates are applied to every attribute in the same row. In
the most likely use case, where there is more than one set of ACL domain and name in a registered
table, each set of ACL domain and name is relevant to a subset of attributes in the same row. That is,
each set of ACL domain and name should be used to enforce security for a subset of, but not all of,
the attributes in a row. From the preceding example, if registered_table_1 has other attributes like
loan_amount from a loan type and customer_account from a customer type, the following SELECT
statement only returns rows that satisfy both ACL predicates:
SELECT customer_account, loan_amount
WHERE (<predicate_for_table_1>) AND
CHECK_ACL('loan_acl_domain',
'loan_acl_name', '<user_name>')
AND
CHECK_ACL('customer_acl_domain',
'customer_acl_name', '<user_name>')
For this example, customer_account is returned when the ACL domain and name
(customer_acl_domain, customer_acl_name) allow it, regardless of whether the other ACL domain
and name (loan_acl_domain, loan_acl_name) allow it or not. Similarly, loan_amount is returned
when the ACL domain and name (loan_acl_domain, loan_acl_name) allow it, regardless of the values
in customer_acl_domain and customer_acl_name.
The ID function
The ID function is used to identify a folder or cabinet whose contents you want to search. You can
use the ID function in FOLDER and CABINET predicates and IN DOCUMENT and IN ASSEMBLY
clauses. The syntax is:
ID('object_id')
The object_id argument identifies the folder, cabinet, or virtual document you want to search. The
object must reside in the current repository.
Use the literal 16-character representation of the object’s ID. For example, the following statement
returns all documents in the specified folder with titles containing the characters Pond:
SELECT * FROM "dm_document"
WHERE FOLDER (ID('099af3ce800001ff'))
AND "title" LIKE '%Pond%'

The MFILE_URL function
The MFILE_URL function returns URLs to content files or renditions associated with a document.
Only those objects on which the user has at least Read permission are returned if MFILE_URL
is included in a selected values list.
The syntax of MFILE_URL is:
MFILE_URL('format',page_no,'page_modifier')
The arguments are ANDed together to determine which URLs are returned. Only the format
argument is required. page_no and page_modifier are optional.
The format argument restricts the returned URLs to those pointing to files in the specified format.
Enclose the format value in single quotes. If you specify format as an empty string, Content Server
takes the format value from the returned object’s a_content_type property.
The page_no argument restricts the returned URLs to those pointing to content files associated with
given page number. If you do not include page_no, the value defaults to 0, which represents the first
content file associated with an object. To indicate that all content pages should be considered, define
page_no as -1. If you define page_no as -1, Content Server returns all URLs that satisfy the other
arguments regardless of the page with which the content files are associated.
The page_modifier argument restricts the returned URLs to those pointing to content files that
have the specified page_modifier. The page_modifier argument must be enclosed in single quotes.
If you specify a value for page_modifier (other than an empty string), all the returned URLs point
to renditions, as content files for primary pages have no page modifier. To return URLs for content
files that have no page modifier, define page_modifier as an empty string. (Content Server sets the
page_modifier property to an empty string by default when a user adds a content file to an object
or saves a rendition without a page modifier.) If you don’t include page_modifier, Content Server
returns the URLs that satisfy the other arguments regardless of whether their content files have a
page modifier or not.
Examples
The following statement returns the URLs to the jpeg_th renditions of the first content pages of
documents owned by ronaldh that have a page modifier of image1.
SELECT MFILE_URL('jpeg_th',0,'image1')
FROM "dm_document" WHERE "object_owner"='ronaldh'
The following example returns the owner name and object ID of all documents that have the subject
prairie chickens. For each returned document, it also returns URLs to the primary content files
associated with the document.
SELECT owner_name,r_object_id,MFILE_URL('',-1,'')
FROM "dm_document" WHERE "subject"='prairie_chickens'

Predicates
Predicates are used in WHERE clauses to restrict the objects returned by a query. The WHERE
clause defines criteria that each object must meet. Predicates are the verbs within the expression
that define the criteria.
For example, the following statement returns only documents that contain the value approved
in their keywords property:
SELECT "r_object_id", "object_name", "object_owner"
FROM "dm_document"
WHERE ANY "keywords" = 'approved'
In this example, the equal sign (=) is a predicate. It specifies that any object returned must have a
keywords property value that equals (matches) the specified word (in this case, the word approved).
DQL recognizes these predicates:
• Arithmetic operators (refer to Arithmetic operators, page 32)
• Comparison operators (refer to Comparison operators, page 32)
• Column and property predicates (refer to Column and property predicates, page 33)
• SysObject predicates (refer to SysObject predicates, page 38)
Arithmetic operators
Arithmetic operators perform arithmetic operations on numerical expressions. Table 4, page 32 lists
the arithmetic operators that you can use as predicates.
Table 4. Arithmetic operators
Operator Operation
+ Addition
– Subtraction
/ Division
* Multiplication
** Exponentiation
Comparison operators
Comparison operators compare one expression to another. Table 5, page 33 lists the comparison
operators that can be used as predicates.

Table 5. Valid comparison operators for DQL statements
Operator Relationship
= Equal
>= Greater than or equal to
<= Less than or equal to
> Greater than
< Less than
<> Not equal
!= Not equal
Column and property predicates
Column and property predicates let you compare values in a registered table’s columns or an object’s
properties to defined criteria. These predicates are divided into two groups. One group is used only
when the search criterion specifies a column or single-valued property. The other group is used when
the search criteria specifies a repeating property. Before you use the predicates related to NULL
values, please read NULLs, default values, and DQL, page 335.
Predicates for columns and single-valued properties
The predicates in this group allow you to compare a value in a table column or single-valued property
to defined criteria. Table 6, page 33 lists the predicates in this group.
Table 6. Predicates for columns and single-valued properties
Predicate Description
IS [NOT] NULL Determines whether the property is assigned a value. This
predicate is useful only for registered table columns.
IS [NOT] NULLDATE Determines whether the property is assigned a null date.
IS [NOT] NULLSTRING Determines whether the property is assigned a null string.
IS [NOT] NULLID Determines whether the property is assigned a null ID value.
(Introduced in release 6.6)
IS [NOT] NULLINT Determines whether the property is assigned a null integer
value.
[NOT] LIKE pattern [ESCAPE
character]
Determines whether the property is like a particular pattern.
(Refer to The ESCAPE character, page 37 for information
about the ESCAPE clause option.)
[NOT] IN value_list Determines whether the property is in one of a particular list
of values.

[NOT] EXISTS (subquery) Determines whether the property matches a value found by
the subquery.
comparison_op SOME (subquery)
comparison_op ANY (subquery)
Determines whether the comparison between the property
value and the results of the subquery evaluate to TRUE in
any case.
comparison_op ALL (subquery) Determines whether the comparison between the property
value and the results of the subquery evaluate to TRUE in all
cases.
For example, the following statement selects all documents that have a title that starts with the
word Breads:
WHERE "title" LIKE 'Breads%'
This next example selects all workflows that are supervised by one of the users named in the predicate:
SELECT "r_object_id", "supervisor_name" FROM "dm_workflow"
WHERE "supervisor_name" IN ('carrie','davidk','holly')
ORDER BY 2
Predicates for repeating properties
The predicates for repeating properties let you compare the values in repeating properties to some
defined criteria. The basic syntax for repeating property predicates is:
[NOT] [ANY] predicate
The ANY keyword must be used whenever an expression references a repeating property predicate
unless the query also includes the DQL hint, ROW_BASED. If ROW_BASED is included in the query,
it is not necessary to include the ANY keyword with repeating property predicates.
Table 7, page 34 lists the predicates in this group.
Table 7. Predicates for repeating properties
attr_name [NOT] LIKE pattern
[ESCAPE character]
Evaluates to TRUE if any value of the repeating property is
[not] like a particular pattern. (Refer to The ESCAPE character,
page 37 for information about the optional ESCAPE clause.)
attr_name IN (value_list) Evaluates to TRUE if any value of the property matches a
value in the value_list. Value_list is a comma-separated list of
values.
[IN|EXISTS] attr_name IN
(subquery)
Evaluates to TRUE if any value of the property matches a
value returned by the subquery.
IN and EXISTS are database hints that may enhance
performance by changing the structure of the generated SQL
query. Refer to Appendix A, Using DQL Hints, for more
information about these optional hints.

attr_name IS [NOT] NULL Evaluates to TRUE if any value of the property is [not] null.
attr_name IS [NOT] NULLDATE Evaluates to TRUE if any value of the property is [not] a
nulldate.
attr_name IS [NOT] NULLSTRING Evaluates to TRUE if any value of the property is [not] a null
string.
attr_name IS [NOT] NULLID Evaluates to TRUE if any value of the specified repeating
property is [not] a null ID value. (Introduced in release 6.6).
attr_name IS [NOT] NULLINT Evaluates to TRUE if any value of the specified repeating
property is [not] a null integer value.
attr_name comparison_op
value_expression
Evaluates to TRUE if the comparison operation is TRUE for
any value of the property.
For example, the following statement returns the object names of all documents with an author
whose name begins with a letter in the first half of the alphabet:
WHERE ANY "authors" <= 'M'
You can use logical operators to build more complex predicate expressions, such as:
[NOT] ANY predicate AND|OR [NOT] ANY predicate {AND|OR [NOT] ANY predicate}
For example, the following statement selects all documents that have Ingrid as an author and
a version label of 1.0:
SELECT "r_object_id", "object_name" FROM "dm_document"
WHERE ANY "authors" = 'Ingrid' AND
ANY "r_version_label" = '1.0'
The predicate statement returns all objects that meet the specified criteria regardless of where the
qualifying value is found in the property’s value list. For example, look at the following statement:
WHERE ANY "authors" IN ('jeanine','harvey') AND
ANY "keywords" = 'working'
This statement returns the name of all documents with either jeanine or harvey as a value anywhere
in the authors property values list and with the keyword working as a value in any position in the list
of values in the keywords property.
In some circumstances, you may want only objects for which the specified values are found in the
same respective positions. For example, assume that the search finds an object for which jeanine is the
value at position 3 in the authors property. You want the name of this object only if the keyword
working is in position 3 of the keywords property.
You can impose this restriction by enclosing the predicate expression in parentheses:
[NOT] ANY (predicate AND|OR [NOT]predicate {AND|OR [NOT]predicate})
For example, the statement below returns the names of documents for which either jeanine and
working or harvey and working (or both) are values of the authors and keywords properties,
respectively, and occupy corresponding positions in the properties.
WHERE ANY ("authors" IN ('jeanine','harvey')
AND "keywords" = 'working')

Notes:
• To return values at a matching index position level, the keyword ANY is outside the parentheses.
• You cannot specify an index position at which to look for values. You can specify only that the
values be found in the same respective positions.
• Using the logical operator OR returns the same results as using the basic syntax, because the
server returns the object if either predicate is true.
For more information and examples of querying repeating properties, refer to Repeating properties
in queries, page 339.
Pattern matching with LIKE
The [NOT] LIKE predicate for both single-valued and repeating properties lets you identify a pattern
to match the property value. This pattern is specified as a character string literal. For example, the
following predicate returns all objects whose subject contains the word Cake:
subject LIKE '%Cake%'
When you use LIKE, the value in the property must match the string exactly, including the case of
each character. If you use NOT LIKE, then the comparison is TRUE for any value that does not
match exactly.
Sometimes, however, you may not know the text of the character string. For example, you might
want to retrieve all documents that have the word Cake in their title but not know all of the titles.
For these instances, Documentum provides two pattern-matching characters that serve as wildcards.
The two characters are:
• The percent sign (%)
• The underbar (_)
You can include the pattern-matching characters anywhere in a character string.
The percent sign
The percent sign replaces 0 or more characters. For example, suppose the predicate contains:
"subject" LIKE 'Breads%'
This returns any object whose subject begins with the word Breads, regardless of the how many
characters or words follow Breads.
To search for any object that contains a particular string in any position, use a predicate that contains
a percent sign both before and after the literal text. For example, the following predicate returns all
objects whose subject contains the string ’work’:
subject LIKE '%work%'
By default, this predicate returns all objects whose subject contains "work” in any position, including
where the string appears as part of a larger string. For example, it returns any object whose subject
contains "networking” or "workers”.
However, for full-text queries only, a behavior change was implemented beginning in release 6.0 to
make them more useful and perform better. The fulltext query:

object_name like '%mall%'
is interpreted as searching for the whole-word mall instead of a word-fragment mall. The following
query will behave differently for Content Server 6.0 and pre-6.0:
select r_object_id from dm_document where object_name like '%mall%' enable(ftdql)
For pre–6.0, the query produces the following hits:
small.txt
the mall document.txt
For 6.0 and later, the query produces hits only for:
the mall document.txt
This change was done to produce more meaningful hits, since word-fragments often produce a mass
of unwanted hits and cause performance issues for the full-text engine. Additionally, studies of
customer queries showed that users often use this form to look for whole words.
The underbar
The underbar replaces one character. For example, suppose the predicate contains:
"subject" LIKE 'Bread_'
This returns any object whose subject is the single word Bread, followed by a space, an s (Breads), or
any other single, printable character.
Matching cases
You can use the UPPER and LOWER functions to retrieve an object that matches a value regardless of
the case of the letters in the value. For example, to retrieve any object with the word cake in the title,
regardless of its position in the title or the case:
UPPER("title") LIKE '%CAKE%'
Using the UPPER function changes all titles to uppercase for the comparison so that the case of the
characters is not a factor. You can use the LOWER function the same way:
LOWER("title") LIKE '%cake%'
The ESCAPE character
There may be occasions when the pattern you want to match includes a percent sign (%) or an
underbar (_). To match either of those characters literally, you must specify an escape character and
insert that character before the percent sign or underscore in the pattern. Use the optional ESCAPE
clause to specify the escape character.
For example, suppose you want to find all documents whose object names contain an underscore.
Because the underscore is interpreted as a wild card by default, you must define and use an escape
character in the query:
SELECT "r_object_id" FROM "dm_document"
WHERE "object_name" LIKE '%_%' ESCAPE ''

In the above example, the backslash is defined as the escape character. Placed in the pattern directly
ahead of the underscore, it tells the server to treat the underscore as a literal character rather than a
wild card.
In addition to the wildcard characters, an escape character can also escape itself. For example,
suppose you want to match the string %_ and that you have defined the backslash as the escape
character. Your LIKE predicate would look like this:
LIKE '%_' ESCAPE ''
In the above example, the backslash character escapes not only the percent sign and underscore but
also itself.
Escape characters can escape only the two wildcard characters and themselves. If you insert an escape
character before any other character in a pattern, it generates an error.
You can use any printable character as an escape character.
SysObject predicates
Three predicates restrict the search specified in a FROM clause.
When you specify an object type in a FROM clause, the server examines that type and its subtypes
for any objects that fulfill the conditions specified in the rest of the query. However, sometimes
you may want to limit the search to specific subtypes, folders, or cabinets. These three predicates
allow you to do so. They are:
• TYPE
• FOLDER
• CABINET
The TYPE predicate
The TYPE predicate restricts the search to objects of a single type or one of its subtypes. The syntax is:
TYPE(type_name)
The type_name argument must identify a subtype of a type specified in the FROM clause.
For example, the following statement retrieves all documents of the type legal_doc or accounting_doc
or a subtype:
WHERE TYPE("legal_doc") OR TYPE("accounting_doc")
Using the TYPE predicate provides one way to select from more than one type. For example, to
retrieve all documents or workflow process definitions that have been created after a particular
date, you could use the following statement:
SELECT "r_object_id", "object_name", "owner_name", "r_creation_date"
FROM "dm_sysobject" WHERE TYPE("dm_document") OR TYPE("dm_process")

The FOLDER predicate
The FOLDER predicate identifies what folders to search. The syntax is:
[NOT] FOLDER(folder_expression {,folder_expression} [,DESCEND])
The folder_expression argument identifies a folder in the current repository. You cannot search a
remote folder (a folder that does not reside in the current repository). Valid values are:
• An ID function
• The ID function (described in The ID function, page 30) identifies a particular folder.
• A folder path
A folder path has the format:
/cabinet_name{/folder_name}
Enclose the path in single quotes. Because cabinets are a subtype of folder, you can specify a
cabinet as the folder.
• The keyword DEFAULT
The keyword DEFAULT directs the server to search the user’s default folder. Note that a user’s
default folder is the same as the user’s default cabinet (because cabinets are a subtype of folders).
The DESCEND keyword directs the server to search the specified folder or folders and any local
folders directly or indirectly contained within that folder or folders. The predicate does not search
any contained, remote folders. The specified folder or folders may have no more than 25,000 nested
folders if you include the DESCEND keyword.
When the search finds a remote folder, it returns the object ID of the associated mirror object in the
current repository. It does not return the object’s ID in the remote repository.
DESCEND applies to all folders specified in the predicate. If you want to search one folder without
descending but descend through another folder, include two folder predicates in your statement
and OR them together. For example:
FOLDER ('/cakes/yeasted', DESCEND) OR FOLDER (DEFAULT)
The keyword NOT directs the server not to search a particular folder.
The CABINET predicate
The CABINET predicate restricts the search to a particular cabinet. Its syntax is:
[NOT] CABINET(cabinet_expression [,DESCEND])
The cabinet_expression argument must identify a cabinet that resides in the current repository.
Valid values are:
• An ID function
The ID function (described in The ID function, page 30) must specify a cabinet ID.
• A folder path
The folder path must identify a cabinet. Its format is:
/cabinet_name

Enclose the path in single quotes.
The keyword DESCEND directs the server to search the specified cabinet and any folders directly
or indirectly contained in the cabinet that reside in the current repository. The predicate does not
search any contained, remote folders.
When the search finds a remote folder, it returns the object ID of the associated mirror object in the
current repository. It does not return the object’s ID in the remote repository.
The keyword NOT directs the server not to search a particular cabinet.
Logical operators
Logical operators are used in WHERE clauses. DQL recognizes three logical operators:
• AND
• OR
• NOT
AND and OR
The AND and OR operators join two or more comparison expressions to form a complex expression
that returns a single Boolean TRUE or FALSE. The syntax for these operators is:
expression AND | OR expression
Two expressions joined using AND return TRUE only if both expressions are true. For example, the
following complex expression returns TRUE when both of its component expressions are true, and
FALSE when only one of them is true:
"title" = 'Yeast Breads of England' AND "subject" = 'Breads'
Similarly, the following expression also returns TRUE only if both conditions are true:
"subject" = 'Breads' AND ANY "authors" IN ('clarice','michelle','james')
Two expressions joined using OR return TRUE if either expression is true. For example, the following
expression returns true when either comparison is true:
"subject" = 'Cheeses' OR "owner_name" = 'donaldm'
NOT
The NOT operator reverses the logic of an expression. The following expression is a simple example:
"subject" = 'Cheeses'
When the server encounters this expression, it is looking for objects that have a subject property with
the value Cheeses. The server returns TRUE if the subject value is Cheeses and FALSE if it is not.
Now, look at the same expression with the NOT operator:
NOT("subject" = 'Cheeses')

In this instance, the server looks for the objects with subject properties that do not contain the value
Cheeses. The server returns TRUE if the subject is not Cheeses and FALSE if the subject is Cheeses.
Order of precedence
You can join any number of expressions together with logical operators. Content Server imposes no
limit. (Note that your underlying RDBMS may impose a limit.) The resulting complex expression is
evaluated from left to right in the order of precedence of the operators. This order, from highest to
lowest, is:
NOT
AND
OR
To illustrate, look at the following example:
NOT expression1 AND expression2 OR expression3 AND expression4
The server resolves this expression as follows:
1. Evaluates NOT expression1
2. ANDs the result of Step 1 with the result of expression2
3. ANDs the results of expression3 and expression4
4. ORs the results of Steps 2 and 3
You can change the order of evaluation by using parentheses. For example:
NOT expression1 AND (expression2 OR expression3) AND expression4
The server resolves this expression as follows:
1. Evaluates NOT expression1
2. ORs the results of expression2 and expression3
3. ANDs the results of Step 1 and Step 2
4. ANDs the results of Step 3 with the result of expression4
Similarly, you can use parentheses to change the precedence of the NOT operator:
NOT(expression1 AND expression2 AND expression3)
The complex expression inside the parentheses is evaluated first and the NOT operator applied to
its result.

DQL reserved words
DQL reserved words , page 396, lists the DQL reserved words. If you use any of these as object or
property names, you must enclose the name in double quotes whenever it is used in a DQL statement.

Chapter 2
DQL Statements
This chapter contains descriptions of the DQL statements. The description of each statement includes:
• Syntax
• Argument descriptions (if any)
• Return value (if appropriate)
• Detailed information about required permissions and usage
• Related statements
• Example
Quoting Object Type Names and Property Names: Documentum recommends that you put double
quotes around all object type names and property names referenced in applications. Doing that will
ensure that the names will not conflict with any DQL reserved words or any words reserved by your
underlying RDBMS. Documentum object type names and property names will not generate conflicts,
but using the quotes in applications will make sure that no conflicts are generated by user-defined
type or property names. To encourage this best practice, the DQL examples in our manuals use the
double quotes.
Comments within DQL Statements: DQL does not accept inline comments. An inline comment is a
comment embedded within a DQL statement.

Abort
Abort
Purpose
Cancels an explicit transaction.
Syntax
ABORT [TRAN[SACTION]]
Description
The ABORT statement terminates a transaction that was opened with the BEGIN TRAN statement.
Any changes made while the transaction was open are rolled back. They are not saved to the
repository.
The ALTER TYPE statement does not participate in transactions. If you alter a type as part of an
explicit transaction and then issue an ABORT statement, the changes you made to the type are
not reversed.
You can include either keyword, TRAN or TRANSACTION.
Related statements
Begin Tran, page 67
Commit, page 71

Alter Group
Alter Group
Purpose
Modifies a user group.
Syntax
ALTER GROUP group_name ADD members
ALTER GROUP group_name DROP members
ALTER GROUP group_name SET ADDRESS email_address
ALTER GROUP group_name SET PRIVATE TRUE|FALSE
Arguments
Table 8. ALTER GROUP argument descriptions
Variable Description
group_name Identifies the group you want to modify. Use the group’s
name. The name can be a string, which must be enclosed in
single quotes, or an identifier, which need not be in quotes. In
addition, if the name contains German umlauts, enclose the
name in single quotes. For example,
ALTER GROUP 'köln' ADD 'dmadmin';
Note: In API, names containing German umlauts need not
be enclosed in single quotes.
members Identifies users, groups, or both to add or drop from the
group. You can specify user names representing individual
users or names representing groups. Use a comma-separated
list to specify multiple names. Alternatively, you can use a
SELECT statement to identify the members (illustrated in
Examples, page 46).
When you are identifying an individual user, the name must
be the value of the user’s user_name property.
email_address Defines an electronic mail address for a group. Specify this as
a character string literal. You can use any email address that
is valid for your environment. To remove a group’s email
address, specify email_address as an empty string.
Permissions
To alter a group, you must be either the group’s owner or a user with Superuser user privileges.

Alter Group
Description
Each ALTER GROUP statement can perform only one type of operation. That is, you cannot add
and drop members in the same statement. Nor can you set an email address in the same statement
that adds or drops members.
The SET PRIVATE clause
SET PRIVATE sets the is_private property for the group. Setting the property to TRUE makes the
group a private group; setting it to FALSE makes the group a public group.
Related statements
Create Group, page 72
Drop Group, page 97
Examples
The following example adds two users to the group called superusers:
ALTER GROUP superusers ADD steve,chip
The next example uses a SELECT statement to identify which users to add to the engineering group:
ALTER GROUP engineering ADD (SELECT "user_name"
FROM "dm_user" WHERE "user_os_name" LIKE '%eng%')
The final example defines an email address for the engineering group:
ALTER GROUP engineering
SET ADDRESS 'engineering@lion.stellar.com'

Alter Aspect
Alter Aspect
Purpose
Adds, drops, or changes fulltext-indexed properties for an aspect.
Syntax
ALTER ASPECT aspect_name ADD
[(property_def {,property_def})]
[[NO]OPTIMIZEFETCH]
ALTER ASPECT aspect_name FULLTEXT SUPPORT ADD ALL
ALTER ASPECT aspect_name FULLTEXT SUPPORT DROP ALL
ALTER ASPECT aspect_name FULLTEXT SUPPORT ADD property_list
ALTER ASPECT aspect_name FULLTEXT SUPPORT DROP property_list
Arguments
Table 9. ALTER ASPECT argument descriptions
aspect_name Identifies the aspect for which the properties are defined. Use
the object_name of a dmc_aspect_type object.
[NO] OPTIMIZEFETCH Indicates whether to duplicate the names and values
of the properties in the object’s property bag when the
aspect is attached to an object that has a property bag.
OPTIMIZEFETCH directs Content Server to store the values
in the object’s property bag. NO OPTIMIZEFETCH directs
Content Server not to duplicate the properties into the
property bag. The default is OPTIMIZEFETCH for releases
before 6.7, and is NO OPTIMIZEFETCH for release 6.7 and
later.
property_def Defines a property for the aspect. Each property definition
has the following format:
property_name domain [REPEATING] [(property_
modifier_list)]
property_name names the property and domain defines
its datatype. The property name must consist of ASCII
characters. The domain can be any valid DQL datatype.
If the datatype is a character string datatype, the domain
specification must also include the length.
Note: Reserved words should not be used as property names
in aspects.

Alter Aspect
REPEATING defines the property as a repeating property.
property_modifier_list defines data dictionary information for
the property. Valid property modifiers are:
update_modifier
value_assistance_specification
mapping_table_specification
default_specification
constraint_specification
Separate multiple modifiers with commas.
You cannot include a property modifier if the statement is
inside an explicit transaction.
Refer to Property modifiers, page 83, for descriptions of the
property modifiers.
You can specify a maximum of 30 properties in an ALTER
ASPECT statement.
property_list Identifies the properties to be indexed or dropped from
indexing. The properties must be properties defined for the
aspect identified in aspect_name.
Permissions
You must have Superuser privileges to use this statement.
Description
Use the ALTER ASPECT statement to add properties to aspects or to drop or modify properties
already attached to an aspect. The statement is also used to define or modify the fulltext indexing
of aspect properties.
Adding properties
You can add a maximum of 30 properties to an aspect in a single operation. If you wish to add more
than 30 to a particular aspect, you must issue multiple ALTER ASPECT statements.
You cannot add properties to an aspect whose name includes a dot. For example, if the name of the
aspect is "com.mycompany.policy”, then you cannot define properties for that aspect.
If the ALTER ASPECT statement is inside an explicit transaction, you cannot include property
modifiers (data dictionary information) in the definition.
If you specify OPTIMIZEFETCH when you first add properties to an aspect, you must specify it also
if you add more properties later. Similarly, if you specify NO OPTIMIZEFETCH, when you first add
properties to an aspect, you must specify NO OPTIMIZEFETCH also if you add more properties later.

Alter Aspect
When you add a property to an aspect that is already attached to one or more objects, the value for
that property in those objects is the null value (nulldate, nullstring, and so forth).
Dropping properties
Dropping properties of an aspect also removes the property from objects to which the aspect is
attached.
[NO] OPTIMIZEFETCH
You can choose to optimize performance for fetching or querying aspect properties by including
the OPTIMIZEFETCH keyword in the ALTER ASPECT statement. That keyword directs Content
Server to duplicate the properties and their values into the property bag of any object to which the
aspect is attached, if the object has a property bag.
When aspects were first introduced, it was assumed that better performance in fetching qualifiable
aspect properties would result by using OPTIMIZEFETCH when adding properties to an aspect.
OPTIMIZEFETCH was set to be the default for the ALTER ASPECT DQL command. However, as we
accumulated more experience, it became clear that NO OPTIMIZEFETCH gives better performance.
Beginning in release 6.7, the default is changed from OPTIMIZEFETCH to NO OPTIMIZEFETCH for
the ALTER ASPECT DQL command. We recommend that you use the NO OPTIMIZEFETCH option
explicitly, in releases earlier than 6.7, when using this command.
For more information about property bags, non-qualifiable properties, and how they are stored in the
object type tables, refer to EMC Documentum Content Server Fundamentals Guide.
Full-text indexing of aspect properties
Properties associated with aspects are not fulltext-indexed by default. If you wish to index them, you
must issue an ALTER ASPECT statement to identify the aspects you want indexed. Table 10, page 49,
describes the syntax options for specifying full-text indexing and what each defines.
Table 10. Syntax variations for full-text-indexing of aspect properties
Syntax Description
FULLTEXT SUPPORT
ADD ALL
Defines all properties of the aspect for indexing
FULLTEXT SUPPORT
ADD property_list
Defines for indexing only those aspect properties listed in property_list.
FULLTEXT SUPPORT
DROP ALL
Stops indexing of all properties of the aspect
FULLTEXT SUPPORT
DROP property_list
Stops indexing of those aspect properties listed in property_list.
When you add or drop indexing for aspect properties, only new objects are affected. The index is not
updated to add or drop aspect property values for aspects attached to existing objects.
Related statements
None

Alter Aspect
Examples
ALTER ASPECT grant_validation
ADD(grant_amount integer, disbursement_date date,
approved_by string(32))NO OPTIMIZEFETCH

Alter Type
Alter Type
Purpose
Modifies an object type.
Syntax
ALTER TYPE type_name
[FOR POLICY policy_id STATE state_name]
type_modifier_list [PUBLISH]
MODIFY (property_modifier_clause)[PUBLISH]
ADD property_def {,property_def}[PUBLISH]
DROP property_name {,property_name}[PUBLISH]
ALTER TYPE type_name ALLOW ASPECTS
ADD|SET|REMOVE DEFAULT ASPECTS aspect_list
ALTER TYPE type_name ENABLE PARTITION
ALTER TYPE type_name SHAREABLE [PUBLISH]
ALTER TYPE type_name FULLTEXT SUPPORT [
NONE |
LITE ADD ALL
LITE ADD property_list |
BASE ADD ALL |
BASE ADD property_list
]
Arguments
Table 11. ALTER TYPE argument descriptions
type_name Identifies the type to alter. Use the name defined for the type
when it was created.
If you are altering a type to allow aspects, type_name must be
the top-most type in the type’s hierarchy. Refer to Allowing
aspects, page 57, for more information.
If you are altering the type to add default aspects, the type
may not be lightweight object type and the type must have
the r_aspect_name property.

Alter Type
If you are altering the type to enable data partitioning,
type_name must be the top-most type in the type’s hierarchy.
Refer to Allowing Partitioning, page 58, for more information.
If you are altering the type to modify fulltext support, using
the FULLTEXT SUPPORT keywords, type_name must be the
name of a lightweight type.
policy_id Identifies the default lifecycle for the type. Use the policy’s
object ID.
Including the FOR POLICY clause requires at least Version
permission on the lifecycle identified by policy_id.
state_name Identifies one state in the default lifecycle. Use the state’s
name as defined in the dm_policy object.
type_modifier_list Lists one or more specifications that set or alter type data
dictionary information for the type. Valid type modifiers are:
update_modifier
component_specification
type_drop_clause
Separate multiple type modifiers with commas.
Refer to Type modifiers, page 87 for information about the
syntax and use of each specification.
property_modifier_clause Defines the change you want to make to the property. The
syntax for this clause is:
property_name domain [ SPACEOPTIMIZE ]
or
property_name (property_modifier_list)
domain is any valid DQL datatype.
property_modifier_list lists one or more modifiers that set or
alter data dictionary information for the property. Valid
property modifiers are:
update_modifier
property_drop_clause

Alter Type
Refer to Property modifiers, page 83 for information about
the syntax and use of each property modifier.
property_def Defines the properties you want to add to the type.
property_def has the following syntax:
property_name domain [REPEATING]
[[NOT]QUALIFIABLE | SPACEOPTIMIZE]
(property_modifier_list)
property_name is the name of the property and domain is any
valid DQL datatype. The keyword REPEATING defines the
property as a repeating property.
By default, a property is qualifiable. Include NOT
QUALIFIABLE if you wish to define the property as
non-qualifiable. The length of a NOT QUALIFIABLE
property with a string datatype must be less than the value
in the max_nqa_string key in the server.ini file if that is set.
Refer to QUALIFIABLE and NOT QUALIFIABLE properties,
page 55, for more information about QUALIFIABLE.
SPACEOPTIMIZE allows you to assign true NULLs as a
property value.
property_modifier_list lists one or more modifiers that define
data dictionary information for the property. Refer to the
preceding description of property_modifier_clause for a list of
valid property modifiers.
property_name For the DROP option, property_name identifies the property
that you want to remove from the specified type.
aspect_list Comma-separated list of aspects. Use the value of the
object_name property of their dmc_aspect_type objects. You
are not required to quote the aspect list; however, if you do,
use single quotes.
Adding and setting default aspects, page 57, describes how
default aspects are added to an object type. Removing default
aspects, page 58, describes how default aspects are removed
from an object type.

Alter Type
Permissions
To issue an ALTER TYPE statement that changes locale-specific data dictionary information, your
session locale must match exactly one of the locales identified in the dd_locales property of the
repository configuration (docbase config object).
Table 12, page 54, lists the ALTER TYPE operations and describes who can perform them and on
which types.
Table 12. Alter Type operations
Alteration Who can do it To which types
Set default ACL Type owner or Superuser All types
Set default storage area Type owner or Superuser All types
Set or drop data dictionary
information
Type owner or Superuser All types, with the exception
that value assistance and
constraints may not be added
to system-defined object types.
Add read/write properties Type owner or Superuser User-defined only
Add read-only properties Superuser User-defined only
Drop read/write properties Type owner or Superuser User-defined only
Lengthen character string
properties
Type owner or Superuser User-defined only
Allow aspects Superuser Only user-defined object types
that have no supertype (that is,
user-defined object types that
are the top-most type in their
hierarchy).
Add, set, or remove default
aspects
Superuser Any object type that is not a
lightweight object type and
that has the r_aspect_name
property.
Enable type to be partitionable Type owner or Superuser Only user-defined object types
that have no supertype (that is,
user-defined object types that
are the top-most type in their
hierarchy).
Make type shareable Type owner or Superuser User-defined only
Allow properties to store true
NULL values.
Type owner or Superuser User-defined only. (However,
see Appendix B, Database
footprint reduction of
dmr_content objects.)
(For the definition of a read and write or read-only property, refer to the EMC Documentum Content
Server System Object Reference manual.)

Alter Type
Description
This section contains information about using the statement.
General notes
You cannot change the definition of a system-defined type. You can only set a variety of defaults
and data dictionary information. You can change the definition of a user-defined type. (Refer to
Table 12, page 54 for a summary of valid operations, the types on which they can be performed,
and who can perform them.)
You cannot issue an ALTER TYPE statement in an explicit transaction.
In release 6.0, a dynamic type feature was added so that non-intrinsic types will be updated
immediately. (Intrinsic type names start with dm_, dmi_, and dmr_ but do not include these
types: dm_message_archive, dm_message_container, dm_acs_config, dm_validation_descriptor,
dm_cont_transfer_config, and aspect attribute types.) In previous releases, after you execute an
ALTER TYPE statement, the changes were not reflected to users until the global type cache was
updated. The updating was automatic, but could take several minutes. Users currently in a session
with the repository will not see the type change if they performed operations on any instance of the
type in their session, prior to the change. (When a user performs an operation on an object, the system
caches the object type definition for that object.)
QUALIFIABLE and NOT QUALIFIABLE properties
A qualifiable property is a standard property. It is stored in a column in the appropriate table in the
database in which the repository resides. Queries can reference qualifiable properties in selected
value lists and in expressions in qualifications. Properties are qualifiable by default.
A property defined as NOT QUALIFIABLE is stored in a property bag. A non-qualifiable property
can be referenced in selected value lists, but may not be referenced in expressions or qualifications
in a query.
If a property is a string datatype and it is defined as NOT QUALIFIABLE, its length must be less than
the value in the max_nqa_string key in the server.ini file if that key is set.
SPACEOPTIMIZE properties
In releases prior to 6.6, Content Server did not allow you to assign true (actual) NULLs as a property
value, as described in the section, NULLs, default values, and DQL, page 335 . Use SPACEOPTIMIZE
in the property_modifier_clause to change the property, or in the property_def clause to add a property
that allows true NULL values. If you want to modify a string property to allow actual NULL values,
you must specify the domain as string(0). If you wish to also lengthen the string value, you must do it
in another property_modifier clause. The property values stored before the property was modified will
not change. The change only affects newly stored values.
For Oracle installations, SPACEOPTIMIZE can modify all properties. For SQL Server and DB2, only
the character and string properties and the ID properties can be set to SPACEOPTIMIZE.
For example, if you created a property named "attribute1” in "my_type1” with domain of STRING(32),
you could alter that property to use true NULLs with the following statement:
ALTER TYPE "my_type1" MODIFY (attribute1 STRING(0) SPACEOPTIMIZE)

Alter Type
Note: The recommendation is that ID attributes should only have SPACEOPTIMIZE applied when
it is expected that most of the time the value will be NULLID (that is, 0000000000000000). This is
because when a non-null ID value is stored, VARCHAR(16) causes 17 bytes of data to be allocated in
SQL Server versus 16 for CHAR(16). This is why we do not put SPACEOPTIMIZE on r_object_id
since every object has a non-null value.
Localization
If you change data dictionary information that is locale-specific, such as label_text, it is changed only
in the current locale (defined in the session’s session_locale property).
If you change data dictionary information that is not locale-specific, such as a constraint definition,
it is changed in all locales.
PUBLISH
Including PUBLISH causes the server to publish the data dictionary information for the object type
or property immediately. Including PUBLISH publishes the changes made in the ALTER TYPE
statement and any other changes to the type or property that are not yet published.
If you do not include PUBLISH, the information is published the next time the Data Dictionary
Publish job runs.
It is not necessary to include PUBLISH if you are dropping a property from the type definition. The
data dictionary information for the property is automatically removed in such cases.
Modifying a type definition
You can modify the definition of a user-defined type by:
• Adding a new property
• Lengthening a character string property
• Dropping a property defined for the type
Adding a new property
Use the following syntax to add a new property to a user-defined type:
ALTER TYPE type_name ADD property_def {,property_def}
property_def defines the new property and has the following syntax:
property_name domain [REPEATING]
[[NOT]QUALIFIABLE](property_modifier_list)
The property name must following the naming rules outlined in the EMC Documentum Content
Server System Object Reference manual.
domain is any valid DQL datatype. The keyword REPEATING defines the property as a repeating
property. The keyword QUALIFIABLE identifies the property as a qualifiable property. If you
include NOT QUALIFIABLE, the property is a non-qualifiable property. QUALIFIABLE is the
default. For information about this property characteristic, refer to QUALIFIABLE and NOT
QUALIFIABLE properties, page 55.
Valid property modifiers are described in Property modifiers, page 83.
Do not add more than 30 properties in one ALTER TYPE statement. If the repository is running
against DB2, the sum total of the lengths of all properties defined for the type is constrained by

Alter Type
the page size defined for the tablespace. Table 17, page 81 lists the tablespace page sizes and the
corresponding maximum row length allowed.
Lengthening a character string property
The character string property must be defined for the object type. It cannot be an inherited property.
The only valid change that you can make is to lengthen the property (or to allow true NULL values).
The change is applied to all the type’s subtypes and to all existing objects of the type and its subtypes.
For example, if a string property called employee_name is currently 24 characters, the following
statement would change the length to 32 characters:
ALTER TYPE "test_result" MODIFY ("patient_name" char(32))
Dropping properties from the type
Only properties that are defined for the type can be dropped. You cannot drop an inherited property.
Dropping a property also removes the property from all subtypes of the type and removes all data
dictionary information for the property.
The drop operation fails if the property is an indexed property for the type and there are existing
indexed objects of the type or its subtypes. (An indexed property is a property whose value is stored
in the full-text index.)
Note: Some relational database management systems (for example, DB2) don’t support dropping
columns from a table (any property is a column in a table in the underlying RDBMS). For those
databases, if you drop a property, the corresponding column in the table representing the type is
not actually removed. If you later try to add a property to that type that has the same name as the
dropped property, you will receive an error message.
Allowing aspects
Altering an object type to allow instances of the type to accept aspects performs the following
actions on the object type:
• Adds the repeating property, r_aspect_name, to the type definition if the type does not have
the property already
• Adds the i_property_bag property to the type definition if the type does not have the property
already
Adding and setting default aspects
A default aspect is an aspect that is associated with an object type and, by default, associated with
each instance of the type that is created after the aspect is added to the type. Default aspects defined
for an object type are also default aspects for subtypes of that type.
ALTER TYPE supports two key words to add default aspects to an object type definition: ADD and
SET. The ADD keyword appends the specified aspect or aspects to any others already associated
with the object type. The SET keyword replaces the object type’s existing defined default aspects with
the aspects specified in the statement. The SET keyword replaces only the default aspects explicitly
defined for the type. It does not replace any inherited default aspects.
For example, suppose you have an object type named "my_custom_type” with two default aspects,
one that is inherited, named full_validation, and one that is defined for the type: quick_validation.
You want to add an aspect named correct_figures. The following statement adds the correct_figures
aspect to the object type definition while retaining the other two default aspects:
ALTER TYPE my_custom_type ADD DEFAULT ASPECTS correct_figures

Alter Type
After the statement completes, my_custom_type has three default aspects: full_validation, which is
inherited, and quick_validation and correct_figures, which are defined explicitly for my_custom_type.
Now, suppose that your application developers have created a new version of the correct_figures
aspect that also performs a quick validation. This new version is called correct_figures_v2. To remove
the current correct_figures and quick_validation aspects and replace them with the new, combined
aspect, use the following statement:
ALTER TYPE my_custom_type SET DEFAULT ASPECTS correct_figures_v2
my_custom_type now has the following default aspects: full_validation and correct_figures_v2.
The SET keyword replaced both aspects defined for the type with the one aspect specified in the
statement, but left the inherited aspect alone.
When you add an aspect to an object type, only the new instances of the type created after the
addition are affected. The aspect is not attached to existing instances of the type.
Removing default aspects
Use the REMOVE keyword to remove default aspects from an object type definition. You can
remove any default aspect defined specifically for the object type. You cannot remove a default
aspect inherited by the object type.
The statement removes the aspects specified in aspect_list. For example, the following statement
removes the aspect named conditional_validation from my_custom_type:
ALTER TYPE my_custom_type REMOVE DEFAULT ASPECT conditional_validation
An error is returned if the specified aspect is not associated with the specified object type definition.
When you remove a default aspect from an object type, existing instances of the type are not affected.
The aspect remains attached to those instances of the type. You must remove the aspect from the
existing type instances explicitly after removing the default aspect from the type definition.
Allowing Partitioning
Altering an object type to be partitionable adds an additional attribute, i_partition, to the type and its
subtypes. Since an object creates database entries in the tables for its type and all its supertypes, the
whole type hierarchy must be either partitionable or non-partitionable. Predefined types are already
partition enabled (or not), so only user-defined supertypes can be modified to be partitionable. For
example, the following statement makes my_custom_type a partitionable type:
ALTER TYPE my_custom_type ENABLE PARTITION
An error is returned if the type is already partitionable, or the type is not a user-defined type that
has no supertype.
If you alter a type to be partitionable, you must use the Administrative method
PARTITION_OPERATION to take advantage of data partitioning.
Making a type shareable
Altering a type to be shareable adds additional properties, i_sharing_type, i_orig_parent, and
allow_propagating_changes to the type, and marks it as a shareable type. An error will occur if you
attempt to alter an already shareable type or a lightweight type to be shareable.

Alter Type
Fulltext support for lightweight types
Use the FULLTEXT SUPPORT arguments to specify which properties are included in the fulltext
index for lightweight types:
• FULLTEXT SUPPORT—Default, all lightweight attributes are indexed
• FULLTEXT SUPPORT NONE—No attributes are indexed
• FULLTEXT SUPPORT LITE ADD ALL—All lightweight attributes are indexed
• FULLTEXT SUPPORT LITE ADD property_list—Selected lightweight attributes are indexed only
(cannot select parent attributes)
• FULLTEXT SUPPORT BASE ADD ALL—All lightweight and parent attributes are indexed
• FULLTEXT SUPPORT BASE ADD property_list—Selected lightweight or parent attributes are
indexed
Type modifiers
Type modifiers set or change some properties of the type info object and data dictionary information
for the type. For example, you can set the default storage area, ACL, or default lifecycle or add
constraints to the type. You can also drop data dictionary information.
You cannot include a type modifier if the ALTER TYPE statement is executing inside an explicit
transaction.
update_modifiers
Update modifiers set or remove property values in the dm_nls_dd_info, dm_dd_info, and
dmi_type_info objects for the type. The dm_nls_dd_info and dm_dd_info objects hold data dictionary
information for the type. The dmi_type_info object holds non-structural information about the
type. You can set properties in the dmi_type_info object that define the type’s default ACL, default
permissions, default storage area, and default group.
An update modifier is also used to define a default lifecycle for the type.
Setting or removing data dictionary values
You can set any property in the dm_nls_dd_info and dm_dd_info objects that are applicable to
types. Use one of the following statement clauses:
SET property_name[[index]]=value
APPEND property_name=value
INSERT property_name[[index]]=value
REMOVE property_name[[index]]
TRUNCATE property_name[[index]]
The property_name is the name of a property defined for the dm_nls_dd_info or dm_dd_info object
type. The nls_dd_info or dm_dd_info property must be applicable to object types and settable by
users.
Include index if the property is a repeating property. Its interpretation varies:
• For a SET operation, the index defines which value to set.
If the operation is adding a new value to the property, the index must identify the next available
position in the list. If the SET operation is replacing an existing value, specify the index of the
existing value.
• For an INSERT operation, the index defines where to insert the new value.

Alter Type
Existing values are renumbered after the insertion.
• For a REMOVE operation, the index defines which value to remove.
• For a TRUNCATE operation, the index defines the starting position for the truncation.
All values in the list, beginning at that position, are truncated.
The APPEND statement clause doesn’t require an index value for repeating properties because it
automatically puts the new value at the end of the repeating property’s value list.
You must enclose the index in square brackets.
value is a literal value. If you use a SET statement clause and the property is single-valued, value can
be NULL.
You cannot use the SET, INSERT, or APPEND statement clauses to alter any of the following
properties:
• From dm_dd_info type
parent_id, default_value, cond_value_assist, cond_computed_expr, val_constraint, unique_keys,
foreign_keys, primary_key
• From the dm_nls_dd_info type
parent_id
Defining the default ACL
A type’s default ACL does not define access to the type. Users can assign a type’s default ACL to any
object of the type they create. Use the following syntax to set a type’s default ACL:
SET DEFAULT ACL acl_name [IN acl_domain]
The value of acl_name is the ACL’s object name. The acl_domain value is the name of its owner. The
owner’s name is either the user who created the ACL or the alias dm_dbo, representing the repository
owner. The combination of the ACL name and domain uniquely identifies the ACL within the
repository. (For more information about ACLs and their names and implementation, refer to the
EMC Documentum Content Server Administration and Configuration Guide or EMC Documentum Content
Server Fundamentals Guide.)
If either the name or domain includes a character (such as a space) that requires you to enclose the
string in single quotes, then you must enclose both strings in single quotes.
If the default ACL is NULL, the server uses the default ACL defined for the type’s supertype as the
default. To set the default ACL to NULL, use the following syntax:
ALTER TYPE type_name SET DEFAULT ACL NULL
If you set the default ACL to NULL, the server automatically sets the default_owner_permit,
default_group_permit, and default_world_permit properties for the type to NULL also.
Defining the default storage area
The default storage area for a type is where Content Server stores all content files associated with
objects of that type. (Users can change the storage area for an individual object.)
To set the default storage area for a type, use the following syntax:
ALTER TYPE type_name SET DEFAULT STORAGE[=]storage_area_name

Alter Type
storage_area_name is the name of the storage object representing the storage area.
Defining the default group
To set the default group for a type, use the following syntax:
ALTER TYPE type_name SET DEFAULT GROUP[=]group_name
group_name can be a character string literal or an identifier. If it is a character string literal, you must
enclose it in single quotes.
To set the default group to NULL, use:
ALTER TYPE type_name SET DEFAULT GROUP[=]NULL
Defining a default lifecycle
A lifecycle describes the life cycle of an object. Typically, the SysObject type and its subtypes have
default lifecycles. If an object type has a default lifecycle defined for it, users or applications can
attach the default simply by specifying the keyword Default in the Attach method. (An object is not
attached to a default lifecycle automatically. Users or the application must explicitly issue an Attach
method. Defining a default lets users or applications attach an object to the default without requiring
them to know the name or object ID of the default. Also, it allows you to change the default without
requiring you to rewrite and recompile any applications that reference the default.)
To set the default lifecycle for a type, use the following syntax:
SET DEFAULT BUSINESS POLICY[=]chronicle_id
[VERSION version_label]
chronicle_id is the object ID of the root version of the lifecycle. The VERSION clause identifies which
version of the lifecycle you want to apply to the type. If you do not specify a version, the default is
the CURRENT version.
A mapping table specification is typically used to map a user-friendly character string value to an
underlying value that may not be as readable or easy to understand. A mapping table can also be
used to map localized or type-specific values to an underlying value.
For example, Desktop Client uses a mapping table defined at the type level to display a list of the
display config objects appropriate for the object type. The actual names of the display config objects
are mapped to more user-friendly strings. (Display config objects represent subsets of type properties
that have a common display definition.)
If you define a mapping table at the type level, the mappings apply to that object type and its subtypes.
You can define the Check constraint in a type modifier list.
You cannot add a constraint to a system-defined object type.
The syntax for adding a constraint to a type is:
ADD constraint_specification
The FOR POLICY clause can be included when you are defining a check constraint. The clause
defines the check constraint as applicable only when instances of the type are in the specified lifecycle

Alter Type
state. You must have at least Version permission on the lifecycle identified by policy_id to include the
FOR POLICY clause.
If you include the FOR POLICY clause, type_name must be the primary type for the lifecycle identified
in policy_id. The policy_id is the object ID of the dm_policy object for the lifecycle. state_name is the
name of the state for which you are defining the constraint. State names are defined in the lifecycle’s
dm_policy object.
To define a check constraint that applies to all states for an object, do not include the FOR POLICY
clause. Any string literals in the check constraint must be ASCII characters.
Component specifications identify which components can operate on objects of the type. Use the
following syntax to identify a component for a type:
Include the optional FOR POLICY clause to define the component as applicable only when objects
of the type are in the specified state. You must have at least Version permission on the lifecycle
identified by policy_id to include the FOR POLICY clause.
If you include the clause, type_name must be the primary type for the lifecycle identified in policy_id.
policy_id is the object ID of the dm_policy object for the lifecycle. state_name is the name of the state for
which you are defining the component. State names are defined in the lifecycle’s dm_policy object.
The format of a valid component specification is described in component_specification, page 89.
type_drop_clause
A type drop clause removes constraint and component definitions defined for a type. You cannot
remove inherited constraints and component definitions.
• To drop check constraints, use:
DROP CHECK
DROP CHECK drops all check constraints defined for the type. If you include the FOR POLICY
clause, the statement removes the check constraints defined for the specified state. You must have
at least Version permission on the lifecycle identified by policy_id to include the FOR POLICY
clause.
Note: To drop a single check constraint, use an update modifier to remove or truncate the
val_constraint[x], val_constraint_enf[x], and val_constraint_msg[x] properties, where x is the
index position of the check constraint you want to remove.
• To drop a component, use:
DROP COMPONENTS
DROP COMPONENTS drops all components defined for the type. If you include the FOR POLICY
clause, the statement removes the components defined for the specified state. You must have at
least Version permission on the lifecycle identified by policy_id to include the FOR POLICY clause.

Alter Type
Property modifiers
Property modifiers set or drop data dictionary information for a property. Data dictionary
information includes constraints, value assistance, default value definitions, and mapping
information. Property modifiers set properties of dm_nls_dd_info and dm_dd_info objects.
You cannot include a property modifier if the ALTER TYPE statement is executing in an explicit
transaction.
update_modifiers
An update modifier lets you directly set a property in a dm_dd_info or nls_dd_info object. Use the
following syntax to include an update modifier:
MODIFY (property_name (update_modifier))
Including the FOR POLICY clause makes the change to the property effective only when instances
of the type are in the specified lifecycle state. You must have at least Version permission on the
lifecycle identified by policy_id.
The policy_id is the object ID of the dm_policy object for the lifecycle. The state_name is the name
of the state for which you are defining the assistance. State names are defined in the lifecycle’s
dm_policy object.
For a description of valid update modifiers for a property, refer to update_modifier, page 83.
A value assistance specification identifies one or more values for a property. Typically, value
assistance is used to populate a pick list of values for the property when it appears as a field on a
dialog box.
You cannot add value assistance to a system-defined object type.
Use the following syntax to add or change the value assistance specification for a property:
MODIFY (property_name (value_assistance_specification))
Including the FOR POLICY clause provides value assistance only when objects of the type are in
the specified lifecycle state. You must have at least Version permission on the lifecycle identified
by policy_id.
The policy_id is the object ID of the dm_policy object for the lifecycle. The state_name is the name
of the state for which you are defining the assistance. State names are defined in the lifecycle’s
dm_policy object.
For a description of valid value assistance specifications, refer to value_assistance_modifier, page 84.
A mapping table specification typically maps a list of character string values to a list of integer values.
This is useful when you want to provide users with an easily understood list of values for a property
that is an integer data type. Use the following syntax to add or change a mapping table specification:

Alter Type
MODIFY (property_name (mapping_table_specification))
If you include the FOR POLICY clause, the mapped values are only available when instances of the
type are in the specified lifecycle state. You must have at least Version permission on the lifecycle
identified by policy_id.
When you include the clause, type_name must be the primary type for the lifecycle identified in
policy_id. The policy_id is the object ID of the dm_policy object for the lifecycle. The state_name is the
name of the state for which you are defining the mapping. State names are defined in the lifecycle’s
dm_policy object.
For a description of valid mapping table specifications, refer to mapping_table_specification, page 85.
A default specification defines the default value for a property. When a new object of the type is
created, the server automatically sets the property to the default value if no other value is specified
by the user.
You cannot add a property and specify a default value for it in the same ALTER TYPE statement. You
must use two ALTER TYPE statements: one to add the property and one to set its default value.
Use the following syntax to set or change a default specification:
MODIFY (property_name ([SET] default_specification))
The default value can be specified as:
• A literal value
• One of the following keywords: USER, NOW, TODAY, TOMORROW, YESTERDAY
• NULL (This option is not valid for repeating properties.)
For a single-valued property, the default specification syntax is:
DEFAULT[=]default_value
For a repeating property, the default specification syntax is:
DEFAULT[=](default_value {,default_value})
For example, the following ALTER TYPE statement sets the default value for one single-valued
property and one repeating property in the object type mytype:
ALTER TYPE "mytype"
MODIFY ("single_attr" (SET default=NOW)),
MODIFY ("rep_attr" (SET default=(USER)))
The default value must be appropriate for the datatype of the property. For example, you cannot
specify the keyword USER for a date property. You cannot specify NULL for a repeating property.
You can specify the following constraints in a property modifier list:
• Not null
• Check
You cannot add a constraint to a system-defined object type.
The syntax for adding a constraint to a property is:

Alter Type
MODIFY (property_name (ADD constraint_specification))
The FOR POLICY clause can be included when you are defining a NOT NULL or check constraint.
The clause defines the constraint as applicable only when instances of the type are in the specified
lifecycle state. To define a NOT NULL or check constraint that applies to all states for the object, do
not include the FOR POLICY clause.
If you include the clause, you must have at least Version permission on the lifecycle identified by
policy_id. The type_name must be the primary type for the lifecycle identified in policy_id. The policy_id
is the object ID of the dm_policy object for the lifecycle. The state_name is the name of the state for
which you are defining the constraint. State names are defined in the lifecycle’s dm_policy object.
For an explanation of each constraint specification, refer to constraint_specification, page 86.
property_drop_clause
An property drop clause removes constraints, value assistance, or mapping information defined for
the property. You cannot drop inherited constraints, value assistance, or mapping information.
• To drop check constraints, use:
DROP CHECK
DROP CHECK drops all check constraints defined for the type. If you include the FOR POLICY
clause, the statement removes the check constraints defined for the specified state. You must have
at least Version permission on the lifecycle identified by policy_id.
Note: To drop a single check constraint, use an update modifier to remove or truncate the
val_constraint[x], val_constraint_enf[x], and val_constraint_msg[x] properties, where x is the
index position of the check constraint you want to remove.
• To drop value assistance, use the following syntax:
MODIFY (property_name (DROP VALUE ASSISTANCE))
• To drop mapping information, use the following syntax:
MODIFY (property_name (DROP MAPPING TABLE))
Related statements
Alter Aspect, page 47
Create Type, page 77
Drop Type, page 98
Examples
This example sets the owner’s object-level permission and the default group for the user-defined
type called legal_document:
ALTER TYPE "legal_document"

Alter Type
SET OWNER PERMIT browse,
DEFAULT GROUP lawyers
This example adds a property to the user-defined type called report_doc:
ALTER TYPE "report_doc"
ADD "monthly_total" integer
The next example changes the length of the client_comments property in the user-defined type
called case_report:
ALTER TYPE "case_report"
MODIFY ("client_comments" string(255))
This next example sets the default ACL for the document object type:
ALTER TYPE "dm_document"
SET DEFAULT ACL generic_doc IN dm_dbo
This next example demonstrates the use of single quotes in setting the ACL:
ALTER TYPE "dm_document" SET DEFAULT ACL 'mktg default'
IN 'marketing'
This final example changes the type my_super to a partitionable type:
ALTER TYPE "my_super" ENABLE PARTITION

Begin Tran
Begin Tran
Purpose
Opens an explicit transaction.
Syntax
BEGIN TRAN[SACTION]
Permissions
Anyone can execute the BEGIN TRAN statement.
Description
The BEGIN TRAN or BEGIN TRANSACTION statement opens an explicit transaction. When you are
working in an explicit transaction, none of the changes you make to files or property values in objects
are saved until you issue a COMMIT statement to commit the changes.
Related statements
Abort, page 44
Commit, page 71

Change...Object
Change...Object
Purpose
Changes the object type of one or more objects.
Syntax
CHANGE current_type [(ALL)] OBJECT[S]
TO new_type[update_list]
[IN ASSEMBLY document_id [VERSION version_label] [DESCEND]]
[SEARCH fulltext search condition]
[WHERE qualification]
Arguments
Table 13. CHANGE...OBJECT argument descriptions
current_type Identifies the object type of the objects to change. Use the
type’s name.
new_type Specifies the new object type for the objects. Use the type’s
name. Valid new types depend on the object’s current type.
Refer to Rules and constraints on changing types, page 69 for
a complete description.
update_list Specifies one or more change operations to perform on the
objects you are changing.Valid formats are:
set property_name = value set property_
name[[index]] = value append [n] property_name
= value insert property_name[[index]] =
value remove property_name[[index]] truncate
property_name[[index]]
If you specify more than one operation, use commas to
separate the clauses. Refer to The update operations, page
162 for details of these options.
IN ASSEMBLY clause Limits the statement to those objects that belong to a
particular assembly. For details about the syntax and use,
refer to The IN ASSEMBLY clause, page 139.
SEARCH clause Restricts the candidates for change to those that satisfy the
full-text search condition. Refer to The SEARCH clause, page
140 for a description of its syntax and use.
WHERE clause Defines a qualification used to restrict what objects are
changed. Refer to The WHERE clause, page 142 for a
description of a valid qualification.

Change...Object
Return value
The CHANGE...OBJECT statement returns a collection identifier. The collection has one query
result object, with one property called objects_changed. The property contains the number of
objects changed.
Permissions
To use this statement, you must have Delete permission for the objects you want to change.
To use the update_list option to change an object’s owner (owner_name property), you must be have
Superuser user privileges or Change Ownership privilege.
Description
The CHANGE...OBJECT[S] statement lets you change one or more objects from one custom object
type to another.
Note that when you change an object from a subtype to a supertype, you lose the values of the
properties that are inherited from the supertype.
Objects that meet the criteria in the statement are deleted from the repository and recreated as objects
of the specified new type. The recreated objects retain their old object IDs and all their previous
relationships. For example, if a document is annotated, changing that document to another type
of document does not delete its annotations.
You cannot execute CHANGE...OBJECT on an object that is immutable.
Rules and constraints on changing types
There are some constraints when you change an object’s type:
• The object’s current type and the new type must have the same type identifier.
The type identifier is the first two characters in an object ID. This value is inherited from the type’s
supertype. For example, all dm_document subtypes have 09 as their type identifier.
• The old and new types cannot be at the same level in the type hierarchy. The new type must be a
supertype or subtype of the current type. That is, the new type must be in a direct hierarchical
line from the current type.
You can move objects up or down in the object hierarchy but not laterally. For example, suppose
TypeA and TypeB are both subtypes of mybasetype, and TypeC is a subtype of TypeB. You
can change objects of Type B to mybasetype or to TypeC. You cannot change objects of TypeA
directory to TypeB because these two types are peers in the hierarchy. Nor can you change objects
of TypeC to TypeA because TypeA is not a supertype or subtype of TypeC.
• You cannot change an object type that is not a subtype of dm_sysobject using DQL.
When you change an object to a type that is higher in the hierarchy, the server drops any properties
from the old type that are not in the new type. Similarly, when you move an object to a type that is
lower in the hierarchy, the server adds any properties in the new type that were not in the old type.
Unless you use the update_list option to set these new properties, the server assigns them default
values appropriate for their datatypes.

Change...Object
The optional clauses are applied in the following order:
• The SEARCH clause
• The IN ASSEMBLY clause
• The WHERE clause
Using the (ALL) option
Including (ALL) directs Content Server to consider all versions of the object for the change. If you
do not include (ALL), the server only considers the CURRENT version for the change. ALL must
be enclosed in parenthesis.
Related statements
Create...Object, page 74
Delete...Object, page 93
Update...Object, page 158
Examples
This example changes all special_document documents for which the subject is new book proposal to
the document subtype book_proposal:
CHANGE "special_document" OBJECTS TO "book_proposal"
SET "department" = 'marketing'
WHERE "subject" = 'new book proposal'

Commit
Commit
Purpose
Commits the changes made during an explicit transaction and closes the transaction.
Syntax
COMMIT [TRAN[SACTION]]
Permissions
Anyone can execute the COMMIT statement.
Description
The COMMIT statement closes a transaction that was opened with the BEGIN TRAN statement and
commits to the repository any changes made to objects or files during the transaction.
You can include either TRAN or TRANSACTION.
Related statements
Abort, page 44
Begin Tran, page 67

Create Group
Create Group
Purpose
Creates a user group.
Syntax
CREATE [PUBLIC|PRIVATE] GROUP group_name
[WITH][ADDRESS email_address] [MEMBERS members]
Arguments
Table 14. CREATE GROUP argument descriptions
group_name Defines the name of the group that you are creating. This
name must be unique among all group names and user
names in the repository. You can specify the name as an
identifier or as a character string literal. In addition, if the
name contains German umlauts, enclose the name in single
quotes. For example,
CREATE GROUP 'köln';
Note:
• In API, names containing German umlauts need not be
enclosed in single quotes.
• Content Server stores all group names in lowercase.
email_address Specifies the electronic mail address of the group. Use a
character string literal. You can specify any email address
that is valid for your environment.
members Specifies users to be included in the group. You can
specify user names representing individual users, names
representing other groups, or both. The names must appear
as a comma-separated list. Alternatively, you can use a
SELECT statement to populate the group.
When you are specifying an individual user, the name must
be the value of the user’s user_name property.
Return value
The CREATE GROUP statement returns a collection whose result object has one property,
new_object_ID, which contains the object ID of the new group.

Create Group
Permissions
You must have Create Group, Sysadmin, or Superuser user privileges to create a group.
Description
When you create a group, you are its owner and can alter or delete the group.
Public and private groups
The PUBLIC keyword creates the group as a public group. The PRIVATE keyword creates the group
as a private group. When a user with Sysadmin or Superuser user privileges creates a group, the
group is public by default. When a user with Create Group user privileges creates a group, the
group is private by default.
The public or private setting is stored in the is_private property for the group. The setting is not used
by Content Server. All groups, public or private, are visible in the dm_group type. This property is
provided for use by user-written applications.
Related statements
Alter Group, page 45
Drop Group, page 97
Examples
The following example creates a group called supers whose members are all the users with the
Superuser user privilege:
CREATE GROUP supers MEMBERS
(SELECT "user_name" FROM "dm_user"
WHERE "user_privileges" >= 16)
The next example creates a group that includes all of Ron’s co-workers but not Ron:
CREATE GROUP rons_baby_gift WITH MEMBERS
WHERE "user_name" != 'Ron')
This final example creates a group and defines an email address for that group:
CREATE GROUP client_group
WITH ADDRESS 'john@jaguar.docu.com'
MEMBERS john,regina,kendall,maria

Create...Object
Create...Object
Purpose
Creates an object.
Syntax
CREATE type_name OBJECT update_list
[,SETFILE 'filepath' WITH CONTENT_FORMAT='format_name']
{,SETFILE 'filepath' WITH PAGE_NO=page_number}
Arguments
Table 15. CREATE...OBJECT argument descriptions
type_name Identifies the type of object to create. Specify the name of
the object type. You can use any valid type in the repository
with the exception of the object types that represent
aspect properties. type_name cannot be the name of a type
representing a type describing aspect properties.
update_list Specifies one or more operations you want to perform on the
new object. Valid formats are:
name[[index]] = value append [n]property_name
property_name[[index]] [un]link 'folder path'
move [to] 'folder path'
If you include multiple clauses in the update list, use commas
to separate the clauses.
SETFILE clause WITH CONTENT_
FORMAT option
Adds the first content file to the new object. Refer to WITH
CONTENT_FORMAT option, page 75 for details.
SETFILE clause WITH PAGE_NO
option
Adds additional content to the object. Refer to WITH
PAGE_NO option, page 75 for details.
Return value
The CREATE...OBJECT statement returns a collection whose result object has one property,
object_created, which contains the object ID of the new object.

Create...Object
Permissions
Anyone can issue the CREATE...OBJECT statement. However, you must have Superuser privileges to
include the SETFILE clause.
Description
The CREATE...OBJECT statement creates and saves a new object. As part of the process, you can set
some of the object’s properties, specify a storage location for the object, and associate one or more
content files with the object.
If you do not use the link update option to define a storage location, the new object is stored in
your default folder or cabinet.
The SETFILE clauses
A SETFILE clause adds content to the new object. You must have Superuser privileges to include the
clause in the statement. Using the SETFILE clause is subject to the following conditions:
• The content must be a file. It cannot be a block of data in memory.
• The content file must be located in a directory visible to Content Server.
• You cannot add a file created on a Macintosh machine.
• The content cannot be stored in content-addressed storage or in a turbo store storage area.
Any object capable of having content may have multiple associated content files. All files must have
the same content format. The content format is defined when you add the first content file to the
object. All subsequent additions must have the same format. Consequently, specifying the format for
content additions after the first file is not necessary. Instead, you must specify the content’s position
in the ordered list of content files for the object.
To add the first content, use the SETFILE clause with the WITH CONTENT_FORMAT option.
To add additional content, use the SETFILE clause with the PAGE_NO option.
You can’t include both options in a single SETFILE clause.
WITH CONTENT_FORMAT option
Use this SETFILE option to add the first content file to a new object. The syntax is:
SETFILE 'filepath' WITH CONTENT_FORMAT='format_name'
The filepath must identify a location that is visible to Content Server.
The format name is the name found in the name property of the format’s dm_format object.
WITH PAGE_NO option
Use this SETFILE option to add additional content to a new object. The syntax is:
SETFILE 'filepath' WITH PAGE_NO=page_number
The page number identifies the file’s position of the content file within the ordered contents of the
new object. You must add content files in sequence. For example, you cannot add two files and
specify their page numbers as 1 and 3, skipping 2. Because the first content file has a page number of
0, page numbers for subsequent additions begin with 1 and increment by 1 with each addition.

Create...Object
You cannot use the SETFILE clause with the PAGE_NO option in a CREATE...OBJECT statement
unless the statement contains a prior SETFILE clause with the CONTENT_FORMAT option.
Related statements
Change...Object, page 68
Examples
The following example creates a new document and sets its title and subject:
CREATE "dm_document" OBJECT
SET "title" = 'Grant Proposal',
SET "subject" = 'Research Funding'
This example creates a new document, sets its title, and adds two content files. The example is
shown for both Windows and UNIX platforms.
On Windows:
SETFILE 'c:proposalsgrantreq.doc'
WITH CONTENT_FORMAT='msww',
SETFILE 'c:proposalsbudget.doc' WITH PAGE_NO=1
On UNIX:
SETFILE 'u12/proposals/grantreq.doc
WITH CONTENT_FORMAT='msww',
SETFILE 'u12/proposals/budget.doc' WITH PAGE_NO=1

Create Type
Create Type
Purpose
Creates an object type.
Syntax
To create a type:
CREATE TYPE type_name
[WITH] SUPERTYPE parent_type
[type_modifier_list] [PUBLISH]
To create a partitionable type:
CREATE PARTITIONABLE TYPE type_name
[WITH] SUPERTYPE NULL
To create a shareable type:
CREATE SHAREABLE TYPE type_name
[WITH] SUPERTYPE parent_type [PUBLISH]
To create a lightweight type:
CREATE LIGHTWEIGHT TYPE type_name
SHARES shareable_type
[AUTO MATERIALIZATION |
MATERIALIZATION ON REQUEST |
DISALLOW MATERIALIZATION]
[FULLTEXT SUPPORT [
NONE |
LITE ADD ALL
BASE ADD ALL |
]
[PUBLISH]
Arguments
Table 16. CREATE TYPE argument descriptions
type_name Names the new type. Use any valid name that is unique
among the other user-defined type names in the repository.
Types with names beginning with dm can only be created by
a user with Superuser user privileges.
The type name must consist of ASCII characters.

Create Type
property_def Defines a property for the new type. You can define up to 30
properties. Each property definition has the following format:
property_name domain [REPEATING] [[NOT]
QUALIFIABLE | SPACEOPTIMIZE] [(property_
modifier_list)]
property_name names the property and domain defines
its datatype. The property name must consist of ASCII
characters. The domain can be any valid DQL datatype. If
the datatype is a character or string datatype, the domain
specification must also include the length.
The domain specification is one of the following;
• bool[ean]
• char[acter]
• date
• double
• float
• ID
• integer
• smallint (Oracle uses NUMBER(5))
• string
• time
• tinyint (not supported on DB2, Oracle uses NUMBER(3))
REPEATING defines the property as a repeating property.
By default, a property is qualifiable. Include NOT
QUALIFIABLE if you wish to define the property as
non-qualifiable. The length of a NOT QUALIFIABLE
property with a string datatype must be less than the value
in the max_nqa_string key in the server.ini file if that is set.
Refer to QUALIFIABLE and NOT QUALIFIABLE properties,
page 81, for more information about QUALIFIABLE.
SPACEOPTIMIZE allows you to assign true NULLs as a
property value.

Create Type
property_def continued property_modifier_list defines data dictionary information for
the property. Valid property modifiers are:
update_modifier
You cannot include a property modifier if the statement is
inside an explicit transaction.
Refer to Property modifiers, page 83 for descriptions of the
property modifiers.
parent_type Identifies the supertype of the new type. Valid supertypes
for non-shareable types are:
• dm_category
• dm_email_message
Note: dm_email_message is a deprecated object type.
• dm_message_archive
• dm_relation
• dm_state_extension
• dm_state_type
• dm_sysobject and its subtypes
• dm_taxonomy
• dm_user and its subtypes
• user-defined types
Valid supertypes for shareable types are:
• dm_sysobject and its subtypes
• user-defined types
To create a type with no supertype, specify parent_type as
NULL. This requires the Superuser user privilege.

Create Type
type_modifier_list Defines data dictionary information for the type. Valid type
modifiers are:
update_modifier
mapping table specification
You cannot include a type modifier if the statement is inside
an explicit transaction.
Refer to Type modifiers, page 87 for descriptions of the type
modifiers.
shareable_type A previously defined shareable type that will be the parent
of the lightweight type
property_list A comma separated list of properties
Return value
The CREATE TYPE statement returns a collection whose result object has one property,
new_object_ID, which contains the object ID of the new object type.
Permissions
You must have Create Type, Sysadmin, or Superuser privileges to create a new object type.
To define read-only properties for a new type (properties whose names begin with r_) or to create a
type that has no supertype, you must have Superuser privileges.
If the statement sets locale-specific information for the new type or properties, your session
locale must match exactly one of the locales defined in the dd_locales property of the repository
configuration (docbase config object).
Description
The user who creates a type becomes the type’s owner.
You cannot include a CREATE TYPE statement in an explicit transaction.
Do not define more than 30 properties in a single CREATE TYPE statement. If the type has more
than 30 properties, use ALTER TYPE to add the additional properties. The total number of properties
in the new type cannot exceed the supported number of columns in a table in the underlying
RDBMS. On DB2, the sum total of the lengths of the properties defined for the type cannot exceed the
maximum row length set by the page size of the tablespace. Table 17, page 81 lists the tablespace
page sizes and the corresponding maximum row length allowed.

Create Type
Table 17. DB2 tablespace page sizes and associated maximum row lengths
Tablespace Page Size Maximum Row Length, in bytes
4K 4005
8K 8001
16K 16293
32K 32677
Additionally, be sure to follow the naming rules for properties described in the EMC Documentum
Content Server System Object Reference manual.
QUALIFIABLE and NOT QUALIFIABLE properties
A qualifiable property is a standard property. It is stored in a column in the appropriate table in the
database in the repository. You can reference qualifiable properties in selected value lists and in
expressions in qualifications in queries. Properties are qualifiable by default.
A property defined as NOT QUALIFIABLE is stored in a property bag. A non-qualifiable property
can be referenced in selected value lists, but not in qualifications except under certain circumstances.
If a property is a string datatype and it is defined as NOT QUALIFIABLE, its length must be less than
the value in the max_nqa_string key in the server.ini file if that key is set.
SPACEOPTIMIZE properties
In releases prior to 6.6, Content Server did not allow you to assign true (actual) NULLs as a property
value, as described in the section, NULLs, default values, and DQL, page 335 . Use SPACEOPTIMIZE
in the property_def clause to add a property that allows true NULL values. For Oracle installations,
SPACEOPTIMIZE can modify all properties. For SQL Server and DB2, only the character and string
properties and the ID properties can be set to SPACEOPTIMIZE.
For example, to create a type, my_type_a, with a property named attr1 of string(32) and using true
NULLs, use the following statement:
CREATE TYPE "my_type_a" ( attr1 STRING(32) SPACEOPTIMIZE) SUPERTYPE "dm_document"
Specifying an ACL for the type
If one or more servers in the repository is using type-based ACL inheritance, it is recommended that
you make sure that the object type has an ACL defined for the type. The ACL does not determine
who can use the type, but serves as the default ACL for objects of the type when users create the
objects without naming an ACL for the objects. After you create the object type, use ALTER TYPE to
set the acl_name and acl_domain properties for the object type if needed.
Note: Default ACL inheritance is defined in the server config object, in the default_acl property. If
that property is set to 2, for type-based ACL inheritance, when a user creates an instance of the

Create Type
type and does not explicitly assign an ACL to the object, the server will assign the ACL associated
with the object type.
Localization
When you create a new type, any data dictionary information that you define for the type or its
properties in the statement is published in all locales identified in the dd_locales property of the
repository configuration (docbase config object).
PUBLISH
Including PUBLISH causes the server to publish the data dictionary information for the object type
immediately. Publishing creates the dd type info , dd attr info, and dd common info objects that store
the published data dictionary information for the object type.
If you don’t include PUBLISH, the information is published the next time the Data Dictionary
Publish job runs.
PARTITIONABLE Types
A partitionable type (and its subtypes) has an additional attribute, i_partition. Since an object creates
database entries in the tables for its type and all its supertypes, the whole type hierarchy must be
either partitionable or non-partitionable. Predefined types are already partition enabled (or not), so
only user-defined supertypes can be created as partitionable.
If you create a partitionable type, you must use the Administrative method PARTITION_OPERATION
to take advantage of data partitioning.
SHAREABLE Type
A shareable type is created for use with lightweight SysObjects. A shareable object acts as the
parent to the lightweight SysObject. Creating a shareable type creates additional non-inherited
properties, i_sharing_type, i_orig_parent, and allow_propagating_changes to the type, in addition to
the properties explicitly specified, and marks it as a shareable type. See EMC Documentum Content
Server Fundamentals Guide for a discussion of shareable and lightweight SysObjects.
LIGHTWEIGHT Type
A lightweight object shares its SysObject properties with other lightweight objects, reducing storage
requirements for the object. A shareable object acts as the parent to the lightweight SysObject, and
contains the properties shared among the lightweight types. Creating a lightweight type creates the
i_sharing_parent property, in addition to the custom properties explicitly specified, and marks it as
a lightweight type. This property links the child lightweight object with its shareable parent. See
EMC Documentum Content Server Fundamentals Guide for a discussion of shareable and lightweight
SysObjects.
Note: The maximum length of light-weight type name for a repository that uses Oracle is 26.
Fulltext support for lightweight types
Use the FULLTEXT SUPPORT arguments to specify which properties are included in the fulltext
index:
• FULLTEXT SUPPORT—Default, all lightweight attributes are indexed
• FULLTEXT SUPPORT NONE—No attributes are indexed
• FULLTEXT SUPPORT LITE ADD ALL—All lightweight attributes are indexed

Create Type
• FULLTEXT SUPPORT LITE ADD property_list—Selected lightweight attributes are indexed only
(cannot select parent attributes)
• FULLTEXT SUPPORT BASE ADD ALL—All lightweight and parent attributes are indexed
• FULLTEXT SUPPORT BASE ADD property_list—Selected lightweight or parent attributes are
indexed
Property modifiers
Property modifiers define data dictionary information for a property. Defining property modifiers
sets properties in data dictionary-related objects for the type. It does not set any dm_type or
dm_type_info properties for the type.
update_modifier
An update_modifier specification can be:
SET property_name[[index]]=value
APPEND property_name=value
INSERT property_name[[index]]=value
REMOVE property_name[[index]]
TRUNCATE property_name[[index]]
property_name is the name of a property defined for the dm_nls_dd_info or dm_dd_info object type.
The dm_nls_dd_info or dm_dd_info property must be applicable to object type properties and
settable by users. If the property is a dm_nls_dd_info property, only the dm_nls_dd_info object
specific to the current locale is updated.
Include index if the property is a repeating property. Its meaning varies:
• For a SET operation, the index defines which value to set.
If the operation is adding a new value to the property, the index must identify the next available
position in the list. If the SET operation is replacing an existing value, specify the index of the
existing value.
• For an INSERT operation, the index defines where to insert the new value.
Existing values are renumbered after the insertion.
• For a REMOVE operation, the index defines which value to remove.
• For a TRUNCATE operation, the index defines the starting position for the truncation.
All values in the list, beginning at that position, are truncated.
The APPEND statement clause doesn’t require an index value for repeating properties because it
automatically puts the new value at the end of the repeating property’s value list.
You must enclose the index in square brackets.
value is a literal value. If you use a SET statement clause and the property is single-valued, value can
be NULL.
You cannot use the SET, INSERT, or APPEND formats against the following properties:
• From dm_dd_info type
parent_id, default_value, cond_value_assist, cond_computed_expr, val_constraint, unique_keys,
foreign_keys, primary_key
• From the dm_nls_dd_info type

Create Type
parent_id
value_assistance_modifier
A value assistance modifier identifies one or more values for the property. Typically, value assistance
is used to populate a pick list of values for the property when it appears as a field on a dialog box. A
value assistance modifier has the following format:
VALUE ASSISTANCE IS
[IF (expression_string)
va_clause
(ELSEIF (expression_string)
va_clause)
ELSE]
va_clause
[DEPENDENCY LIST ([property_list])
You can include multiple ELSEIF clauses.
expression_string
expression_string defines a Boolean condition. If it returns TRUE, the server executes the associated
va_clause to return values for the property. expression_string can be an expression or a complete
user-defined routine. It must be written in Docbasic and must return a Boolean value. Any string
literals in the expression or routine must consist of ASCII characters. You must have Sysadmin or
Superuser user privileges to provide a user-defined routine for expression_string.
For expressions, the syntax is:
expression [LANGUAGE docbasic]
For user-defined routines, the syntax is:
routine_name
([routine_parameter {,routine_parameter}])
FROM OBJECT(object_id)
[LANGUAGE docbasic]
routine_parameter is a property name.
The FROM clause must identify an object whose content file contains the source code for the routine.
va_clause
The va_clause provides the values for the property. It has two possible formats. One format provides
a list of literal values. The second defines a query to return the values. Each format allows you to
provide an estimate of the expected number of property values and indicate whether the values
represent the complete list of allowed values for the property.
• To provide a list of literal values, use the syntax:
LIST (literal_list)
[VALUE ESTIMATE=number]
[IS [NOT] COMPLETE]
• To provide values from a query, use the syntax:
QRY 'query_string'
[QRY ATTR = property_name]
[ALLOW CACHING]
[VALUE ESTIMATE=number]
[IS [NOT] COMPLETE]

Create Type
query_string is a DQL query. You can use the special token, $$, in the query string. When the query
is executed, a single dollar sign will appear in the query in place of the token.
The QRY ATTR clause defines which of the properties in the selected values list to display.
Typically, query_string has only one property in its selected values list—the property for which
you are providing the value assistance. However, some situations may require you to put more
than one property in the selected values list. In such cases, the server assumes that the first
property selected is the property for which you are providing value assistance. If this is not the
case, you must include the QRY ATTR clause to define which property is the property for which
you are providing assistance.
The ALLOW CACHING clause permits clients to cache the query results.
dependency clause
The DEPENDENCY LIST clause identifies the properties on which expression_string depends. The
default is all properties named in the expression_strings.
A mapping table specification most commonly maps a list of descriptive character strings to a
corresponding list of integers. For example, assume an object type has an integer property called
country in which each value represents a different country. If you display that property in a dialog
box, you want users to see a name for each country rather than an integer value. Using a mapping
table specification, you can map the integer values to corresponding country names.
The syntax is:
MAPPING TABLE (map_element {,map_element})
where map_element is:
VALUE=value_string
[DISPLAY=display_string]
[COMMENT=description_string]
For example:
MAPPING TABLE (VALUE=4
DISPLAY=Spain
COMMENT='Added to list in first quarter 98')
The default for DISPLAY is the data value as a character string literal. The default for COMMENT is
a NULL string.
A default specification provides a default value for a property. When a user creates a new instance
of the type, Content Server automatically sets the property to the default value if no other value is
specified by the user.
The default value can be specified as:
• A literal value
• One of the following keywords: USER, NOW, TODAY, TOMORROW, YESTERDAY
• NULL (This option is not valid for repeating properties.)
For a single-valued property, the syntax is:
DEFAULT[=]default_value
For a repeating property, the syntax is:

Create Type
DEFAULT[=](default_value {,default_value})
If you specify the default value with a keyword that indicates a date or time, the keyword is specified
using the DATE function:
DATE(keyword)
For example, the following statement creates object type mytype with one single-valued date property
and one repeating string property and defines a default value for each:
CREATE TYPE "mytype" ("single_attr" date (default=date(NOW)),
"rep_attr" string(32) repeating (default=(USER)))
WITH SUPERTYPE "dm_document"
The default value must be appropriate for the datatype of the property. For example, you cannot
specify the keyword USER for a date property. You cannot specify NULL for a repeating property.
A constraint_specification defines constraints for a property. For example, you can define a check
constraint to provide data validation for the property.
You can define a CHECK constraint in a property modifier list:
If you define a check constraint in a property modifier list, only the single property for which the
constraint is defined can be part of the constraint. If the constraint must include multiple properties,
define it in the type modifier list.
For each constraint, you can define an error message to display if the constraint is violated.
Any constraint you define is inherited by all the type’s subtypes.
The EMC Documentum Content Server Fundamentals Guide contains the expanded information about
the constraints and what they do.
Check
Check constraints are used most commonly for data validation. The constraint consists of an
expression or routine that must return TRUE. If it returns FALSE, the property’s value violates the
constraint. To define a check constraint, use:
CHECK(expression_string)
expression_string can be an expression or a complete user-defined routine. The expression or routine
must be written in Docbasic and return a Boolean value. Any string literals in the expression or
routine must consist of ASCII characters. To specify a routine, you must have Sysadmin or Superuser
user privileges.
[REPORT 'message_string' [ON VIOLATION]]
For user-written routines, the syntax is:
routine_name
FROM OBJECT(object_id |
PATH folder_path[VERSION version_label])
[LANGUAGE docbasic]

Create Type
Identify the object using the object’s ID or a folder path for an object. If you use a folder path, you can
include a version label. If you don’t include a version label, the label CURRENT is assumed.
message_string is a string literal that contains the error message you want displayed if the constraint is
violated. You can include the following special tokens in the message to parameterize it:
• $value(property_name)
When the message is displayed, $value(property_name) is replaced with the value entered by the
user in the field associated with the property. The property name must be enclosed in parentheses
with no spaces before the opening parenthesis. For example:
The security type must be folder" or cabinet and you
entered $value(security_val).
• $$
Use $$ in a message string to display a single dollar sign in the message.
Type modifiers
Type modifiers define data dictionary information for the type. For example, you can set the type’s
default lifecycle or define constraints for the type. Type modifier set properties in the data dictionary
objects for the type. No properties are set in the dm_type or dm_type_info objects for the type.
update_modifier
You can use the same update modifiers for types as for properties (refer to Property modifiers, page
83). In addition, you can use the following to set the default lifecycle for a type:
SET DEFAULT BUSINESS POLICY[=]
chronicle_id [VERSION version_label]|
NULL|
NONE
chronicle_id is the object ID of the root version of the lifecycle. The VERSION clause identifies which
version of the lifecycle you want to use. The default is the CURRENT version.
If you set the default lifecycle to NULL, the type inherits the default lifecycle from its supertype.
If you set the default lifecycle to NONE, the type has no default lifecycle.
You can define a Check constraint in a type modifier list.

Create Type
For each constraint, you can define an error message to display if the constraint is violated.
Constraints are defined in a type modifier list if the constraint includes multiple properties. If only a
single property defines the constraint, the constraint is typically defined in the property modifier list
for the property rather than the type modifier list.
The EMC Documentum Content Server Fundamentals Guide contains more information about the
constraints and their use.
Check
Check constraints are used most commonly for data validation. The constraint consists of an
expression or routine that must return TRUE. If it returns FALSE, the property’s value violates the
constraint. Check constraints are defined at the type level when expression_string references two or
more properties. To define a check constraint, use:
CHECK(expression_string)
expression_string can be an expression or a complete user-defined routine. The expression or routine
must be written in Docbasic and return a Boolean value. To specify a routine, you must have
Sysadmin or Superuser user privileges.
For user-written routines, the syntax is:
routine_name
FROM OBJECT (object_id |
PATH folder_path[VERSION version_label])
[LANGUAGE docbasic]
Identify the object using the object’s ID or a folder path for an object. If you use a folder path, you can
include a version label. If you don’t include a version label, the label CURRENT is assumed.
message_string is a string literal that contains the error message you want displayed if the constraint is
violated. You can include two special tokens in the message to parameterize it:
• $value(property_name)
When the message is displayed, the server replaces $value(property_name) with the value entered
by the user in the field associated with the property. The property name must be enclosed in
parentheses with no space before the opening parenthesis. For example:
The security type must be folder or cabinet and you
entered $value(security_val).
• $$
Use $$ in a message string to display a single dollar sign in the message.

Create Type
Component specifications identify which components can operate on objects of the type. The
syntax is:
COMPONENTS (component_id_list)
component_id_list contains one or more entries with the following syntax:
component_classifier=object_id|NONE
component_classifier is a character string that represents a qualified component (a dm_qual_comp
object). It consists of the component’s class name and an acronym for the build technology used to
build the component. For example, an component classifier might be Checkin.ACX or Checkin.HTML.
object_id is the object ID of the qualified component represented by the classifier. If you specify NONE
instead of an object ID, the component is not available for the type.
Related statements
Alter Type, page 51
Drop Type, page 98
Examples
The following example creates a new subtype called employee, sets the label text for most properties
and constraints in the property modifier list and the type modifier list, and publishes the data
dictionary information.
CREATE TYPE "employee"
(
"emp_ssn" string(10)
(CHECK ('Len(emp_ssn)=10'
LANGUAGE docbasic)
REPORT 'The SSN must have exactly 10 digits.'
ENFORCE BY APPLICATION,
SET "label_text"='Social Security Number'),
"emp_first_name" string(32)
(SET "label_text"='First Name',
NOT NULL),
"emp_middle_name" string(32)
(SET "label_text"='Middle Name'),
"emp_last_name" string(32)
(SET "label_text"='Last Name',
NOT NULL),
"emp_disambiguator" integer,
"emp_department" integer
(
NOT NULL,
SET "label_text"='Department Code')
)
WITH SUPERTYPE NULL

Create Type
UNIQUE ("emp_first_name","emp_middle_name",
"emp_last_name","emp_disambiguator"),
SET "label_text"='Employee Record'
PUBLISH
The following example creates a subtype of dm_document and publishes its information in the
data dictionary:
CREATE TYPE "legal"
("lawyer" CHAR(30),"case_number" INT,
"defendants" CHAR(30) REPEATING)
WITH SUPERTYPE "dm_document" PUBLISH
The following example creates a user-defined type with no supertype:
CREATE TYPE "my_base_type"
("author" CHAR(30), "title" CHAR(145), "doc_id" ID)
WITH SUPERTYPE NULL
The following example creates a subtype of the user-defined type my_base_type:
CREATE TYPE "acctg" ("accounting" CHAR(30) REPEATING)
WITH SUPERTYPE "my_base_type"

Delete
Delete
Purpose
Removes rows from a registered table.
Syntax
DELETE FROM table_name WHERE qualification
Arguments
Table 18. DELETE argument descriptions
table_name Identifies the registered table from which you are removing
rows. Use the name of table in the underlying RDBMS.
qualification Defines the conditions used to restrict the rows that are
deleted. Refer to The WHERE clause, page 142 for a
description of the valid forms of a qualification.
Return value
The DELETE statement returns a collection whose result object has one property, rows_deleted, that
contains the number of rows deleted.
Permissions
To delete a row, the following conditions must be true:
• Your object-level permission for the dm_registered object in the repository that represents the
RDBMS table must be at least Browse.
• Your table permission for the dm_registered object that represents the table must be
DM_TABLE_DELETE.
• The user account under which Content Server is running must have the appropriate RDBMS
permission to delete from the specified table. (The actual name of this permission will depend on
your RDBMS.)
(For more information about security and object-level and table permissions, refer to the EMC
Documentum Content Server Administration and Configuration Guide.)

Delete
Description
The DELETE statement deletes rows from a registered table. A registered table is a table in the
underlying RDBMS that has been registered with Content Server. (Refer to Register, page 110 for
information about registering tables.) All rows for which the WHERE clause qualification evaluates
to TRUE are deleted.
Related statements
Insert, page 107
Register, page 110
Select, page 116
Unregister, page 153
Update, page 155
Examples
The following example deletes the rows in the registered table authors_table that contain the name of
any author who is not also found in the authors property of a document:
DELETE FROM "authors_table"
WHERE NOT EXISTS (SELECT * FROM "dm_document"
WHERE ANY "authors" = authors_table.name)

Delete...Object
Delete...Object
Purpose
Deletes objects from the repository.
Syntax
DELETE [PUBLIC]type_name[(ALL)]
[correlation_variable]
[WITHIN PARTITION (partition_id {,partition_id})
OBJECT[S]
[IN ASSEMBLY document_id [VERSION version_label]
[NODE component_id][DESCEND]]
Arguments
Table 19. DELETE...OBJECT argument descriptions
type_name Identifies the type of object to remove from the repository.
Valid type_names are:
dm_assembly and its subtypes
dm_user and its subtypes
dm_group and its subtypes
dm_sysobject and its subtypes
user-defined types with NULL supertypes and their subtypes
correlation_variable Defines a qualifier for the type name that is used to clarify
property references.
WITHIN PARTITION clause Restricts the statement to objects in particular partitions.
partition_id identifies the object partition. The property,
i_partition, contains the partition value for an object.
IN ASSEMBLY clause Restricts the statement to objects in a particular assembly.
document_id identifies the document with which the assembly
is associated. Use a literal object ID:
ID('object_id')
version_label specifies a particular version of the document.
Use a symbolic or an implicit version label. If you do not
include a version label, the server uses the version identified
by the document_id argument.

Delete...Object
component_id specifies a particular component in the
assembly. Including the NODE option restricts the statement
to the specified component. Use a literal object ID to identify
the component:
ID('object_id')
The DESCEND keyword directs the server to delete not
only all directly contained node components, but also any
indirectly contained components that meet the criteria. If
you do not include this keyword, the server deletes only the
directly contained components that meet the criteria in the
statement.
SEARCH clause Restricts the statement to objects that meet the SEARCH
clause fulltext search condition. Refer to The SEARCH clause,
page 140 for a detailed description of the SEARCH clause.
WHERE clause Restricts the statement to objects that meet the qualification.
The qualification is an expression that evaluates to TRUE
or FALSE. Refer to The WHERE clause, page 142 for a
description of the valid forms of a qualification.
Return value
The DELETE...OBJECT statement returns a collection whose result object has one property,
objects_deleted, that contains the number of objects deleted.
Permissions
You must have Delete permission on an object to delete the object.
Description
This section contains information about using this statement.
General notes
Deleting objects is subject to the following conditions:
• The object cannot belong to a frozen assembly or a frozen or immutable virtual document.
• If the compound_integrity property in the server’s server config object is set to TRUE, the object
cannot belong to any virtual document, regardless of the whether the document is immutable
or not.
The type_name argument specifies the type of object to delete. The server searches all objects of
the specified type and any subtypes for objects that meet any additional criteria you define in the
statement.

Delete...Object
The keyword PUBLIC restricts the query to those objects with the r_is_public property set to TRUE.
If the query contains a SEARCH clause, the full-text search is restricted to those documents for
which ISPUBLIC is TRUE. When the server queries or searches only public objects, it uses only
the setting of r_is_public for security checks.
The keyword ALL directs the server to consider all versions of each object. If you do not include
ALL, the server only considers versions with the symbolic label CURRENT. You must enclose ALL
in parentheses.
After the server finds all objects that meet the defined criteria, it deletes those for which you have
Delete permission. If any of the objects that are otherwise eligible for deletion belong to a frozen
assembly or an unchangeable virtual document, then the entire statement is rolled back and no
objects are deleted.
The optional clauses, WITHIN PARTITION, IN ASSEMBLY, SEARCH, and WHERE, restrict the
statement’s scope. The WITHIN PARTITION clause restricts the operation to particular partitions.
The IN ASSEMBLY clause restricts the operation to a particular virtual document assembly or node
(component) within the virtual document. The SEARCH clause restricts the operations to indexed
objects that meet the fulltext search condition. The WHERE clause restricts the operations to objects
that meet the specified qualification. The clauses are applied in the following order:
• SEARCH
• IN ASSEMBLY
• WHERE
You cannot use DELETE...OBJECT to destroy record objects. Use the Destroy API command instead.
The IN ASSEMBLY clause
Including the IN ASSEMBLY clause generates an error if the compound_integrity property in the
server’s server config object is set to TRUE. This property controls whether objects contained in a
virtual document may be deleted from the repository.
If the document identified by document_id does not have an associated assembly, the statement
fails with an error.
Related statements
Examples
This example deletes all old documents owned by Joe:
DELETE "dm_document" OBJECTS
WHERE "r_modify_date" < date('01/01/1970')
AND "owner_name" = 'joe'
The following statement deletes all documents that contain the word yeast but not the word rolls:
SEARCH DOCUMENT CONTAINS 'yeast' AND NOT'rolls'

Delete...Object
This statement deletes all documents that contain the words yeast and bread and those that contain
the words cake and wedding:
SEARCH DOCUMENT CONTAINS 'yeast' AND 'bread'
OR 'cake' AND 'wedding'
The following statement deletes all workflows that have either Janine or Jeremy as a supervisor:
DELETE "dm_workflow" OBJECTS
WHERE "supervisor_name" = 'janine' OR
"supervisor_name" = 'jeremy'
This example also deletes all workflows that have Janine or Jeremy as a supervisor:
DELETE "dm_workflow" OBJECTS
WHERE "supervisor_name" IN ('janine','jeremy')

Drop Group
Drop Group
Purpose
Removes a group from the repository.
Syntax
DROP GROUP group_name
Syntax description
Table 20. DROP GROUP syntax description
group_name Identifies the group to remove from the repository. You can
specify the name as an identifier or as a character string literal.
Permissions
You must be the owner of a group or have Superuser user privileges to drop a group.
General notes
When you drop a group, Content Server also removes any registry objects that identify the group as
the subject of an audit or event notification request.
Related statements
Alter Group, page 45
Create Group, page 72
Example
This example drops the group called rons_baby_gift:
DROP GROUP rons_baby_gift

Drop Type
Drop Type
Purpose
Removes a user-defined object type from the repository.
Syntax
DROP TYPE type_name
Arguments
Table 21. DROP TYPE argument descriptions
type_name Identifies the object type to remove.
Permissions
You must own the type or have Superuser privileges to drop a type.
Description
The type you identify must meet the following conditions:
• No objects of the type can exist in the repository.
• The type cannot have any subtypes.
Dropping a type also removes data dictionary information for the type and its properties. The
information is removed when one of the following occurs:
• The IDfSession.publishDataDictionary method is executed for the entire repository.
• The Data Dictionary Publisher job runs.
Note: After you drop a type, you may not be able to create a type by the same name immediately.
Allow time for the system to synchronize the type cache before attempting to recreate the type.
Related statements
Alter Type, page 51
Create Type, page 77
Examples
DROP TYPE "my_base_type"

Execute
Execute
Purpose
Executes administration methods.
Syntax
EXECUTE admin_method_name [[FOR] object_id]
[WITH argument = value {,argument = value}]
Arguments
Table 22. EXECUTE argument descriptions
Argument Description
admin_method_name Specifies which administration method you want to execute.
Table 23, page 100 lists the methods.
Administration method names are not case sensitive.
object_id Identifies the object on which you want the function to
operate. Use the object’s object ID.
WITH clause Defines one or more arguments and their values for the
administration method.
Valid arguments differ for each method. Refer to Chapter
3, Administration Methods, for a description of each
administration method and its arguments.
Return value
The EXECUTE statement returns a collection. The properties of the query result object in the
collection depend on which administration method was executed. Refer to the description of each
method in Chapter 3, Administration Methods, for details.
Permissions
The privileges required depend on the administration method you are executing. Refer to the
description of the individual method in Chapter 3, Administration Methods, for information.

Execute
Description
The EXECUTE statement is the DQL equivalent of the IDfSession.apply method. You can use
EXECUTE to invoke any of the administration methods that you can invoke using the apply method
except PING and WEBCACHE_PUBLISH. These cannot be executed using the EXECUTE statement.
Table 23, page 100, lists the administration methods that you can invoke with the EXECUTE statement
and briefly describes the task that each performs.
Table 23. Administration methods by category for the EXECUTE statement
Category of Operation Function Description
Process Management CHECK_SECURITY, page
181
Checks a user or group’s
permissions level for one or more
objects.
GET_INBOX, page 216 Returns items in user’s Inbox.
MARK_AS_ARCHIVED,
page 248
Sets the i_is_archived
property of a dm_audittrail,
dm_audittrail_acl, or
dm_audittrail_group object
to T.
PURGE_AUDIT, page 287 Deletes audit trail entries from
the repository.
RECOVER_AUTO_TASKS,
page 296
Recovers work items claimed by
a workflow agent master session
but not yet processed.
ROLES_FOR_USER, page
308
Returns the roles assigned to a
user in a particular client domain.
Execute procedures DO_METHOD, page 192 Executes system-defined
procedures such as lpq or
who or user-defined procedures.
HTTP_POST, page 224 Directs the execution of a method
to an application server.
Content storage management CAN_FETCH, page 173 Determines whether content
in a distributed storage area
component can be fetched by the
server.
CHECK_RETENTION_
EXPIRED, page 177
Finds SysObjects in
content-addressed storage
that have an expired retention
period or no retention period.

Execute
CLEAN_LINKS —
Deprecated, page 184
This method is deprecated.
DFC Version 6 does not
support linked storage
areas or link record objects.
Consequently, linked storage
areas, link records and this
supporting method are
deprecated.
Provides maintenance for linked
store storage areas.
On Windows platforms, cleans up
unneeded link record objects and
resets file storage object security.
On UNIX platforms, cleans up
unneeded linkrecord objects,
directories, and links associated
with linked storage areas.
DELETE_REPLICA , page
188
Removes a replica from a
distributed storage area.
DESTROY_CONTENT, page
190
Removes a content object and its
associated file from the repository.
(Do not use this for archiving; use
PURGE_CONTENT instead.)
GET_FILE_URL, page 214 Returns the URL to a content file.
GET_PATH, page 221 Returns the path to a particular
content file in a particular
distributed storage area
component.
IMPORT_REPLICA, page
228
Imports an external file as a
replica of content already in the
repository.
MIGRATE_CONTENT, page
251
Moves content files from one
storage area to another.
PURGE_CONTENT, page
293
Deletes a content file from a
storage area. (Used as part of the
archiving process.)
PUSH_CONTENT_ATTRS,
page 294
Sets the content metadata in a
content-addressed storage system
for a document stored in that
storage system.
REGISTER_ASSET, page 298 Queues a request for the creation
of a thumbnail, proxies, and
metadata for a rich media content
file. The request is queued to
Media Server.
This method is only available or
useful if you have Documentum
Media Transformation Services
running.

Execute
REPLICATE, page 303 Copies content in one component
of a distributed storage area to
another area.
RESTORE_CONTENT, page
306
Moves a file or files from archived
storage to the original storage
location.
SET_CONTENT_ATTRS,
page 313
Sets the content_attr_name and
content_attr_value properties in
the content object associated with
the content file.
SET_STORAGE_STATE,
page 320
Sets the state of a storage area to
off-line, on-line, or read-only.
TRANSCODE_CONTENT,
page 324
Queues a request for a content
transformation to Media Server.
running.
Database Methods DB_STATS, page 186 Provides database operation
statistics for a session.
DROP_INDEX, page 199 Drops an index.
EXEC_SQL, page 203 Executes SQL statements.
EXPORT_TICKET_KEY,
page 205
Exports a repository’s login ticket
key.
FINISH_INDEX_MOVES,
page 207
Completes an interrupted object
type index move operation.
GENERATE_PARTITION_
SCHEME_SQL —
Creates an SQL script to partition
a repository.
IMPORT_TICKET_KEY,
page 230
Imports a login ticket to the
repository.
MAKE_INDEX, page 245 Creates an object type index.
MOVE_INDEX, page 275 Moves an object type index from
one tablespace to another.
Note: Not supported on DB2.
PARTITION_OPERATION,
page 277
Partitions a repository.
REORGANIZE_TABLE,
page 301
Reorganizes a database table for
query performance.
RESET_TICKET_KEY, page
305
Generates a login ticket key for
the repository.

Execute
UPDATE_STATISTICS, page
327
Updates the statistics for a
database table.
Full-Text Methods ESTIMATE_SEARCH, page
201
Returns the number of results
matching a particular SEARCH
condition.
MARK_FOR_RETRY, page
249
Finds all queue items representing
objects that failed indexing and
marks them as pending indexing.
MODIFY_TRACE, page 273 Sets the tracing level for full-text
querying operations.
Session Management CHECK_CACHE_CONFIG,
page 174
Requests a consistency check on a
particular cache config object.
GET_LAST_SQL, page 219 Returns the last SQL statement
issued.
GET_SESSION_DD_
LOCALE, page 223
Returns the locale in use for the
current session.
LIST_AUTH_PLUGINS,
page 232
Lists the authentication plugins
loaded by Content Server.
LIST_RESOURCES, page 233 Provides information about
the server operating system
environment.
LIST_SESSIONS, page 237 Provides information about
current, active sessions.
LIST_TARGETS, page 240 Lists the connection brokers
defined as targets for the server.
The information is returned
in a collection with one result
object whose properties list the
connection brokers defined as
targets for the server.
LOG_ON, page 243 and
LOG_OFF, page 242
Turn server logging of
information about RPC calls
on or off.
SET_APIDEADLOCK, page
310
Sets a deadlock trigger on
a particular API method or
operation.
SET_OPTIONS, page 317 Turn various tracing options on
or off.
SHOW_SESSIONS, page 322 Provides information about
current, active sessions and
a user-specified number of
timed-out sessions.

Execute
The EXECUTE statement is not case sensitive. Also, you do not have to specify the datatype of the
arguments for each function. The statement determines the datatype from the value you assign
to the argument.
Examples
Refer to the individual descriptions of the method in Chapter 3, Administration Methods, for
examples.

Grant
Grant
Purpose
Gives one or more user privileges to one or more users.
Syntax
GRANT privilege {,privilege} TO users
Arguments
Table 24. GRANT argument descriptions
privilege Identifies the privilege you want to grant. Valid privileges are
SUPERUSER
SYSADMIN
CREATE TYPE
CREATE CABINET
CREATE GROUP
CONFIG AUDIT
PURGE AUDIT
VIEW AUDIT
users Identifies the users to whom you want to a privilege or
privileges. The user name must belong to an individual user.
You can specify one or more users as a comma-separated list
of user names or use a SELECT statement to identify the user
names. (Refer to Examples, page 106, for the use of a SELECT
statement.)
The user name must be the value is found in the user_name
property of the dm_user object associated with the user.
Permissions
To grant Superuser or Sysadmin user privileges, you must have Superuser privileges.
To grant Create Group, Create Type, or Create Cabinet user privileges, you must have Superuser
or Sysadmin user privileges.
To grant Config Audit, Purge Audit, Or View Audit privileges, you must be the repository owner or a
Superuser. Repository owners and Superusers cannot grant these privileges to themselves.

Grant
Description
Granting a privilege to a user who already has that particular privilege does not generate an error.
Related statements
Revoke, page 114
Examples
The following example grants the Create Type user privilege to the users donna and carol:
GRANT CREATE TYPE TO donna,carol
This example grants the Create Cabinet user privilege to all individual users (that is, to those user
names that do not represent a group):
GRANT CREATE CABINET TO
WHERE "r_is_group" = FALSE)
The final example grants two privileges to the users jim and mike:
GRANT SYSADMIN,SUPERUSER TO jim,mike

Insert
Insert
Purpose
Inserts a row into a registered table.
Syntax
INSERT INTO table_name [(column_name {,column_name})]
VALUES (value {,value}) | dql_subselect
Arguments
Table 25. INSERT argument descriptions
table_name Identifies the registered table in which you want to insert
new rows.
column_name Identifies a column in the table to receive an assigned value.
value Specifies the value assigned to a column. The number of
values specified must equal the number of columns named or
the number of columns in the table. Refer to Assigning values
to columns, page 108 for more information.
dql_subselect Specifies a SELECT statement. The returned values are
inserted into the table.
Return value
The INSERT statement returns a collection whose result object has one property, rows_inserted, that
contains the number of rows inserted into the table.
Permissions
To use the INSERT statement, the following conditions must be true:
• Your object-level permission for the dm_registered object representing the RDBMS table must be
at least Browse.
• Your table permission for the dm_registered object representing the table must be
DM_TABLE_INSERT.
permission to insert data into the specified table. (The actual name of this permission will depend
on your RDBMS.)
For more information about security and object-level and table permissions, refer to the EMC
Documentum Content Server Administration and Configuration Guide.

Insert
Description
Use the information in this section to execute this statement.
Assigning values to columns
Which value is inserted in each column you specify is determined by position. That is, the first
column you specify in the statement receives the first value. The second column receives the second
value, and so forth. Consequently, if you are inserting a value in every column, it is not necessary to
specify the columns; the server automatically inserts the values in each column in turn. However, if
you omit the column names, the number of values must equal the number of columns in the table.
If you specify column names, you can insert values into a subset of the table’s columns. Also, it is not
necessary to specify the column names in the same order in which they appear in the table. Columns
that are not specified in the INSERT statement receive default values. (Refer to the documentation
for your underlying RDBMS for information about the default values assigned to columns in this
situation. Different systems have different rules. For example, some database management systems
only allow you to default nullable columns. In such systems, all non-nullable columns must be
specified.)
Defining values
There are two possible ways to define the values to assign to columns. You can use the VALUES
clause or you can use a DQL subselect statement.
The VALUES clause has the syntax:
VALUES value {,value}
One value must be specified for each column named in the statement. Each value must be a literal
value appropriate for the datatype of the column to which it is being assigned. For example, assume
there is a registered table called customer_accounts and that you want to insert values into four
of its columns: customer_name, acct_number, open_date, and balance. The customer_name and
acct_number columns are character string datatypes. The open_date column is a date datatype,
and the balance column is a floating point datatype. The following statement inserts values into
these columns:
INSERT INTO "customer_accounts" ("customer_name","account_number","open_date,
balance") VALUES ('Henrietta Hornsby','01264',date('4/2/1993'),125.00)
The value Henrietta Hornsby is inserted into the customer_name column, the value 01264 is inserted
into the account_number column, and so forth.
Similarly, when you use a subselect statement, the returned values must be equal in number to the
number of columns named in the INSERT statement and the values must be appropriate for the
column into which they will be inserted. (Refer to Select, page 116 for the subselect statement’s
syntax.)
Related statements
Delete, page 91
Register, page 110
Select, page 116
Update, page 155

Insert
Examples
This example saves the object names and creation dates of the contents of the flowers folder into
the objects table:
INSERT INTO "objects" ("o_name", "c_date")
SELECT "object_name", "r_creation_date"
FROM "dm_document"
WHERE FOLDER ('/public/subject/flowers')

Register
Register
Purpose
Registers a table from the underlying RDBMS with the repository.
Syntax
REGISTER TABLE [owner_name.]table_name
(column_def {,column_def})
[[WITH] KEY (column_list)]
[SYNONYM [FOR] 'table_identification']
Arguments
Table 26. REGISTER argument descriptions
owner_name Identifies the table’s owner.
Use the owner’s RDBMS user name. If the owner is a
repository user, this value may be found in the user’s
user_db_name property.
If the RDBMS is Oracle or DB2 and the owner is the DBA,
you can use the alias dm_dbo.
If the RDBMS is MS SQL Server and the owner is the DBA,
you can use either dbo or dm_dbo as an alias.
owner_name is optional if the current user is the table’s owner.
The default value is the current user.
table_name Identifies the RDBMS table to register with the repository.
You can specify the table’s actual name or a synonym. The
name must consist of ASCII characters.
If you specify the table’s actual name, do not include the
SYNONYM clause in the statement.
For Oracle or DB2, if you specify a synonym, that synonym
must be previously defined through the RDBMS. Including
the SYNONYM clause in the REGISTER statement is optional.
For MS SQL Server, if you specify a synonym, you must
include the SYNONYM clause in the statement.
For MS SQL Server, you cannot register a remote table.

Register
(Refer to The SYNONYM clause, page 112 for details.)
If you do not have Superuser user privileges, you must be
the owner of the table.
column_def Describes columns in the table. The format for a column
definition is:
column_name datatype [(length)]
where column_name is the name of the column in the table and
datatype is the DQL datatype that corresponds to the column’s
RDBMS datatype. The column name must consist of ASCII
characters. Valid datatypes in a column definition are:
float, double
integer, int
tinyint
smallint
char, character, string
date, time
You must also specify a length for columns that have a
character, char, or string datatype (for example, char(24)).
column_list Identifies the columns in the table on which indexes have
been built. Use a comma-separated list of column names.
table_ identification The name of the table in the underlying RDBMS.
You must include the SYNONYM clause if you run against
MS SQL Server and you specify a synonym as the table name
in the statement.
Including the SYNONYM clause is optional if you are
running against Oracle or DB2 and specify a synonym as the
table name in the statement.
(The SYNONYM clause, page 112 contains the details.)
Return value
When issued through IDQL, the REGISTER statement returns the object ID of the dm_registered
object for the table. If you issue the statement through IAPI (using the Query method), the statement
returns a collection whose result object has one property, called new_object_ID, that contains the
object ID of the dm_registered object for the table.
Note: IAPI is a deprecated feature.

Register
Permissions
To register a table, you must own the table or have Superuser user privileges. If folder security is
enforced in the repository, you must also have at least Write permission on the System cabinet.
Description
This section contains usability notes.
General notes
When you execute the REGISTER statement, the server creates an object of type dm_registered (a
SysObject subtype) that represents the RDBMS table in the repository. This object is automatically
linked to the system (/system) cabinet.
A dm_registered object can be manipulated like any other SysObject or SysObject subtype with one
exception: you cannot version a dm_registered object.
You do not have to include all of the columns in the underlying RDBMS table in the statement. You
can specify a subset of the underlying table’s columns. The column definitions you provide need not
match the column definitions for the underlying table. When the server creates the dm_registered
object for the table, it uses your column definitions.
The REGISTER statement automatically assigns a default ACL to the dm_registered object. (The
default is determined by the value in the default_acl property of the server’s server config object.)
The statement also sets default table permissions. The table permissions are set to SELECT for the
owner. The group and world are not given any default table permissions. You can change these by
setting the properties directly. Note that table permissions for registered tables are not hierarchical.
(For a complete description of registered table permits, refer to the EMC Documentum Content Server
Administration and Configuration Guide.)
Changes to the definition of an underlying table are not automatically reflected in its corresponding
dm_registered object. If the underlying table is modified and you want the dm_registered object to
match the underlying table definition, you must unregister the table and then register it with new
column definitions.
The SYNONYM clause
Use the SYNONYM clause to register RDBMS tables that have synonyms in the underlying tables. (A
synonym for an RDBMS table must be created independently in the RDBMS before you register the
table. The Register statement does not create a synonym.)
The SYNONYM clause records the actual name of the table that corresponds to the table’s synonym.
The name is stored in the synonym_for property of the table’s dm_registered object.
When you run against Oracle or DB2, Content Server passes the synonym directly to the database
server. The actual table name, as specified in the SYNONYM clause and recorded in the synonym_for
property is purely informational. For example, suppose johndoe has a table called myremotetable
in the remote Oracle database londonremote, and that this table has the synonym remote1. To
register this table, he uses:
REGISTER TABLE johndoe."remote1" ("columnA" int)
SYNONYM FOR johndoe.myremotetable@londonremote

Register
After he registers the table, he can use the synonym in DQL statements and it is passed directly to the
database server. For example, issuing the following DQL:
SELECT * FROM johndoe."remote1"
generates the following SQL:
When you run against MS SQL Server, Content Server substitutes the value you specified in the
SYNONYM clause (recorded in the synonym_for property) for the table name in the SELECT,
INSERT, UPDATE and DELETE statements. For example, suppose johndoe issues the following
REGISTER statement:
REGISTER TABLE johndoe."remote1" ("columnA" int)
SYNONYM FOR londonserver.londonremote.johndoe.myremotetable
After he registers the table, he issues the following SELECT statement:
Content Server substitutes the actual table name for the synonym and the following SQL is generated:
SELECT * FROM londonserver.londonremote.johndoe.myremotetable
Special note for Oracle platforms
If you create a database link between the database on which the Documentum repository is installed
and another database, and then the use SYNONYM feature to obtain information from that second
database, both databases must be in the same codepage.
Related statements
Delete, page 91
Insert, page 107
Update, page 155
Examples
The following statement registers the RDBMS table named departments:
REGISTER TABLE "departments"
("dept_name" CHAR(30), "dept_code" INT)
KEY ("dept_code")

Revoke
Revoke
Purpose
Removes one or more user privileges from one or more users.
Syntax
REVOKE privilege {,privilege} FROM users
Arguments
Table 27. REVOKE argument descriptions
privilege Specifies the privilege to revoke. Valid privileges are:
SUPERUSER
SYSADMIN
CREATE TYPE
CREATE CABINET
CREATE GROUP
CONFIG AUDIT
PURGE AUDIT
VIEW AUDIT
users Specifies the users from whom you are revoking the specified
privilege or privileges. The user name must belong to
an individual user. You can specify one or more users as
a comma-separated list of user names or use a SELECT
statement to identify the user names. (Refer to Examples,
page 115 for the use of a SELECT statement.)
The user name must be the value is found in the user_name
property of the dm_user object associated with the user.
Permissions
To revoke the Sysadmin or Superuser user privilege, you must have Superuser user privileges.
To revoke the Create Group, Create Type, or Create Cabinet user privilege, you must have either
Superuser or Sysadmin user privileges.
To revoke Config Audit, Purge Audit, Or View Audit privileges, you must be the repository owner or
a Supersuser. Repository owners and Superusers cannot revoke these privileges from themselves.

Revoke
Description
The statement succeeds even if a user does not have the privilege being revoked.
Related Statements
Grant, page 105
Examples
The following example revokes the Create Cabinet and Create Type privileges from the users john
and howard:
REVOKE CREATE CABINET, CREATE TYPE FROM john, howard
The next example demonstrates the use of a subselect statement to identify the users. This example
revokes the Superuser user privilege from all users except haroldp:
REVOKE SUPERUSER FROM
WHERE "user_privilege" >= 16 AND "user_name" != 'haroldp')

Select
Select
Purpose
Retrieves information from the repository, the database, or both.
Syntax
A SELECT statement may be either a standard SELECT or an FTDQL SELECT statement. The syntax
for an FTDQL SELECT statement is a subset of the standard syntax. For a brief description of an
FTDQL SELECT statement, refer to FTDQL SELECT statements, page 121.
The syntax for a standard SELECT statement is:
SELECT [FOR base_permit_level][ALL|DISTINCT] value [AS name] {,value [AS name]}
FROM [PUBLIC] source_list
[WITHIN PARTITION (partition_id{,partition_id})
| IN DOCUMENT clause
| IN ASSEMBLY clause]
[SEARCH [FIRST|LAST]fulltext_search_condition
[IN FTINDEX index_name{,index_name}]
[GROUP BY value_list]
[HAVING qualification]
[UNION dql_subselect]
[ORDER BY value_list]
[ENABLE (hint_list)]
The syntax for an FTDQL SELECT statement is:
[SEARCH fulltext_search_condition
[ORDER BY SCORE]
Arguments
Table 28. SELECT argument descriptions
base_permit_level Defines a minimum permission level for the returned objects.
If specified, the query returns only those objects for which the
user has at least the specified permit level.
Valid values are: NONE, BROWSE, READ, NOTE, VERSION,
WRITE, and DELETE.
If unspecified, the default is BROWSE.
There are no FTDQL constraints on this value.

Select
value Identifies the information to retrieve. Valid values are:
• Property and column names
• Scalar, aggregate, and date functions
• MFILE_URL function
• Arithmetic expressions
• The keywords CONTAIN_ID, CONTENTID, DEPTH,
ISCURRENT, ISPUBLIC, ISREPLICA, OBJTYPE,
PARENT, SCORE, SUMMARY, SYSOBJ_ID, TEXT,
THUMBNAIL_URL, USER
• An asterisk (*)
Multiple values can appear in any order. If the SELECT
statement is a subselect or a subquery, you can only include
one value.
There are constraints on the values in the selected values
list for FTDQL queries. For details, refer to the Usage Notes
sections describing each kind of selected value.
AS name Defines a name for the result object property that will contain
the retrieved value. This argument is optional. If you omit it,
the system provides a default name. (Refer to The AS clause,
page 132 for information about the default name.)
name must consist of ASCII characters
You may include the AS name option in an FTDQL SELECT
statement.
source_list Identifies the object types and RDBMS tables to search. You
can specify any combination of object types, RDBMS tables,
and inline views in a standard SELECT statement.
Types are specified using the following syntax:
type_name [(ALL)|(DELETED)|(LITE)] [correlation_
variable] [WITH passthrough_hint_list]
type_name cannot reference an object type that represents
aspect properties.
Tables are specified using the following syntax:
[owner_name.]table_name [correlation_variable]
[WITH passthrough_hint_list]

Select
passthrough_hint_list is a list of hints for one or more
databases. The hint list for an individual database has the
following format:
db_keyword('hint'{,'hint'})
Valid db_keywords are: ORACLE, SQL_SERVER, and DB2.
hint is any valid hint accepted by the particular RDBMS.
If you include hints for multiple databases, the hint lists for
each must be separated by commas and the entire set (for all
databases) enclosed in parentheses. Passthrough hints, page
363 describes using passthrough hints fully.
On Oracle and DB2, a passthrough hint in the source list in a
subquery is applied to the entire SELECT statement.
The RDBMS table must be a registered table unless the user
issuing the SELECT is a superuser. Only superusers can
query unregistered RDBMS tables. Additionally, there are
several constraints on specifying more than one type in the
source list.
Inline views are specified using the following syntax:
( SELECT ... ) [correlation_variable]
An inline view is a subquery statement. In other words, it
is a SELECT statement enclosed in parenthesis. The results
from the subquery act like a temporary table in the select list.
An optional correlation_variable can be specified, just as with
tables or types.
For complete information about specifying the source_list
clause, including LEFT OUTER JOIN, refer to Source list,
page 133. For a description of the constraints imposed on the
source list by an FTDQL query, refer to FTDQL constraints
on the source list, page 137.

Select
IN DOCUMENT clause Identifies the target of the SELECT statement as a particular
virtual document. This clause can assemble a virtual
document or identify the components of a virtual document.
If you include this clause, you cannot specify the IN
ASSEMBLY clause. Refer to The IN DOCUMENT clause, page
137 for the syntax and usage of the IN DOCUMENT clause.
This clause may not be included in an FTDQL SELECT
statement.
IN ASSEMBLY clause Identifies the target of the SELECT statement as an assembly
of a virtual document. The server uses the components of the
assembly as the definition of the virtual document and applies
any other specified conditions to those components, rather
than searching the document’s hierarchy for components.
This clause may not be included in an FTDQL SELECT
statement.
fulltext_search_condition Defines criteria for searching of the full-text index. Refer to
The SEARCH clause, page 140 for a description of its syntax
and use.
You can include a fulltext search condition in an FTDQL
SELECT statement.
qualification Restricts returned results to objects meeting the conditions in
the qualification. Refer to The WHERE clause, page 142 for
a full description of the valid forms of a qualification for a
WHERE clause or a HAVING clause.
dql_subselect Defines an additional DQL SELECT statement. The columns
returned by the statement must correspond to those returned
by the outermost SELECT statement in number and datatype.
A dql_subselect may not be included in an FTDQL SELECT
statement.
value_list Defines an order in which to group or sort the returned
results. The values in this list must also be selected values.
Refer to The GROUP BY clause, page 146 for the specific use
of this in each clause.
A GROUP BY clause may not be included in an FTDQL
SELECT statement.
hint_list One or more standard or passthrough DQL hints.
Refer to the description of the source list for the format of a
passthrough hint.
Standard hints are the following:

Select
CONVERT_FOLDER_LIST_TO_OR
FETCH_ALL_RESULTS N
FORCE_ORDER
FTDQL and NOFTDQL
FT_CONTAIN_FRAGMENT
GROUP_LIST_LIMIT N
HIDE_SHARED_PARENT
IN and EXISTS
OPTIMIZE_ON_BASE_TABLE
OPTIMIZE_TOP N
RETURN_RANGE
RETURN_TOP N
ROW_BASED
SQL_DEF_RESULT_SET N
TRY_FTDQL_FIRST
UNCOMMITTED_READ
DM_LEFT_OUTER_JOIN_FOR_ACL
N is an integer value and level_1 and level_2 are optimization
levels.
For a brief description of the standard hints, refer to Table 29,
page 149. The ROW_BASED hint affects the selected values
list and the WHERE clause qualification. For information
about its effects, refer to the descriptions of those portions of
the syntax. Information about ROW_BASED is also found
in Appendix A, Using DQL Hints, which contains full
information about all the standard hints.
ROW_BASED may not be included in an FTDQL query, nor
may it be used in queries that reference a lightweight object
type in the FROM clause. All other hints are acceptable in
FTDQL queries or in queries against a lightweight object type.
Description
This section contains usability information for the Select statement.
General notes
The SELECT statement retrieves information from object types and RDBMS tables. By default, the
statement returns only objects for which the user has at least Browse permission. If you want to
enforce a higher level of permission on the returned objects, you can include the FOR base_permit_level
clause. For example, suppose you issue the following SELECT statement:
SELECT FOR VERSION "r_object_id","object_name","owner_name"
FROM "dm_document" WHERE "subject"='budget_proposal'

Select
The query returns all budget proposal documents for which the currently logged-in user has at least
Version permission.
You can execute the statement as a standalone statement and indirectly, as part of a variety of other
DQL statements. For example, the following CREATE GROUP statement uses a SELECT statement to
select the users that will populate the new group:
CREATE GROUP supers MEMBERS
(SELECT "user_name" FROM "dm_users"
WHERE "user_privilege" >= 16)
(For more information about CREATE GROUP, refer to Create Group, page 72 .)
FTDQL SELECT statements
An FTDQL SELECT statement is a SELECT statement whose syntax conforms to a particular set of
rules that allow the query to be executed directly against the fulltext index. If the statement conforms
to FTDQL syntax, the statement is run solely against the fulltext index; the repository is not queried.
This feature provides enhanced performance for the query. The general syntax accepted for an
FTDQL query is shown in the formal syntax description. The syntax constraints on particular clauses
enforced for an FTDQL query are listed in the Usage Notes sections describing each clause. (A
summary of the rules is found in Table 119, page 393.)
An FTDQL SELECT statement must be a standalone statement. That is, SELECT statements
embedded as a subselect in another DQL statement are never executed as FTDQL SELECT statements.
Nor are unioned SELECT statements executed as FTDQL SELECT statements. All FTDQL SELECT
statements must include one of the following:
• A SEARCH clause
• The keyword SCORE, SUMMARY, or TEXT in the list of selected values
• The DQL hint ENABLE(FTDQL)
A query that conforms to the FTDQL syntax rules is automatically executed as an FTDQL query.
If you are wondering whether a particular query conforms to the rules, you can include the
ENABLE(FTDQL) hint in the query. If the query conforms, the hint has no effect and the query is
executed as an FTDQL query. If the query does not conform, Content Server returns an error.
If you do not want a particular query to execute as an FTDQL query even though it conforms to the
FTDQL syntax, include the ENABLE(NOFTDQL) in the query.
A standard query queries both the fulltext index and the database. Such a query typically contains a
SEARCH clause and a WHERE clause. If the WHERE clause is not compliant with the rules of
FTDQL or the query contains an explicit ENABLE(NOFTDQL), Content Server executes the SEARCH
clause against the fulltext index and the WHERE clause against the database and then returns the
intersection of the results.
Note: A query that does not conform to the FTDQL syntax and does not include the ENABLE(FTDQL)
hint is executed as a standard SELECT query. It does not return an error.
If an FTDQL query contains references to aspect properties, those properties must be indexed.
Aspect properties are not indexed by default. You must explicitly declare them for indexing. Use
Documentum Application Builder or an ALTER ASPECT statement to do so.
FTDQL metadata queries that contain LIKE predicates with pattern matching or on metadata that
contains underscores in the values may return different results than non-FTDQL queries on the same

Select
metadata. This occurs because the index server processes FTDQL queries against metadata and the
database server processes non-FTDQL queries against metadata.
FTDQL queries are not affected by the distinct_query_results key in the server.ini. Setting this key
to T (TRUE) does not affect how an FTDQL query is processed.
Referencing non-qualifiable properties
Non-qualifiable properties (those stored in a property bag) can be referenced in SELECT statements.
They may appear in the selected values list. However, they may not be referenced in an expression.
Nor may they appear in a WHERE clause unless the query is an FTDQL query.
Referencing aspect properties
To reference an aspect property in a SELECT statement, you must qualify the property name with
the name of the aspect for which it is defined. For example, suppose you have an aspect named
grant_validation, with a property defined for it named grant_amount. To select the grant amount,
reference the property as follows in the selected values list:
grant_validation.grant_amount
For example:
SELECT grant_validation.grant_amount FROM education_grants
....
If the query has multiple object types in the FROM clause, any aspect properties referenced
in the selected values must be also qualified with the type name. For example, assuming that
grant_validation and salary_calc are names of aspects:
SELECT education_grants.grant_validation.grant_amount,
dm_user.research_personnel.salary_calc
FROM education_grants, dm_user
...
The query result column that contains the selected aspect property value is named with the fully
qualified name as specified in the selected values list.
The ALL and DISTINCT keywords
The optional ALL and DISTINCT keywords determine whether the SELECT statement returns
duplicate rows. ALL returns all rows, including any duplicates. DISTINCT does not return duplicate
rows. If neither keyword is included, the default is determined by the server’s default behavior. (The
server’s default behavior is determined by the distinct_query_results flag in the server’s server.ini
start-up file.)
The DISTINCT keyword is ignored if the SELECT statement also includes an IN DOCUMENT or IN
ASSEMBLY clause. The server issues a warning if the statement includes both the DISTINCT keyword
and an IN DOCUMENT or IN ASSEMBLY clause. The warning is also issued if the server’s default
setting is not to return duplicates and the query contains an IN DOCUMENT or IN ASSEMBLY clause.
Selecting property values
Select object property values by specifying the property’s name as a selected value. The properties
you specify must belong to an object type identified in the FROM clause. The properties can be
either single-valued or repeating properties. Repeating properties, page 123, contains guidelines for
selecting repeating property values.

Select
You cannot select any of the following properties if you are querying audit trail entries unless you
have Superuser or View_audit privileges:
• acl_name • object_name
• acl_domain • object_type
• property_list • owner_name
• property_list_id • session_id
• chronicle_id • version_label
The following statement returns the value of title, a single-valued property, for all documents in
the repository:
SELECT "title" FROM "dm_document"
This next example returns the value of authors, a repeating property, from all documents that have
bread as their subject:
SELECT "authors" FROM "dm_document"
WHERE "subject"='bread'
You can use an asterisk (*) to select a predefined set of system-defined properties for the object. Refer
to The asterisk (*) as a selected value, page 131 for details.
Selecting non-qualifiable properties
Non-qualifiable properties can be referenced in the selected values list. However, you can reference
them only as an individual selected value, by name. They cannot be referenced in expressions or
functions, such as an aggregate function, in the selected values list.
Selecting aspect properties
You can select values of properties defined for aspects attached to objects. If the properties are
repeating properties, there are some constraints on the query. For a listing of these, refer to Repeating
properties, page 123.
Repeating properties
Including a repeating property in the selected values list is subject to the following constraints:
• You cannot select a repeating property and the name of a column from a registered table unless
you also include the DQL hint, ROW_BASED.
Note: The ROW_BASED hint may not be included in FTDQL queries or queries that reference a
lightweight object type in the FROM clause.
• You cannot select a repeating property if there are multiple object types identified in the FROM
clause unless you also include the DQL hint, ROW_BASED.

Select
• If the repeating property is an aspect property or from a lightweight object or its shared parent,
the following constraints also apply:
— If the property is used in an expression, all properties referenced in the expression must be
repeating properties from the same aspect or the same lightweight object type.
— If the property is referenced in an aggregate function, you must include a GROUP BY clause
that references r_object_id and only r_object_id.
— The query must include r_object_id in the selected values list.
— The query may not be a subquery.
You can select both repeating and single-valued properties in the same statement. If you do, the
format of the returned results varies depending on how the query is written. The results can be
returned with a separate result row for each value returned for a selected repeating property or
the result rows can be returned in a master-detail format, with each row representing the values,
including all repeating values, for one object.
To return results by object ID
To obtain query result objects that include all selected repeating property values for one object in
one query result object, include the r_object_id property in the selected values list. Additionally, the
source list in the FROM clause may contain only item.
To ensure that the results are ordered in either ascending or descending order, you can include an
ORDER BY clause also. Typically, results are returned in ascending order by default.
Note: Refer to The ORDER BY clause, page 148 for more information about using the ORDER By
clause.
For example, the following statement returns one result object that contains the names of all three
authors in the authors property:
SELECT "r_object_id","object_name","authors"
FROM "dm_document"
WHERE "title" = 'Breads of France'
ORDER BY "r_object_id"
This statement returns the following object:
r_object_id object_name authors
object_id French_breads Jean Connie Corwin
To return each repeating value in an individual result object
To obtain result objects that contain one repeating property value in each object, include the DQL
hint, ROW_BASED, in the query.
For example, suppose you want to obtain the object ID and authors of a document whose object_name
is French Bread, and you want the results in a separate row for each author. The following query
that includes the ROW_BASED hint returns a separate row for each returned value for the authors
property:

Select
SELECT "r_object_id","authors" FROM dm_document
WHERE "title" = 'Breads of France'
ENABLE(ROW_BASED)
Assuming that there are three authors, named Jean, Connie, and Corwin, the returned rows are:
r_object_id authors
0900000334fa21c3 Jean
0900000334fa21c3 Connie
0900000334fa21c3 Corwin
Additionally, if the query has a WHERE clause qualification that references a value in the selected
repeating property, then the query will return only the row that contains that value. For example,
suppose you execute the following statement:
SELECT "r_object_id","title","authors" FROM dm_document
WHERE "authors" = 'JPierpont'
ENABLE(ROW_BASED)
The query returns one row:
r_object_id title authors
0900000334fa21c3 English Muffins JPierpont
FTDQL requirements
The requirements for single or repeating properties in a selected values list for an FTDQL query are
the same as those for a standard query.
Selecting column values
Select column values by specifying column names as values. The columns must belong to a RDBMS
table that you identify in the FROM clause. You can use column names to select some or all of the
column values, or you can use an asterisk to retrieve all column values.
Note: The account from which Content Server was installed must have SELECT privileges on the
underlying table in the RDBMS before you can use DQL to select from an RDBMS table. Additionally,
to query a registered table, you must have DM_TABLE_SELECT table permission and at least Browse
object-level permission for the dm_registered object representing the table. If the SELECT statement
includes a SEARCH clause, your object-level permission must be at least Read.
FTDQL requirements
The requirements for column values in a selected values list for an FTDQL query are the same as
those for a standard query.
Scalar, aggregate, and date functions as selected values
You can include scalar, aggregate, and date functions to manipulate returned values.

Select
A scalar function operates on one value. The DQL SELECT statement accepts three scalar functions:
• UPPER, which returns the uppercase form of a character string
• LOWER, which returns the lowercase form of a character string
• SUBSTR, which returns a subset of a specified character string
An aggregate function operates on a set of values and returns one value. The DQL SELECT statement
accepts five aggregate functions:
• COUNT, which returns the number of values in a group of values
• MIN, which returns the minimum value in a group of values
• MAX, which returns the maximum value in a group of values
• AVG, which returns the average value of a group of values
• SUM, which returns the total of all values in a group of values
For example, the following statement returns the number of documents owned by the user horace:
WHERE "owner_name"='horace'
If an aggregate function references a repeating property from a lightweight object or its shared parent
or from an aspect, the query must also have r_object_id in the selected values list and must include a
GROUP BY clause that references r_object_id and only r_object_id.
An aggregate function cannot reference a non-qualifiable property.
You cannot include an aggregate function if your statement includes an IN DOCUMENT clause
unless the statement also includes the USING ASSEMBLIES clause and an assembly exists for the
virtual document itself. (Refer to EMC Documentum Content Server Fundamentals Guide for information
about assemblies.) However, in this case, you may prefer to use the IN ASSEMBLY clause instead of
the IN DOCUMENT clause with USING ASSEMBLIES. Refer to The IN ASSEMBLY clause, page 139
for more information.
A date function manipulates a date in some manner. To use a date function in the selected values list,
use the AS clause to provide an alias for the returned value in the result object. For example:
SELECT DATETOSTRING("r_modify_date",'yy-mm-dd") AS ModifyDate
FROM "dm_document"
It is recommended that you use the AS clause to alias a return value for any function referenced in the
selected values list. The AS clause, page 132, has information about the AS clause.
FTDQL requirements
You can include scalar, aggregate, and date functions in the selected values list of an FTDQL query.
MFILE_URL function as a selected value
The MFILE_URL function returns URLs for the content files and renditions associated with the
objects returned by the SELECT statement. The function has three arguments that are used to
control which URLs are returned. For a description of the function and its arguments, refer to The
MFILE_URL function, page 31.
You can include the MFILE_URL function in the selected values list of an FTDQL query.

Select
Arithmetic expressions as selected values
You can include arithmetic expressions to perform calculations on the returned values. Arithmetic
expressions are expressions that use arithmetic operators to form the expression, such as:
property A + property B
column C * column D
(2 * column B) - property F
property X + 4
The following statement returns the total of all rental charges paid in June, including regular rents
and late charges:
SELECT "rent_received" + "late_charge_total" AS month_total
FROM "yearly_rent_records"
WHERE "mon" = 'june'
The expression cannot reference a non-qualifiable property.
FTDQL requirements
Arithmetic expressions are not allowed in the selected values list in an FTDQL query.
Keywords as selected values
Keywords are words that have a special meaning for the server. The majority are full-text keywords.
That is, the values they return are property values stored in the full-text index or values computed
during a full-text search. Three keywords return information about an object’s relationships to other
objects in a virtual document. The remaining keywords determine whether an object is a replica,
return the current user name, and return the URL to an object’s thumbnail file.
It is not necessary to include a SEARCH clause in the query to include a full-text key in the selected
values list. However, for some keywords such as TEXT, the returned values are not useful unless a
SEARCH clause is included.
Full-Text keywords
The following keywords return information about full-text indexes.
• CONTENTID
The CONTENTID keyword returns the object ID of the content object representing the content
file that matches the select criteria. (A content object links a content file to all documents that
contain the file.)
• ISCURRENT
The ISCURRENT keyword is a Boolean keyword that returns 1 (TRUE) if the object is the current
version and 0 (FALSE) if the object is not the current version.
• ISPUBLIC
The ISPUBLIC keyword is a Boolean keyword that returns 1 (TRUE) if the object is a public object
(its r_is_public property is TRUE) and 0 (FALSE) if the object is not a public object.
• OBJTYPE
The OBJTYPE keyword returns the r_object_type of the selected object.
• SCORE

Select
The SCORE keyword returns a document’s relevance ranking as determined by the index
server. If an ORDER BY clause is not included in the query, the returned document order is
by descending score.
• SYSOBJ_ID
The SYSOBJ_ID keyword returns the object ID of the objects returned with a fulltext search. It is
the object ID stored in the fulltext index for the object.
• SUMMARY
The SUMMARY keyword returns a summary of each document returned by the SELECT
statement. The summary is up to 4096 bytes from the document, chosen by the Full-Text
Engine. For example, the following statement returns a summary of the document whose name
is Company History:
SELECT SUMMARY FROM "dm_document"
WHERE "object_name"='Company History'
• TEXT
The TEXT keyword returns the variant text forms to which a specified term was match, to return a
particular document. The values are returned as a repeating property. For example, if the search
string specifies "talk”, the index server may return documents that contain the words "talking” or
"talked”. If so, then TEXT would return those forms of the word "talk”.
FTDQL requirements
All the keywords classified as fulltext keywords can be included in an FTDQL query.
Virtual document keywords
The following keywords return information about relationships in a virtual document.
• CONTAIN_ID
The CONTAIN_ID keyword returns the object IDs of the containment objects associated with the
components of a virtual document. The CONTAIN_ID keyword cannot return containment IDs
for components that are found by searching an assembly. Consequently, if the SELECT statement
includes an IN DOCUMENT clause with the USING ASSEMBLIES option, the containment IDs of
any components found using an assembly will be zeros (’0000000000000000’).
For example, suppose you have a virtual document, A, with one component, B. B is also a virtual
document with four components and an assembly. If you execute a SELECT statement to retrieve
the containment IDs of the components of A and include the DESCEND keyword, you receive the
containment IDs for A’s components at all levels. If the SELECT statement includes DESCEND
and USING ASSEMBLIES, you receive only the containment ID for B. (B’s containment object
links it to A.) Because the direct components of B are found using an assembly, their containment
IDs are not returned.
• DEPTH
The DEPTH keyword returns a component’s level within a virtual document. You can only use the
DEPTH keyword if the SELECT statement includes the IN DOCUMENT clause and the clause
includes the DESCEND keyword.
To illustrate its use, suppose the following query is executed against the virtual document shown
in Figure 1, page 129 :
SELECT "r_object_id", DEPTH FROM "dm_document"

Select
IN DOCUMENT ID('object_id_A') DESCEND
Figure 1. Sample virtual document
The statement returns
Object ID DEPTH
A 0
B 1
D 2
C 1
• PARENT
The PARENT keyword returns the object ID of the object that directly contains a direct or indirect
component of a virtual document. You can only include the PARENT keyword in the value list
if the SELECT statement contains an IN DOCUMENT clause.
For example, suppose the following query is executed against the virtual document shown in
Figure 1, page 129:
SELECT "r_object_id", PARENT FROM "dm_document"
IN DOCUMENT ID('object_id_A') DESCEND
The statement returns the object IDs and parents of each component of the virtual document:
Object ID PARENTS
A A
B A
D B
C A
Note that a virtual document is always considered to be its own parent.
FTDQL requirements
You may not use any of the keywords classified as virtual document keywords in an FTDQL query.
Miscellaneous keywords
The following keywords return miscellaneous information.
• ISREPLICA
The ISREPLICA keyword is a Boolean keyword that returns 1 (TRUE) if the returned object is
a replica and 0 (FALSE) if the object is not a replica. (Replica is the term used to describe any
object in the repository that has been placed in the repository as a copy of an object from another
repository. Only environments using object replication have replicas.)
You can only include the ISREPLICA keyword in the value list of the outermost SELECT
statement. You cannot use ISREPLICA in a subquery.
• USER

Select
The USER keyword returns the current user’s user name.
• THUMBNAIL_URL
The THUMBNAIL_URL keyword returns the URL to a thumbnail file. Use THUMBNAIL_URL
in a SELECT statement’s selected values to return the URL to the thumbnail associated with
the returned objects. For example:
SELECT object_name,THUMBNAIL_URL FROM dm_document
WHERE FOLDER('/Corporate Objectives')
The statement returns the documents in the folder Corporate Objectives. If the content of any of
returned document has an associated thumbnail, the URL to that thumbnail is returned also. If a
document has multiple content pages with thumbnails, the keyword returns only the thumbnail
associated with the first content page that has a thumbnail.
The URL contains the following information:
— The base URL defined for the thumbnail storage area
— A path relative to the storage area identified in the store property that points to the thumbnail
file
— A store property that identifies the storage area containing the thumbnail
If the storage area’s require_ticket property is TRUE, the URL also contains a ticket that
contains an encryption of the path plus a time stamp. The ticket is valid for five minutes, which
means that the URL is good only for five minutes.
For example:
http://myserver.documentum.com:8080/getThumbnail?
path=00232803/80/00/01/Ob.jpg&store=thumbnail_store_01&ticket=8002DWR670X
http://myserver.documentum.com:8080/getThumbnail? is the base URL.
path=00232803/80/00/01/Ob.jpg specifies the path to the file.
store=thumbnail_store_01 identifies the storage area.
ticket=8002DWR670X is the optional ticket.
In some applications, there may be multiple thumbnails generated for a single content object. For
example, thumbnails with different sizes can be associated with a single content object. In those cases,
you may want to use the MFILE_URL function to retrieve the URL. Typically, the page_modifier
property identifies the different thumbnails, so you would use that property to select the desired
thumbnail. See The MFILE_URL function, page 31, for more information.
However, another mechanism is available to affect the query issued to find thumbnail content when
the THUMBNAIL_URL DQL keyword is used. The following server.ini settings should be used to
control what content is returned by the keyword. They require knowledge of the underlying SQL
query issued by the content server to use correctly. The server.ini settings are:
• thumbnail_url_where_clause_extra—specifies an additional predicate that is added to the where
clause of the query
• thumbnail_url_order_by_clause—specifies the order by clause of the query
Beginning with release 6.7, the following query is used to find thumbnail content:
select b.parent_id,storage_id, format, data_ticket
from dmr_content_s a, dmr_content_r b
where a.r_object_id = b.r_object_id and

Select
b.parent_id in (selected_id_list) and page = 0 and
storage_id in (list_of_thumbnail_stores)
custom_predicate
order by order_by_clause
In the query:
• selected_id_list is the list of object IDs for which we are trying to find thumbnail content
• list_of_thumbnail_stores is the list of storage IDs where thumbnail content can be stored
• custom_predicate is:
— nothing—if the thumbnail_url_where_clause_extra server.ini parameter is not set
— and (string)—where string is the value of the parameter
• order_by_clause is:
— the string "full_format desc”, the default value, if the thumbnail_url_order_by_clause server.ini
parameter is not set
— the string specified in the parameter, if set
So, for example, to configure the query to return only objects that have the format jpeg_th and have
the page_modifier medium_jpeg_th, the following line should be added to server.ini:
thumbnail_url_where_clause_extra =
"exists (select x.r_object_id from dmr_content_r x where
x.r_object_id=a.r_object_id and x.page_modifier='medium_jpeg_th')"
The thumbnail_url_order_by_clause is blank in server.ini for this example.
When that line is added to the server.ini, subsequent queries to select the thumbnail content would
be modified to:
select b.parent_id,storage_id, format, data_ticket
from dmr_content_s a, dmr_content_r b
where a.r_object_id = b.r_object_id and
b.parent_id in (selected_id_list) and page = 0 and
storage_id in (list_of_thumbnail_stores)
and (exists (select x.r_object_id from dmr_content_r x where
x.r_object_id=a.r_object_id and x.page_modifier='medium_jpeg_th'))
order by full_format desc
To validate the settings used to modify the behavior of thumbnail_url, use the
VALIDATE_THUMBNAIL_URL_SETTINGS administrative method to check the settings. If the
settings are invalid, the method returns a detailed error message explaining the problem. This
information can be used to fix the settings in server.ini. In DQL, run the following query to execute
this method:
EXECUTE VALIDATE_THUMBNAIL_URL_SETTINGS
FTDQL requirements
The THUMBNAIL_URL keyword is the only keyword in the miscellaneous category that is acceptable
in an FTDQL query. You may not use the IS_REPLICA or USER keywords in an FTDQL query.
The asterisk (*) as a selected value
If the FROM clause references only registered tables or only object types, you can use an asterisk
(*) in the values list.
For registered tables, the asterisk returns all columns in the table.

Select
For lightweight object types:
• If the LITE keyword is included in the query, the asterisk returns all single-valued properties of
the lightweight object.
• If the LITE keyword is not included in the query, the asterisk returns all single-valued properties
of both the lightweight object and its shared parent.
For non-lightweight object types, what is returned by an asterisk depends on whether or not the
ROW_BASED hint is included in the query. If the hint is not included, an asterisk returns values only
for the first object type listed in the FROM clause. The values returned depend on the object type:
• For objects that are subtypes of persistent object type, the asterisk returns:
— All read/write single-valued properties
— The r_object_id property
• If the object type is SysObject or a SysObject subtype, in addition to the properties returned for
a persistent object, the asterisk also returns:
— r_object_type
— r_creation_date
— r_modify_date
— a_content_type
If the query includes the ROW_BASED hint, the asterisk returns the values of all properties of all
object types in the FROM clause. For example, the following query returns all property values from
both object types for returned objects:
SELECT * FROM dm_user,dm_group
If you want to limit the returned values to only the property values for a particular type in the FROM
clause, qualify the asterisk with the type. For example:
SELECT a.*, b.r_object_id FROM dm_document a, mydoc b
WHERE a.subject = b.title ENABLE(ROW_BASED)
In the above example, the asterisk returns the values only for properties from dm_document objects.
You can use an asterisk in this manner only when the ROW_BASED hint is included in the query.
An asterisk does not return values of properties defined for any aspects attached to an object.
FTDQL requirements
You may not use an asterisk as a selected value in an FTDQL query.
The AS clause
The AS clause lets you name the properties of the query result objects that represent the returned
results. For example, suppose you issue the following SELECT statement:
SELECT "a"+"b" AS total FROM "accts"
Total is assigned as the name of the query result object property that contains the returned a + b values.
When the statement returns the value of a property or column name, providing a name for the query
result property is optional. The system assigns the property or column name as the default name. If
you provide a name, the name must consist of ASCII characters.

Select
When the statement includes a function, arithmetic expression, or a keyword, you must specify a
name if you are running against MS SQL Server, and we recommend that you also do so if you
are running against Oracle or DB2.
MS SQL Server does not provide default names when values are functions, arithmetic expressions, or
keywords. For example, suppose you issue the following statement:
WHERE "subject" = 'cake'
Your return value would be:
______________
integer
Notice that there is no name above the integer return value.
For DB2, if you don’t provide a name, the returned column is named with its position in the selected
values list. For example, in the above SELECT example, the count would be returned in a column
named 1.
Because the server does not allow duplicate property names in an object, if you assign the same name
to more than one property, the system ignores all subsequent occurrences of the name and assigns
those properties different names.
Oracle provides a default name, but it is difficult to predict what forms the names for selected
functions, arithmetic expressions, or keywords will have. Consequently, explicitly assigning a name
is recommended.
FTDQL requirements
You may use the AS clause in an FTDQL query.
The FROM clause
The FROM clause defines which items are searched. Both the PUBLIC keyword and the source
list restrict the search.
PUBLIC keyword
The PUBLIC keyword restricts the search to objects for which the r_is_public property is set to
TRUE. If the query contains a SEARCH clause, the full-text search is restricted to those documents
for which ISPUBLIC is TRUE. When the server searches public objects only, it uses the setting of
r_is_public for security checks.
You may include PUBLIC in an FTDQL query.
Including PUBLIC increases the search performance.
Source list
A source list defines which object types and which RDBMS tables to search. The source list for a
standard query can include:
• One or more object types, to a maximum of 20 types. (In releases prior to 6.6, the maximum
was 10.)

Select
Any object type except those internal object types that represent aspect properties may be
specified.
• One or more RDBMS table names
• One or more inline views
The total number of object types and tables that you can include in the FROM clause is constrained by
the RDBMS. Each relational database has a limit on the number of tables that can joined in one query.
In the underlying RDBMS query, for each object type in the FROM clause, all the _s tables in the type’s
hierarchy plus any needed indexes are candidates for inclusion in the underlying query. If the query
references repeating properties, the _r tables are included also.
For registered tables, all object types included in the view’s definition are included in the query.
DQL passthrough hints included in the source list are applied only to the type or table after which the
hint appears. For portability, you can include hints for multiple databases. If you include hints for
multiple databases, only the hints that are applicable to the database receiving the request are used.
For more information about using passthrough hints, refer to Passthrough hints, page 363.
Source list JOIN operations
If more than one element is specified in the source list, separating each element by a comma causes
the types, tables, or inline views to be combined in an implied JOIN operation. However, a LEFT
OUTER JOIN clause can also be used. Use the following syntax:
FROM table | type | inline_view {,table | type | inline_view}
[LEFT [OUTER] JOIN type|table [correlation_variable] ON Boolean_condition
[LEFT [OUTER] JOIN type|table [correlation_variable] ON Boolean_condition]]
In the syntax above, you must supply at least one table, type, or inline view after FROM, and can
include additional comma separated tables, types, or inline views. An implied JOIN will join those
comma separated elements. Following the comma separated list of tables, types, and inline views,
you can optionally specify a LEFT OUTER JOIN clause. In that clause, you must specify one type or
table and a Boolean condition. You cannot use an inline view as the right-hand element in the LEFT
OUTER JOIN clause. One additional optional LEFT OUTER JOIN clause can follow the first one. The
Boolean_condition specifies what property to use to specify the JOIN.
Unlike the standard SQL behavior, a LEFT OUTER JOIN will behave like a simple JOIN, if the
Boolean_condition clause uses a repeating property from the right-hand element of the LEFT OUTER
JOIN.
Additionally:
• If any repeating attribute appears in the ON clause, a ROW_BASED hint must be used.
• If any repeating attributes are used in a select list or ON clause, only matching rows are returned.
The server adds the dm_repeating1_1.r_object_id=<type_name_1>.r_object_id predicate in the
WHERE clause which translates LEFT OUTER JOIN to an INNER JOIN.
• If there are aspect attributes in the select list, no type join can be performed, including LEFT
OUTER JOIN.
Including type names in the source list
Use the following syntax to specify a type name:
type_name [(ALL)|(DELETED)|(LITE)] [correlation_variable]
The type name argument identifies the object type to search. The server searches objects of the specified
type and all subtypes. For example, to search all SysObjects, specify dm_sysobject. The server
searches dm_sysobject and all of its subtypes, such as dm_document, dm_process, and so forth.

Select
The keyword ALL directs the server to search all versions of each object. The keyword DELETED
directs the server to search all versions of each object including any versions for which i_is_deleted
is set to TRUE. (This property is set to TRUE if you delete the object and it is the root version of a
version tree). The keyword LITE directs the server to search only the properties in the lightweight
type, otherwise the server also searches the properties in the shareable parent and the supertype(s). If
you specify LITE, then you must only specify properties in the lightweight type. You must enclose
ALL, DELETED, and LITE in parentheses.
If the FROM clause includes neither ALL nor DELETED, the server searches only the CURRENT
version.
The value of correlation_variable is a qualifier for the type name that is used to clarify property
references in the query.
Any user can specify any object type in a query with two exceptions. The first exception is the type
dmi_audittrail_attrs. To specify that type, you must have either Superuser or View Audit privileges.
The second exception are the object types representing aspect properties. These types cannot be
referenced in the FROM clause by any user.
The server searches only objects within that specified type that are owned by the user or objects to
which the user has access. Whether a user has access depends on the object-level permissions of the
objects. The user must have at least Browse permission on an object for the object to be included in
a query run by the user. If the query includes a SEARCH clause, the user must have at least Read
permission on the object. (For a full description of server security, refer to the EMC Documentum
Content Server Administration and Configuration Guide.)
Including multiple object types is subject to the following constraints:
• The selected values list cannot contain repeating properties unless the ROW_BASED hint is
included in the query.
• If the selected values list contains an asterisk (*), only the first type’s properties will be expanded
unless the ROW_BASED hint is included in the query.
The asterisk (*) as a selected value, page 131, contains details about including an asterisk in the
selected values list.
• The query cannot include a SEARCH clause.
• Unqualified property names in the query are disambiguated from left to right, as the object
types appear in the FROM clause.
• It the query includes an IN DOCUMENT clause, the object types must join in one-to-one
relationship.
• If the query includes an IN DOCUMENT or IN ASSEMBLY clause or a TYPE or FOLDER
predicate, the clause or predicate is applied to only one type in the join, and that type will be the
first object type identified in the FROM clause.
Including table names in the source list
You can include one or more RDBMS table names in the FROM clause of a standard query. The
syntax is:

Select
[owner_name.]table_name [correlation_variable]
The owner_name argument identifies the owner of the table. You must include the owner name if
you are not the owner of the table. If the owner is the DBA of the database, you can use the alias
dm_dbo as the owner name. (Using dm_dbo when appropriate makes your statement portable across
databases.) If the owner name is unspecified, the current user is assumed.
You must be a superuser to include an unregistered RDBMS table in the list. Any user with
appropriate permissions can include a registered table.
table_name is the name of the table that you want to search.
correlation_variable is a qualifier for the table name used to clarify column references in the query.
Including inline views in the source list
You can include one or more inline view in the FROM clause of a standard query, just like a type or
registered table. The syntax is:
( SELECT ... ) [correlation_variable]
The parenthesis are required, as an inline view is the result from a subquery. The correlation_variable
is optional, and is used to refer to the view in other sections of a query statement.
Clarifying type and table names
Type and table names can be the same. In such instances, the server must distinguish which name
identifies a type and which identifies a table. The server uses the following rules for distinguishing
between type and table names:
• If (ALL) or (DELETED) is present, the name is a type name.
• If ownername is present, the name is a table name.
• All other cases are ambiguous references. In such cases, the server checks to see whether the name
is a type name. If it is, the name is assumed to be a type.
The values in the source_list argument are processed from left to right. Consequently, if you include
tables and types with the same name in the FROM clause, be sure to qualify the table name with its
owner’s name. This will ensure that the server does not mistakenly assume that the table name is
the type name.
To illustrate, suppose you have an object type named legal_docs and a registered table of the same
name. The following statement shows one correct way to include them both in one SELECT statement:
SELECT a."r_object_id", b."client"
FROM "legal_docs" a, jolenef."legal_docs" b
The first legal_docs (with correlation name a) is a type name. The server knows the second is a
registered table because an owner name is specified. Note that correlation variables should be used in
this case to clarify column references.
The following statement is valid because billing_docs is not found to be a type and so is assumed to
be a registered table:
SELECT a."r_object_id", b."client"
FROM "dm_document" a, "billing_docs" b
In the following statement, the server assumes that the query references the legal_docs type, not the
registered table, because no owner name is specified and there is a type called legal_docs:
SELECT a."object_id", b."client"
FROM "legal_docs" b, "dm_document" a

Select
FTDQL constraints on the source list
Observe the following rules if you want to execute the query as an FTDQL query:
• You can only reference one object type and that type must be dm_sysobject or a dm_sysobject
subtype.
• You cannot reference a registered table in the source list.
The IN DOCUMENT clause
Note: You may not include an IN DOCUMENT clause in an FTDQL SELECT statement.
The IN DOCUMENT clause lets you use the SELECT statement to perform the following operations:
• Identify only the directly contained components of a virtual document
• Identify the indirectly contained components that reside in the current repository with the virtual
document
If a component is a folder that doesn’t reside in the current repository, the query returns the
object ID of the folder’s mirror object in the repository, but not any folders or documents that
the remote folder contains.
• Assemble a virtual document
• Identify the virtual documents that directly contain a particular object
If the value list for the SELECT statement contains a repeating property and the statement contains an
IN DOCUMENT clause, the server automatically removes any duplicates of the repeating property.
You cannot include an ORDER BY clause when you include an IN DOCUMENT clause.
The syntax of the IN DOCUMENT clause is:
IN DOCUMENT object_id [VERSION version_label]
[DESCEND][USING ASSEMBLIES]
[WITH binding condition]
[NODESORT BY property {,property} [ASC|DESC]]]
Specifying the virtual document
To identify which virtual document to search, use either the document’s object ID or the combination
of an object ID and a version label. When you use only the object ID, the server searches the version
identified by that object ID. When you use a version label in addition to an object ID, the server
searches that version of the object. Note that this version may not be the same as the version
identified by the object ID.
If the SELECT statement is not a subquery, specify the object_id using the special ID function and the
actual object ID of the virtual document. For example:
IN DOCUMENT ID('object_id')...
If the SELECT statement is a subquery, you can also specify the object_id using a correlated r_object_id
property reference. For example:
SELECT "r_object_id" FROM "dm_document" x
WHERE EXISTS
(SELECT * FROM "dm_document"
IN DOCUMENT x."r_object_id"
WHERE "object_name"='Chapt1')

Select
The VERSION option lets you specify a version label. This label can be assigned to any version in the
version tree that contains the specified virtual document. If you do specify a label, the server searches
the specified version rather than the version identified by the object ID.
The DESCEND option
The keyword DESCEND directs the server to search the virtual document’s hierarchy, including all
components directly or indirectly contained in the virtual document. Otherwise, the server searches
only those components that are directly contained. You cannot use the DESCEND keyword in the
IN DOCUMENT clause if the SELECT statement is a subquery.
The USING ASSEMBLIES option
An assembly defines the components of a particular version of a virtual document at a particular
time (the time the assembly was created). Creating an assembly ensures that a particular version of
a virtual document always contains the same material. (Refer to EMC Documentum Content Server
Fundamentals Guide for information about creating assemblies.)
If a component selected for inclusion is a virtual document and the USING ASSEMBLIES clause is
included, the server determines whether an assembly is defined for that component. If so, the server
uses the information in the assembly as the definition of the component’s composition.
Refer to the EMC Documentum Content Server Fundamentals Guide for a full discussion of assemblies.
The WITH option
The WITH option lets you identify which version of a virtual document’s component you want to
include in the document. The syntax of the WITH option is:
WITH binding condition
where binding condition is one or more expressions that resolve to a single Boolean TRUE or FALSE.
The expressions can include comparison operators, literals, qualifiable property names, column
names, scalar and date functions, arithmetic expressions, and any predicates except SysObject
predicates. Multiple expressions in the condition are joined using logical operators.
The condition defined by the WITH option is applied to any component of the virtual document that
was added without a version being specified at the time of its addition. For example, suppose you
add component C to virtual document A without specifying a particular version of component C.
Later, when you assemble virtual document A, you can use the WITH clause condition to determine
which version of component C to include in the document. You may want to include the version
that has the word accounting as a keyword or the version that carries the symbolic label CURRENT.
Each time you assemble the document, you can select a different version of component C. Choosing
a version for inclusion at the time of actual assembly is called late binding. (For more information
about late binding, refer to EMC Documentum Content Server Fundamentals Guide.)
More than one version of a component may qualify for inclusion in the virtual document. To resolve
this, you can include the NODESORT BY option in the IN DOCUMENT clause or you can allow the
server to make a default choice. If you allow the server to make the choice, the server includes the
version with the smallest object ID; that is, the earliest version of the component.
The NODESORT BY option
The NODESORT BY option works in conjunction with the WITH option to identify one version of a
particular component for inclusion in a virtual document. (In some instances, more than one version

Select
of a component may fulfill the condition imposed by the WITH option.) The NODESORT BY option
sorts the versions by the value of one or more properties. The server includes the first version.
The syntax of the NODESORT BY option is:
NODESORT BY property {,property} [ASC|DESC]
where property is any qualifiable property belonging to the component. You can sort in ascending
(ASC) or descending (DESC) order. If you do not specify a sort order, the default is ascending (ASC).
If you use the NODESORT BY option and two or more versions still qualify for inclusion, the server
picks the version having the smallest object ID. This situation could occur if two or more versions
have exactly the same values in the properties specified in the NODESORT BY option.
The IN ASSEMBLY clause
Note: You may not include the IN ASSEMBLY clause in an FTDQL SELECT statement.
The IN ASSEMBLY clause allows you to identify an assembly associated with a particular document.
An assembly is a snapshot of a virtual document at a particular point in time and under conditions
defined when the assembly was created. Using the IN ASSEMBLY clause means that the SELECT
statement queries the virtual document represented by the assembly; however, the server uses only
the components found in the assembly, rather than recursively searching the virtual document’s
hierarchy.
You can also use this clause to identify a single component of a virtual document and any components
contained by that component. The component must be contained in the assembly.
The IN ASSEMBLY clause is more flexible than the IN DOCUMENT clause. For example, you
can include aggregate functions such as COUNT or MAX in the value list when you use the IN
ASSEMBLY clause. You cannot include aggregate functions when a SELECT statement includes
an IN DOCUMENT clause.
The syntax of the IN ASSEMBLY clause is:
IN ASSEMBLY [FOR] document_id [VERSION version_label]
[NODE component_id] [DESCEND]
where document_id identifies the document with which the assembly is associated. Use the
document’s object ID for this argument, specified as an ID literal:
ID('object_id')
Note: You can create an assembly for a virtual document and assign that assembly to another
document. In such cases, the value of the document_id argument identifies the document to which the
assembly is assigned, not the virtual document that the assembly represents.
If the specified document does not have an associated assembly, the statement fails with an error.
The VERSION option lets you specify a particular version of a document. If you include a version
label, the server finds the version of the document carrying that label and uses the assembly
associated with that version. You can specify either a symbolic label or an implicit label.
The NODE option allows you to identify a particular component of a virtual document. Use the
component’s object ID, specified as an ID literal:
ID('component_id')
The component must be contained in the assembly.

Select
You cannot include the NODE option if the SELECT statement contains an aggregate function in the
value list or if the SELECT statement is a subselect or subquery.
The DESCEND keyword directs the server to return not only directly contained components but also
any indirectly contained components that meet the criteria in the statement. If you do not include this
keyword, the server returns only directly contained components of the document or component.
The SEARCH clause
The SEARCH clause identifies a subset of the full-text indexed documents. When you include a
SEARCH clause, Content Server submits the query to the index server, which returns all objects
that meet the SEARCH clause condition. The index server searches both content files and indexed
metadata (property values) for matches with the specified condition.
The syntax of the SEARCH clause for a standard query is:
SEARCH [FIRST|LAST]DOCUMENT CONTAINS [NOT]search_string
[IN FTINDEX index_name {,index_name}]
The search_string syntax is described in SEARCH DOCUMENT CONTAINS, page 140.
FIRST and LAST keywords
The FIRST and LAST keywords determine when the search is conducted. These keywords are only
valid in standard SELECT queries. You may not include them if the query is an FTDQL query.
If you include FIRST, the server runs the full-text search query first and then applies the SELECT
statement. For example, suppose legal_documents is a subtype of dm_document. The following
statement first searches the full-text index for all documents that have the word fiduciary in their
content and then finds the members of that group that are also legal_documents:
SELECT "object_name" FROM "legal_documents"
SEARCH FIRST DOCUMENT CONTAINS 'fiduciary'
If you include LAST, the server first searches the repository for all objects that meet the SELECT
criteria and then applies the criteria in the SEARCH clause. For example, the following statement first
finds all legal_documents authored by G. Oliphant and then finds the members of that group that
contain the word fiduciary:
SELECT "object_name" FROM "legal_documents"
SEARCH LAST DOCUMENT CONTAINS 'fiduciary'
WHERE ANY "authors"='G.Oliphant'
The default behavior is to search the full-text index first. If you include a full-text keyword in the
selected values list, the full-text search is conducted first even if LAST is specified in the SEARCH
clause.
SEARCH DOCUMENT CONTAINS
The SEARCH DOCUMENT CONTAINS syntax is the way to specify a fulltext search condition in a
SEARCH clause. Including a list of words or phrases in a SEARCH DOCUMENT CONTAINS clause
returns those documents whose content contains one, several or all of the specified words or phrases
depending on the syntax you specify. You can also filter out the documents that contain the specified
words or phrases. The documents are returned in order of importance.
The syntax is:
SEARCH [FIRST|LAST]DOCUMENT CONTAINS [NOT] search_string

Select
where search_string has the following syntax:
[NOT]'word' {AND|OR [NOT] 'word'}
word may be a word, a phrase. It may also be a list of words or phrases, or both, separated by spaces
and quoted appropriately. For more information about this syntax, refer to Specifying words and
phrases in SEARCH DOCUMENT CONTAINS, page 141.
The complex syntax using logical operators is not recommended as future support may be dropped
for this syntax and replaced with a different syntax. However, this syntax is currently supported.
Specifying words and phrases in SEARCH DOCUMENT CONTAINS
A word can be any character string that does not include spaces or punctuation. For example, the
following statement finds all indexed documents that contain the word yeast:
SELECT * FROM "dm_documents"
SEARCH DOCUMENT CONTAINS 'yeast'
The SEARCH DOCUMENT CONTAINS clause uses AND as the default logic operator for search
items. For example, the following query returns documents that have both yeast and cake in their
content or metadata:
SEARCH DOCUMENT CONTAINS 'yeast cake'
To search for any word in a list of search items, you have to explicitly use the OR operator. For
example, the following query returns documents that have yeast or cake or both words in their
content or metadata:
SEARCH DOCUMENT CONTAINS 'yeast' OR 'cake'
If you want to include a single quote in the search string, you must escape the quote with another
single quote. For example:
SEARCH DOCUMENT CONTAINS 'O''Malley'
To search on a phrase, enclose the phrase in double quotes inside the quoted search string. For
example, the following query returns all documents that include the phrase "blend butter and sugar”:
SEARCH DOCUMENT CONTAINS '"blend butter and sugar"'
Note: You can use single quotes to enclose a phrase, but if you do so, you must escape the quotes
with single quotes.
You can combine words and phrases in the same search string. For example, the following query
returns documents that contain the word "yeast” or the phrase "blend butter and sugar” or both:
SEARCH DOCUMENT CONTAINS 'yeast "blend butter and sugar"'
You can filter out the documents that contain the specified words or phrases by using the NOT
operator. The following example searches for documents that contain "yeast” but not "butter”:
SEARCH DOCUMENT CONTAINS 'yeast' AND NOT 'butter'
Case sensitivity
All searches conducted by the index server are case insensitive.
For example, suppose you issue the following query:

Select
SELECT owner_name,r_creation_date FROM dm_document
SEARCH DOCUMENT CONTAINS 'budget'
The query returns all documents that contain, either in content or metadata, the word budget in
lowercase or uppercase or in any combination (Budget, buDget, budgeT, and so forth).
Accent and diacritical marks
Some languages use accents and diacritical marks on some characters or syllables in words. Searches
are insensitive to accent and diacritical marks. When you search on a word or phrase, the search
returns all objects that contain the word or phrase, even if some matches also contain an accent or
diacritical mark. Similarly, when you search on a word or phrase that contains such marks, the search
ignores the marks and returns all objects that contain the word or phrase, spelled with or without the
accent or diacritical mark.
For example, suppose you issue the following query:
SEARCH DOCUMENT CONTAINS 'cote'
The query returns all documents that contain, in metadata or content, the word cote, including those
with instances of the word with accents or diacritical marks (côte, côté, and so forth).
Now, suppose you issue the following query that specifies a search term that includes an accent:
SEARCH DOCUMENT CONTAINS 'coté'
That query also returns all documents that contain, in metadata or content, the word cote, including
those with instances of the word with accents or diacritical marks (côte, côté, and so forth).
Asterisk as wildcard character
You can use the asterisk (*) as a wildcard character in a search string. The asterisk matches any string
of letters. You can use it in any position in the string. For example, foo* matches any word that begins
with "foo”. The string *foo* matches any word that contains "foo” in its sequence of letters. The string
faa*foo matches any word that begins with faa and ends with foo.
You may also use the asterisk as a wildcard in a phrase search. For example, suppose you provide
the following search string:
'"n* is th* time*"'
That string matches phrases such as "now is the time”, "nothing is thoroughly timed”, and "November
is thinking time”.
The WHERE clause
Use the WHERE clause to restrict which objects are returned. For example, the following statement
selects the title of every document but only returns the titles of documents owned by the current user:
SELECT "title" FROM "dm_document"
WHERE "user_name" = USER
The syntax of the WHERE clause is:
WHERE qualification
The qualification argument consists of one or more expressions that resolve to a single Boolean TRUE
or FALSE. The expressions can include comparison operators, literals, qualifiable property names,

Select
column names, scalar and date functions, arithmetic expressions, predicates, and full-text keywords.
Multiple expressions are joined together using the logical operators. (Chapter 1, DQL Language
Elements, contains descriptions of all accepted comparison and arithmetic operators, predicates,
functions, and logical operators. Full-text keywords are described in Full-Text keywords, page 127.)
A qualification can also include a subquery:
WHERE ANY "authors" IN
WHERE "r_is_group" = TRUE)
The qualification can be simple or as complex as necessary. For example, the following WHERE
clause is simple:
SELECT * FROM "dm_workflow"
WHERE "supervisor_name" = 'horace'
This next example contains a more complex qualification:
SELECT "r_object_id", "title," FROM "dm_document"
WHERE "r_creation_date" >= DATE('01/01/99','mm/dd/yy')
AND "r_modify_date" <= NOW
The next example references an aspect property. Note that the property name is fully qualified with
the name of the aspect. All references to aspect properties must be qualified with the aspect name.
SELECT "r_object_id", "title" FROM "dm_document"
WHERE ANY AUTHOR IN ("JohnDoe") and grant_validation.grant_amount>100000
General constraint on the qualification
A qualification cannot reference any of the following properties of audit trail objects unless you have
Superuser or View_audit privileges:
• acl_name • object_name
• acl_domain • object_type
• property_list • owner_name
• property_list_id • session_id
• chronicle_id • version_label
Using repeating properties in qualifications
Referencing repeating properties in a qualification is supported. However, there are some rules and
guidelines. This section discusses those rules and guidelines.
ANY keyword use
If an expression references a repeating property, you must include the ANY keyword in the
expression unless the ROW_BASED hint is included in the query. For example, the following query
includes the ANY keyword because it does not include ROW_BASED:
SELECT "r_object_id","title","subject" FROM "dm_document"
WHERE ANY "authors" IN ('gillian')
If ROW_BASED is included, the query may be written without the ANY predicate:
SELECT "r_object_id","title","subject" FROM "dm_document"

Select
WHERE "authors" IN ('gillian') ENABLE(ROW_BASED)
If you are referencing a repeating property and including a subquery, you can use the IN or EXISTS
keyword after the ANY keyword to control the generated SQL query. The syntax is:
WHERE ANY [IN|EXISTS] attr_name IN subquery
Including IN or EXISTS may change the generated SQL query and consequently, enhance
performance. IN and EXISTS, page 353 contains an example that shows the differences between
the options in the generated query.
Referencing repeating properties in compound expressions
A compound expression is multiple expressions linked by logical operators. For example, the
following is a compound expression:
object_name = 'book_proposal' AND owner_name = 'john doe'
If two or more expressions in a compound expression reference repeating properties, the properties
must be from the same object type unless the ROW_BASED hint is included in the query. For
example, the following query contains a compound expression that has two expressions referencing
repeating properties from dm_document:
SELECT r_object_id, object_name FROM dm_document
WHERE ANY authors IN ('joe','john') AND ANY keywords LIKE 'eng%'
However, suppose you want to select from two object types and use a compound expression that
referenced repeating properties from both types. To do that, you must include the ROW_BASED hint:
SELECT a.r_object_id, b.i_acceptance_date
FROM dm_document a, dmi_package b
WHERE a.authors IN ('john doe') AND b.r_package_label = 'APPROVED' ENABLE(ROW_BASED)
Without the ROW_BASED hint, the query would fail with an error stating that the two repeating
properties it references, authors and r_package_label, are not from the same type.
Additionally, when ROW_BASED is included, it is not necessary to include the ANY predicate.
Including ANY does not generate an error in such cases, but it is not necessary.
Referencing repeating properties in comparison expressions
A comparison expression is an expression that uses a comparison operator, such as = or >, to compare
the values of two properties. The following rules apply when comparing a repeating property to
another property in a qualification:
• You can compare a repeating property to a single-valued property from the same or a different
object type.
For example, the following queries are legal:
SELECT r_object_id, title FROM dm_document
WHERE ANY authors=object_owner
SELECT a.r_object_id, b.user_name
FROM dm_document a, dm_user b

Select
WHERE ANY a.authors=b.user_name
• You cannot compare a repeating property to another repeating property unless you include the
ROW_BASED hint.
For example, the following query is illegal:
SELECT a.r_object_id,b.r_object_id,a.title
FROM dm_document a,dm_sysobject b
WHERE ANY a.authors = b.keywords
To make that example a legal statement, you must include the ROW_BASED hint:
SELECT a.r_object_id,b.r_object_id,a.title
FROM dm_document a,dm_sysobject b
WHERE ANY a.authors = b.keywords
ENABLE(ROW_BASED)
FTDQL requirements for where clauses
A WHERE clause in an FTDQL query is subject to the following rules:
• Any repeating properties referenced in the clause must be of type string or ID.
• Only the DQL UPPER and LOWER functions are allowed. The functions SUBSTR and
MFILE_URL and all aggregate functions are not acceptable.
• The following predicates are not allowed:
— BETWEEN
— NOT LIKE
— NOT FOLDER
— [NOT] CABINET
— ONLY
— TYPE
• The IN and EXISTS keywords are not allowed.
• Any valid form of the FOLDER predicate (except NOT FOLDER) may be used. The DESCEND
option is also allowed.
• The following rules apply to the LIKE predicate:
— The LIKE predicate may be used with pattern matching characters.
— The LIKE predicate can be used with an ESCAPE clause, but it is ignored.
• The keywords TODAY, YESTERDAY, TOMORROW may not be used in the DATE function.
• The following rules apply to all expressions in the WHERE clause:
— Expressions may not contain the ISREPLICA or USER keywords.
— Expressions may use any comparison operator.
— Expressions may use an arithmetic operator, but they may not be used to form a compound
expression.
— Expressions that compare one property to another are not allowed. For example, subject=title
is invalid.
— Expressions in the following format are not allowed:

Select
property_name operator('literal' operator 'literal)
— Expressions that force index correspondence between repeating properties are not allowed.
Such expressions AND together expressions that reference repeating properties in the format:
predicate(repeating_attr_expr AND repeating_attr_expr)
(For more information about forcing index correspondence, refer to Forcing index
correspondence in query results, page 342.)
• The following additional rules apply to expressions in the format first_expression operator
second_expression
— The first_expression is limited to one of the following:
property name
upper(property name)
lower(property name)
— The second_expression is limited to one of the following:
literal value
upper(literal value)
lower(literal value)
the DATE() function
— The operator may be any valid operator.
The GROUP BY clause
Note: You may not include a GROUP BY clause in an FTDQL SELECT statement.
The GROUP BY clause groups results. Use it when you include a function in the value list for the
SELECT statement. The value_list argument for the GROUP BY clause must contain all of the values
in the value list for the SELECT statement that are not aggregate function values.
For example, the following statement returns the names of all documents, their owners, and a count
of the documents that each owner owns. The results are grouped by owner name:
SELECT "owner_name", count(*)
FROM "dm_document"
GROUP BY "owner_name"
This next example selects the names of all documents, their owners, and their subjects, and groups
them first by owner and then, within each owner group, by subject:
SELECT "owner_name", "subject", count (*)
FROM "dm_document"
GROUP BY "owner_name", "subject"
The HAVING clause
Note: You may not include a HAVING clause in an FTDQL SELECT statement.
The HAVING clause restricts which groups are returned. It is most commonly used in conjunction
with the GROUP BY clause.
For example, the following statement returns owner names and the number of documents each owner
owns, grouped by owner names, for all owners having more than 100 documents:
SELECT "owner_name", count(*)
FROM "dm_document"

Select
HAVING count(*) > 100
This statement first finds and counts all documents for each owner name. It returns only those groups
(owner’s name and count) for which the count is greater than 100.
You can also use the HAVING clause in a SELECT statement that does not include the GROUP BY
clause when a value in the value list is an aggregate function. For example, the following statement
returns a count of the workflows supervised by haroldk if that count is greater than or equal to 25:
SELECT COUNT(*) FROM "dm_workflow"
WHERE "supervisor_name" = 'haroldk'
HAVING COUNT (*) >=25
If the number of workflows supervised by haroldk is less than 25, the statement returns nothing.
Note that if you do not include a GROUP BY clause when you use a HAVING clause, the only value
permitted in the value list for the SELECT statement is one aggregate function.
The UNION clause
Note: You may not include a UNION BY clause in an FTDQL SELECT statement.
The UNION clause lets you obtain results from more than one SELECT statement. The syntax is:
UNION dql_subselect
The dql_subselect argument must return the same number of properties or columns as the first
SELECT statement, and each property or column must have the same datatype as its corresponding
property or column in that SELECT statement. For example, if the first SELECT statement returns
three properties, then the subselect argument in the UNION clause must also return three properties.
The datatypes of the first properties returned by each must be the same, as must the datatypes
of the second and third properties.
Neither the first SELECT statement nor any subsequent UNION SELECT statements can contain an
IN DOCUMENT clause. The IN ASSEMBLY clause can be included only if the clause is in the first
SELECT statement and all unioned subselect statements.
For all databases except DB2, when you union two or more SELECT statements, the property names
that are returned are derived from the first SELECT statement. For example, suppose you issue the
following statement:
SELECT "name", "address" FROM "current_emp"
UNION SELECT "ret_name", "ret_address" FROM "retired_emp"
The query result objects for this statement have two properties, name and address, taken from the
value list of the first SELECT in the statement.
For DB2 , if corresponding selected values in the unioned SELECT statements don’t have the same
name or alias, the returned property name is an number representing the property’s position in the
selected values list. For example, suppose you issue the following query:
SELECT "name", "address" FROM "current_emp"
UNION SELECT "ret_name", "ret_address" FROM "retired_emp"
The query result objects for this statement have two properties. The first property is named 1,
representing selected values from name and ret_name. The second property is named 2, representing
selected values from address and ret_address.

Select
The server does not return duplicate rows when you union two or more SELECT statements. This is
true even if the ALL keyword is included in the first SELECT statement.
The ORDER BY clause
Use the ORDER BY clause to sort the results of the SELECT statements. The syntax of this clause is:
ORDER BY value [ASC|DESC] {,value [ASC|DESC]}
The value argument must be a property or column name that appears in the value list. You can sort
in ascending (ASC) or descending (DESC) order. If you do not specify a sort order, the default is
ascending.
The primary sort is on the first value specified, the secondary sort is on the second value specified,
and so forth. To illustrate, the following statement returns the owner name, subject, and title of all
documents in the system in ascending order by owner name and the subject:
SELECT "owner_name", "subject", "title"
FROM "dm_document"
ORDER BY "owner_name", "subject"
You can specify a value either explicitly, by name, or by position. For example, the following
statement returns the same results as the previous example, but notice that the ORDER BY clause
specifies values by their position in the SELECT value list:
SELECT "owner_name", "subject", "title"
FROM "dm_document"
ORDER BY 1, 2
The only exception occurs when you union two or more SELECT statements. In such cases, you can
only specify a value by its position in the value list.
FTDQL requirements for ORDER BY
If you include an ORDER BY clause in an FTDQL query, the sort order must be specified by SCORE:
ORDER BY SCORE
You cannot reference SCORE by its position in the selected values list. You must reference it by name.
You may sort in either ascending or descending order.
Including DQL hints
You can include processing hints for DQL and the RDBMS servers in SELECT statements. The hints
for DQL are called standard hints and can be added at the end of the SELECT statement, using the
ENABLE keyword. Hints to the RDBMS server are called passthrough hints and can be added in the
FROM clause, after a table or type name, and at the end of the statement, using the keyword ENABLE.
The ROW_BASED hint may not be included in FTDQL queries or in queries that reference a
lightweight object type in the FROM clause. All other hints are acceptable in FTDQL queries or
when querying lightweight object types.
Table 29, page 149, lists the standard hints and provides brief guidelines for their use. For detailed
information about their implementation, refer to Appendix A, Using DQL Hints .

Select
Table 29. DQL standard hints
Standard hint Description
CONVERT_FOLDER_LIST_TO_OR Changes DQL FOLDER clause to the behavior of release
6.0 and earlier.
FETCH_ALL_RESULTS N Fetches all results from the cursor immediately and then
closes the cursor. Set N to a positive integer number or
0 to return all rows.
Guideline
• If you include FETCH_ALL_RESULTS and other hints
that control the number of rows returned, the last one
listed in the hints is used.
FORCE_ORDER Controls the join order for the tables and types listed
in the source list. If this hint is included, the tables are
joined in the order they are listed.
Guideline
• This hint is ignored on DB2.
[NO]FTDQL FTDQL directs Content Server to execute the query as
an FTDQL query. If the syntax of the query violates the
rules for FTDQL syntax, an error is returned.
NOFTDQL directs Content Server to execute the query
as a standard query.
FT_CONTAIN_FRAGMENT Expands the range of matches for full-text searches that
include a search clause of the form LIKE ’%string%’.
GROUP_LIST_LIMIT N GROUP_LIST_LIMIT defines the maximum number of
groups that may be referenced in the generated SQL
statement to check permissions for the user executing
the query.
The default limit is 250. Using this hint overrides
the default value and the DM_GROUP_LIST_LIMIT
environment variable, if that is set, as well as flushes and
rebuilds the group list cache in the new size. For users
whose group membership number is below the specified
limit, generated SQL statements display group names
inline without subqueries.
HIDE_SHARED_PARENT HIDE_SHARED_PARENT directs Content Server to
return only the rows in the query results that are not
shared parents.

Select
OPTIMIZATION_LEVEL level_1
level_2
Allows you to change the query optimization level. Set
level_1 to the optimization level you want for the current
query and level_2 to the optimization level desired for
the connection after the query completes.
Guidelines
• This hint in only useful against DB2 databases. It is
ignored for other databases. For information about
valid values for level_1 and level_2, refer to your DB2
documentation.
IN and EXISTS Mutually exclusive hints used in a standard WHERE
clause, in a repeating property predicate that contains
a subquery. If you do not include either IN or EXISTS
explicitly, when Content Server translates the query, it
includes one or the other automatically.
OPTIMIZE_ON_BASE_TABLE Optimizes queries that reference properties contained
within one type in the type hierarchy.
OPTIMIZE_TOP N Returns N rows very quickly, then continues with the
remainder of the rows. Set N to positive integer.
Guidelines
• This hint is most useful in conjunction with
RETURN_TOP.
• On an Oracle database, N is ignored.
RETURN_RANGE starting_row
ending_row
[optimize_top_row]
’sorting_clause’
Specifies which rows are returned by a query sorted by
the returned values of specified properties. This hint is
provided as a general way to paginate the results of a
query.
Guidelines
• optimize_top_rows provides the top rows for
optimization.
• sorting_clause has the format
— ’attribute_name [ASC|DESC]
[,attribute_name [ASC|DESC]...]’
It uses ascending by default.

Select
RETURN_TOP N Limits the number of results returned by a query. Set N
to a positive integer number.
Guidelines
• If you include RETURN_TOP and other hints that
control the number of rows returned, the last one
listed in the hints is used.
• This hint is useful because it can limit the effect of a
bad (unbounded) query on the database.
• Sorting the results or including the DISTINCT
keyword reduces or eliminates the benefits of using
RETURN_TOP.
• On a SQL Server database, this hint reduces the
number of rows touched by the query.
• On a DB2 database, using OPTIMIZE_TOP with
RETURN_TOP is recommended.
ROW_BASED Directs Content Server to return query results that
contain repeating property values in a row-based
format—one row for each returned repeating property
value. In addition, the hint also affects the rules
governing:
• Selecting repeating property values
• Using an asterisk as a selected value
• Referencing repeating properties in WHERE clause
qualifications
More information about these effects is found in
Repeating properties, page 123, The asterisk (*) as a
selected value, page 131, and Using repeating properties
in qualifications, page 143.
SQL_DEF_RESULT_SET N Defines a maximum number (N) of rows to return. Set
N to a positive integer number. If set to 0, all rows are
returned.
Guidelines
• On a SQL Server database, this hint forces the use of
default result sets instead of a cursor to return the
rows. On all other databases, the hint only sets the
number of rows to return.
• It is recommended that you set N to a maximum limit,
rather than specifying it as 0.

Select
• If you include SQL_DEF_RESULT_SET and other
hints that control the number of rows returned, the
last one listed in the hints is used.
TRY_FTDQL_FIRST TRY_FTDQL_FIRST directs Content Server to first
execute the query as an FTDQL query and, if a timeout
or resource exceeded error occurs, to retry the query as a
standard query.
If this hint is included, the FTDQL and NOFTDQL hints
are ignored if they are also included.
UNCOMMITTED_READ Ensures that a read only query returns quickly even if
another session is holding locks on the tables queried by
the read only query.
This hint is useful only on SQL Server and DB2 databases.
DM_LEFT_OUTER_JOIN_FOR_ACL If the user is not a super user, this hint ensures that
Content Server generates an SQL query that contains a
LEFT OUTER JOIN on the dm_acl_s table. This operation
improves the performance of DQL queries.
Examples
The following example returns the names and titles of all documents owned by the current user:
SELECT "object_name", "title" FROM "dm_document"
WHERE "owner_name" = USER
The next example selects all documents and their authors with the subject employee_benefits:
SELECT "r_object_id", "authors" FROM dm_document
WHERE "subject" = 'employee_benefits'
ORDER BY 1
The following example returns the names of all objects in the New Books cabinet:
SELECT "object_name" FROM "dm_sysobject"
WHERE CABINET ('/New Books')
ORDER BY "object_name"
The following FTDQL example returns those documents that contain the phrase "for publication:” for
which the user has Write permission:
SELECT FOR WRITE object_name,title,owner_name FROM dm_document
SEARCH DOCUMENT CONTAINS 'for publication'
There are additional examples demonstrating the use of specific clauses in the descriptions of the
clauses.

Unregister
Unregister
Purpose
Removes a registered table from the repository.
Syntax
UNREGISTER [TABLE] [owner_name.]table_name
Arguments
Table 30. UNREGISTER argument descriptions
owner_name Identifies the table’s owner. The default value is the current
user.
Use the owner’s RDBMS user name. If the owner is a
repository user, this value is found in the user’s user_db_name
property.
If the RDBMS is Oracle and the owner is the DBA, you can
use the alias dm_dbo.
If the RDBMS is MS SQL Server and if the owner is the DBA,
you can use the alias dbo.
table_name Identifies a registered table to remove from the repository.
Use the table name as it appears in the RDBMS.
Permissions
You must be the owner of the registered table or have Superuser privileges to unregister a table.
Description
Unregistering a table removes the object of type dm_registered that represents the table in the
repository. It does not remove the table from the underlying RDBMS.
If you attempt to unregister a table that is not registered, you receive an error message.

Unregister
Related statements
Delete, page 91
Insert, page 107
Register, page 110
Examples
The following example unregisters the table called departments:
UNREGISTER TABLE "departments"

Update
Update
Purpose
Updates the rows of a registered table.
Syntax
UPDATE table_name SET column_assignments
Arguments
Table 31. UPDATE argument descriptions
table_name Identifies the registered table to update. (Refer to Register,
page 110 for information about registering tables in the
repository.)
SET clause Specifies the columns to update and the new values to assign
to them. The syntax for column_assignments is:
column_name = value expression
Refer to Identifying columns to update, page 156 for a full
description of the syntax.
WHERE clause Restricts the rows that are updated to those that meet the
criteria in the qualification. Refer to The WHERE clause,
page 142 for a full description of WHERE clauses and
qualifications.
Return value
The UPDATE statement returns a collection whose result object has one property, rows_updated, that
contains the number of updated rows.

Update
Permissions
To use the UPDATE statement, the following conditions must be true:
• Your object-level permission for the dm_registered object in the repository that represents the
RDBMS table must be at least Browse
• Your table permission for the dm_registered object representing the table must be
DM_TABLE_UPDATE
permission to update the specified table. (The actual name of this permission depends on your
RDBMS.)
(For more information about security, object-level and table permissions, refer to the EMC
Description
This section contains usability information.
General notes
The SET clause identifies which columns are updated and the WHERE clause determines which rows
are updated. The server searches for all rows that meet the qualification and updates the specified
columns. If a WHERE clause is not included, then all rows are updated.
Note: The UPDATE statement with more than 100 different repeating attributes is not executed
successfully.
Identifying columns to update
In the SET clause, column_assignment has the format:
column_name = value_expression
where column_name is the name of a column in the table.
The value_expression can be a simple or complex expression but must resolve to a single value. It
can include literal values, other column names, special keywords, date and scalar functions, and
arithmetic expressions. The resulting value must be appropriate for the datatype of the specified
column.
Identifying rows to update
The WHERE clause qualification determines which rows are updated. The qualification consists of one
or more expressions that resolve to a single Boolean TRUE or FALSE. The server updates only those
rows for which the qualification is TRUE. The expressions in a qualification can include comparison
operators, literals, scalar and date functions, column names, arithmetic expressions, and column
predicates. Multiple expressions are joined together using logical operators.
Related statements
Delete, page 91
Insert, page 107

Update
Register, page 110
Examples
This example updates the column called chef_name:
UPDATE "recipes" SET "chef_name" = 'carol'
WHERE "chef_name" = 'carole'

Update...Object
Update...Object
Purpose
Updates an object in the repository.
Syntax
UPDATE [PUBLIC]type_name [(ALL)][correlation_var]
[WITHIN PARTITION partition_id {,partition_id}]
OBJECT[S] update_list
[,SETFILE filepath WITH CONTENT_FORMAT=format_name]
{,SETFILE filepath WITH PAGE_NO=page_number}
Arguments
Table 32. UPDATE...OBJECT argument descriptions
type_name Identifies the type of the object that you want to update.
Valid values are:
dm_assembly and its subtypes
dm_user and its subtypes
dm_group and its subtypes
dm_relation_type
dm_sysobject and its subtypes
If you have Sysadmin or Superuser user privileges, you can
also update the dmr_content object type.
correlation_var Defines a qualifier for the type name that is used to clarify
property references in the statement.

Update...Object
update_list Specifies the update operations to perform on the object.
Valid formats are:
name[[index]] = value append [n] property_name
property_name[[index]] [un]link 'folder path'
If you specify more than one operation, use commas to
separate them.
SETFILE clause WITH CONTENT_
FORMAT option
Adds the first content file to the new object. Refer to WITH
CONTENT_FORMAT option, page 161 for details.
SETFILE clause WITH PAGE_NO
option
Adds additional content to the object. Refer to WITH
PAGE_NO option, page 161 for details.
IN ASSEMBLY clause Restricts the statement to objects in a particular assembly.
document_id identifies the document with which the assembly
is associated. Use a literal object ID:
ID('object_id')
version_label specifies a particular version of the document.
Use a symbolic or an implicit version label. If you do not
include a version label, the server uses the version identified
by the document_id argument.
component_id specifies a particular component in the
assembly. Including the NODE option restricts the statement
to the specified component. Use a literal object ID to identify
the component:
ID('object_id')
The DESCEND keyword directs the server to update not
only all directly contained node components, but also any
indirectly contained components that meet the criteria. If
you do not include this keyword, the server updates only
the directly contained components that meet the criteria in
the statement.

Update...Object
SEARCH clause Restricts the statement to objects that meet the SEARCH
clause fulltext search condition. Refer to The SEARCH clause,
page 140 for a detailed description of a SEARCH clause.
WHERE clause Restricts the statement to objects that meet the qualification.
The qualification is an expression that evaluates to TRUE
or FALSE. It can include comparison operators, literals,
property names, scalar and date functions, special keywords,
predicates, and arithmetic expressions. Multiple expressions
can be joined together using logical operators.
Refer to The WHERE clause, page 142 for a detailed
description of the clause.
Return value
The UPDATE...OBJECT statement returns a collection whose result object has one property, called
objects_updated, that contains the number of objects updated.
Permissions
To update an object, you must have Write permission on the object.
To include the SETFILE clause in the statement, you must have Superuser privileges.
Description
General notes
To update an object, the object: cannot belong to a frozen assembly or a frozen or immutable virtual
document.
Aspects cannot be updated using the UPDATE...OBJECT statement; instead you can use DFC or IAPI.
The type_name argument specifies the type of objects to update. The server searches all objects of
the specified type and any subtypes for objects that meet any additional criteria you define in the
statement.
The keyword PUBLIC restricts the query to those objects with the r_is_public property set to TRUE.
If the query contains a SEARCH clause, the full-text search is restricted to those documents for
which ISPUBLIC is TRUE. When the server queries or searches only public objects, it uses only
the setting of r_is_public for security checks.
The keyword ALL directs the server to consider all versions of each object. If you do not include
ALL, the server only considers the version with the symbolic label CURRENT. You must enclose
ALL in parentheses.
The IN ASSEMBLY, SEARCH, and WHERE clauses restrict the scope of the statement. The IN
ASSEMBLY clause restricts the operation to a particular virtual document assembly or node

Update...Object
(component) within the virtual document. The SEARCH clause restricts the operations to indexed
objects that meet the fulltext search condition. The WHERE clause restricts the operations to objects
that meet the specified qualification. The clauses are applied in the following order:
• SEARCH
• IN ASSEMBLY
• WHERE
If you do not include any of these clauses, the updates are applied to all objects of that type for which
you have Write permission.
If any of the objects that are otherwise eligible for updating belong to a frozen assembly or an
unchangeable virtual document, then the entire statement is rolled back and no objects are updated.
The SETFILE clauses
A SETFILE clause adds new content to the end of the sequence of content files in the object or replaces
an existing content file. You cannot use SETFILE to insert a content file between two current files. You
cannot store the content file you add in a turbo store storage area.
You must have Superuser privileges to include the clause in the statement.
Any object capable of having content may have multiple associated content files. All files must have
the same content format and be ordered within the object.
The content format is defined when you add the first content file to the object. All subsequent
additions must have the same format. Consequently, specifying the format for content additions after
the first file is not necessary. Instead, you must specify the content’s position in the ordered list of
content files for the object.
To add the first content, use the SETFILE clause with the WITH CONTENT_FORMAT option.
To add additional content, use the SETFILE clause with the PAGE_NO option.
You can’t include both options in a single SETFILE clause.
WITH CONTENT_FORMAT option
Use this SETFILE option to add the first content file to an object. The syntax is:
SETFILE 'filepath' WITH CONTENT_FORMAT='format_name'
The format name is the name found in the name property of the format’s dm_format object.
WITH PAGE_NO option
Use this SETFILE option to add additional content to an object or replace existing content. The
syntax is:
SETFILE 'filepath' WITH PAGE_NO=page_number
The filepath must identify a location that is visible to Content Server. The file must have the same file
format as the existing content associated with the object.
The page number identifies the file’s position of the content file within the ordered contents of the
new object. You must add content files in sequence. For example, you cannot add two files and
specify their page numbers as 1 and 3, skipping 2. Because the first content file has a page number of
0, page numbers for subsequent additions begin with 1 and increment by 1 with each addition.

Update...Object
To replace a content file, specify the page number of the file you want to replace.
The update operations
The update_list defines the operations to perform. You can:
• Set the value of a single-valued or repeating property
• Add a value for a repeating property
• Insert a value into the list of values for a repeating property
• Remove a value from a repeating property
• Truncate a repeating property (remove all values)
• Link an object to a folder or cabinet or unlink an object from a folder or cabinet
• Move an object to a new folder or cabinet
Unless you have Superuser user privileges, you can only update read and write properties. These
properties have names beginning with a_ or have no prefix at all. If you have Superuser user
privileges, you can update read-only properties. These properties have names beginning with r_.
You can specify more than one update operation in a single statement. When an UPDATE...OBJECT
statement includes multiple operations, use commas to separate them. For example, the following
statement sets two properties and inserts a value into a repeating property. Notice the commas at the
end of each update operation clause:
UPDATE "dm_document" OBJECTS
SET "title" = 'A Cake Primer',
SET "subject" = 'cake',
INSERT "authors"[3] = 'georgette'
WHERE "r_object_id" = '090007354140004e'
The server processes the update operations in the order listed.
Setting a property value
To set the value of a single-valued or repeating property, use the following syntax:
set property_name[[index]] = value
The property_name argument is the name of the property. If the property is a repeating property, the
index identifies the position of the new value in the property’s list of values. The positions of the
values for a repeating property are numbered from zero. Enclose the index value in square brackets.
The value argument is the value to assign to the property. The value can be a literal or a subquery. If it
is a subquery, it cannot return more than one row. If it returns more than one row, the statement
returns an error.
Appending a value to a repeating property
To append a value to a repeating property, use the following syntax:
append [n]property_name = value
The property_name argument is the name of the property to which to append a new value. The
value argument specifies what value to add to the property’s ordered list of values. The value is
automatically added to the end of the repeating property’s list of values.
The value can be either a literal value or a subquery. The subquery can return any number of rows.
By default, the system appends a maximum of 20 rows. To override the default, use the [n] option

Update...Object
to define how many rows to append. You can use any integer number (to append that number of
rows) or an asterisk (*) (to append all rows).
Inserting a value into a repeating property
To insert a value into the list of values for a repeating property, use the following syntax:
insert property_name[[index]] = value
The property_name argument identifies the property. The index argument defines where to insert the
new value. The positions of all values for a repeating property are numbered from zero. If you do not
include the index, the server automatically inserts the new value in position zero, as the first element
in the ordered list. You must enclose the index in square brackets.
The value defines the value that you want to insert. The value can be either a literal value or a
subquery. If it is a subquery, it cannot return more than one row. If it returns multiple rows, the
statement returns an error.
When you insert a value, all values that follow the inserted value are renumbered. For instance, when
you insert a value at position [5], the value formerly at position [5] is moved up to position [6].
Consequently, if you are inserting more than one value in the property (for example, if the statement
is in a program loop), be sure to increase the index number each time you insert a value.
Removing values from repeating properties
To remove a value from a repeating property, use the following syntax:
remove property_name[[index]]
The property_name argument identifies the property. The index argument indicates the position of the
value to remove in the property’s ordered list of values. The positions of all values for a repeating
property are numbered from zero. If you do not include the index, the server removes the first value
(position 0) in the property. Enclose the index in square brackets.
When you remove a value, the remaining values are renumbered. For instance, when you remove the
value at position [5], the value formerly at position [6] is moved up to position [5]. Consequently, if
you are removing more than one value in the property (for example, if the statement is in a program
loop), be sure to start with the highest index number and decrement the index number each time
you delete a value.
For example, if you want to remove the values at positions 6 and 7, remove 7 first and then 6. If you
remove 6 first, the value at 7 is moved into position 6 and the value at 8 is moved into position 7.
When you remove 7, you are actually removing the value formerly in the position 8.
Truncating a repeating property
To truncate a repeating property (remove its values), use the following syntax:
truncate property_name[[index]]
The property_name argument identifies the property. The index argument specifies where in the
property’s list of values to begin the truncation. For example, if you specify property_name[4], the
server removes all values beginning with property_name[4]. If you do not specify an index level,
the server removes all values.
Linking or unlinking an object
To link an object to a folder or cabinet, use the following syntax in the update_list:
link 'folder path'

Update...Object
Linking an object adds a new link for the object. Current links are not affected.
To unlink an object from a cabinet or folder use the following syntax:
unlink 'folder path'
Unlinking removes only the specified link. Other links are not affected.
The folder path argument specifies the folder or cabinet to which you are linking the object. A folder
path has the following format:
cabinet_name{/folder_name}
Moving an object to a new folder or cabinet
To move an object to a new folder or cabinet, use the following syntax:
The folder path argument specifies the folder or cabinet to which to link the object. A folder path has
the following format:
cabinet_name{/folder_name}
Moving an object removes all current links and adds a link to the specified cabinet or folder.
Related statements
Examples
The following statement deletes all indexed documents that contain the word yeast:
SET "keywords" = 'yeasted'
SEARCH DOCUMENT CONTAINS 'yeast'
This next statement updates all documents that contain the word yeast but not the word rolls:
SET "keywords" = 'pastries'
SEARCH DOCUMENT CONTAINS 'yeast' AND NOT 'rolls'
The following statement updates all cabinets that have either Janine or Jeremy as their owner:
UPDATE "dm_cabinet" OBJECTS
SET "is_private" = TRUE
WHERE "owner_name"='janine' OR owner_name='jeremy'

Chapter 3
Administration Methods
Administrative methods are methods that perform a variety of administrative and monitoring tasks.
(Table 33, page 166, lists the methods by task categories.) They are executed in an application by
invoking either the DQL EXECUTE statement or the IDfSession.apply method. You can also execute
them interactively through Documentum Administrator.
This chapter first describes how to invoke the methods. Then it provides an alphabetical reference
to the methods. For each method, it provides:
• Invoking syntax
Only the DQL syntax is shown. For the IDfSession.apply method syntax, refer to the Javadocs.
• Arguments
• Return Value
• Permissions
• Description
• Related Administration Methods
• Examples
Invoking administration methods
To execute an administration method manually, use Documentum Administrator. All the methods
(except DO_METHOD, WEBCACHE_PUBLISH, and ROLES_FOR_USER) are available through
the Methods facility in repository Management. Many of the methods are also available through
category-specific pages.
Note: Because DO_METHOD is used to execute user-defined procedures, this method is
implemented as part of the Attributes page for user-defined methods.
To execute an administration method in an application, use either the IDfSession.apply method or
the DQL EXECUTE statement. You can also use EXECUTE to invoke an administration method
through IDQL.
The EXECUTE statement is the DQL equivalent of the API Apply method. You can use it to execute
any of the administration methods except PING or WEBCACHE_PUBLISH.
The EXECUTE statement is not case sensitive. You can enter the administration method name and
argument names in uppercase, lowercase, or any combination. You must enclose character string

values (including an object ID in a FOR clause) in single quotes when they are included in the
EXECUTE statement.
For information about using IDfSession.apply to invoke an administration method, refer to the
Javadocs.
Scope of the administration methods
Administrative methods execute in the context of the current repository scope. You cannot connect to
a repository and execute an administration method against a different repository.
Administration method operations
Table 33, page 166, lists the administration methods by category and describes the operation that
you can perform with each method.
Table 33. Administration methods by category
Category of Operation Administration Method Description
Process Management BATCH_PROMOTE, page
171
Promotes multiple objects to their
next lifecyle states.
Note: This method is not available
through Documentum Administrator
or using DQL EXECUTE.
CHECK_SECURITY, page
181
Checks a user or group’s permissions
level for one or more objects.
FIX_LINK_CNT, page 208 Updates the r_link_cnt property for
a specified folder.
GET_INBOX, page 216 Returns items in user’s Inbox.
MARK_AS_ARCHIVED,
page 248
Sets the i_is_archived property of a
dm_audittrail, dm_audittrail_acl, or
dm_audittrail_group to T.
PURGE_AUDIT, page 287 Removes audit trail entries from the
repository.
RECOVER_AUTO_
TASKS, page 296
Recovers work items claimed by a
workflow agent master session but
not yet processed.
ROLES_FOR_USER, page
308
Returns the roles assigned to a user in
a particular client domain.
Execute methods DO_METHOD, page 192 Executes system-defined procedures
such as lpq or who or user-defined
procedures.

HTTP_POST, page 224 Directs the execution of a method to
an application server.
Content storage management CAN_FETCH, page 173 Determines whether content in a
distributed storage area component
can be fetched by the server.
CHECK_RETENTION_
EXPIRED, page 177
Finds SysObjects in content-
addressed storage that have an
expired retention period or no
retention period.
CLEAN_LINKS —
This method is
deprecated. DFC Version
6 does not support linked
storage areas nor link
record objects, which
consequently, deprecates
this method.
Provides maintenance for linked store
storage areas.
On UNIX, this method cleans up
unneeded linkrecord objects and
directories and links associated with
linked storage areas.
On Windows, this method cleans
up unneeded linkrecord objects and
resets file storage object security.
DELETE_REPLICA , page
188
Removes a replica from a distributed
storage area.
DESTROY_CONTENT,
page 190
Removes a content object and its
associated file from the repository.
(Do not use this for archiving; use
PURGE_CONTENT instead.)
GET_FILE_URL, page 214 Returns the URL to a content file.
GET_PATH, page 221 Returns the path to a particular
content file in a particular distributed
storage area component.
IMPORT_REPLICA, page
228
Imports an external file as a replica of
content already in the repository.
MIGRATE_CONTENT,
page 251
Moves content files from one storage
area to another.
PURGE_CONTENT, page
293
Deletes a content file from a storage
area. (Used as part of the archiving
process.)
PUSH_CONTENT_
ATTRS, page 294
Sets the content metadata in a
content-addressed storage system for
a document stored in that storage
system.

REGISTER_ASSET, page
298
Queues a request for the creation of a
thumbnail, proxies, and metadata for
a rich media content file. The request
is queued to the Media Server.
running.
REPLICATE, page 303 Copies content in one component of
a distributed storage area to another
area.
RESTORE_CONTENT,
page 306
Moves a file or files from archived
storage to the original storage
location.
SET_CONTENT_ATTRS,
page 313
Sets the content_attr_name and
content_attr_value properties in the
content object associated with the
content file.
SET_STORAGE_STATE,
page 320
Sets the state of a storage area to
off-line, on-line, or read-only.
TRANSCODE_
CONTENT, page 324
Queues a request for a content
transformation to the Media Server.
running.
Database Methods DB_STATS, page 186 Provides database operation statistics
for a session.
DROP_INDEX, page 199 Drops an index.
EXEC_SQL, page 203 Executes SQL statements.
FINISH_INDEX_MOVES,
page 207
Completes an interrupted move
operation for an object type index.
GENERATE_
PARTITION_SCHEME_
SQL — Deprecated, page
209
Creates an SQL script to partition a
repository. (Deprecated)
MAKE_INDEX, page 245 Creates an object type index.
MIGRATE_TO_LITE,
page 265
Migrates standard SysObjects to
lightweight SysObjects and shareable
parents.

MOVE_INDEX, page 275 Moves an object type index from one
tablespace or segment to another.
Note: This is not supported on DB2.
PARTITION_
OPERATION, page 277
Partitions a repository.
REORGANIZE_TABLE,
page 301
Reorganizes a database table for
query performance.
UPDATE_STATISTICS,
page 327
Updates the statistics in a database
table.
Full-Text Methods ESTIMATE_SEARCH,
page 201
Returns the number of results
matching a particular SEARCH
condition.
MARK_FOR_RETRY,
page 249
Finds all content objects that have a
particular negative update_count and
marks them as awaiting indexing.
MODIFY_TRACE, page
273
Sets the tracing level for full-text
indexing operations.
Session Management CHECK_CACHE_
CONFIG, page 174
Requests a consistency check on a
particular cache config object.
GET_LAST_SQL, page
219
Returns the last SQL statement
issued.
GET_SESSION_DD_
LOCALE, page 223
Returns the locale in use for the
current session.
LIST_AUTH_PLUGINS,
page 232
Lists the authentication plugins
loaded by Content Server.
LIST_RESOURCES, page
233
Provides information about the server
operating system environment.
LIST_SESSIONS, page 237 Provides information about current,
active sessions.
LIST_TARGETS, page 240 Lists the connection brokers defined
as targets for the server.
The information is returned in a
collection with one result object
whose properties list the connection
brokers defined as targets for the
server.
LOG_ON, page 243 and
LOG_OFF, page 242
Turn server logging of information
about RPC calls on or off.
PING, page 286 Determine if a client has an active
server connection.

SET_APIDEADLOCK,
page 310
Sets a deadlock trigger on a particular
API method or operation.
SET_OPTIONS, page 317 Turn various tracing options on or off.
SHOW_SESSIONS, page
322
Provides information about current,
active sessions and a user-specified
number of timed-out sessions.
Web Publishing Management WEBCACHE_PUBLISH,
page 329
Invokes the dm_webcache_publish
method to publish documents to a
Web site.

BATCH_PROMOTE
BATCH_PROMOTE
Purpose
Promotes multiple objects to their next state.
Syntax
This method cannot be executed using the DQL EXECUTE statement.
Arguments
Table 34. BATCH_PROMOTE arguments
Argument Datatype Value Description
ARGUMENTS S comma-separated
list of object IDs
Identifies the objects to promote.
Use the objects’ object IDs. You
must enclose the list of object IDs
in single quotes.
Return value
BATCH_PROMOTE launches a method, dm_bp_batch. If the method executes successfully,
BATCH_PROMOTE returns a collection with one result object. If the method is not successfully
launched, BATCH_PROMOTE returns nothing. If BATCH_PROMOTE returns nothing, use
IDfSession.getMessage to retrieve the error message.
If BATCH_PROMOTE returns a collection, check the value of the method_return_val property of the
result object. A value of 36 indicates that all objects were successfully promoted. Any value other than
36 means that an error occurred and processing stopped at some point. Use an IDfSession.getMessage
to retrieve the error message.
Permissions
The user issuing the method must have the necessary permissions to promote the objects listed
in the argument.
Description
BATCH_PROMOTE allows you to promote multiple objects with one command. The objects can
be attached to different lifecycles and can be in different states.
The method checks the permissions on each object specified to ensure that the user issuing the method
has the correct permissions to promote the object. Then the method checks that associated lifecycle

BATCH_PROMOTE
or lifecycles are valid. Finally, the method executes any entry criteria specified for the objects’ next
states. If BATCH_PROMOTE encounters an error at this stage, the method exits and returns nothing.
If the permissions are correct, the lifecycles are valid, and any entry criteria are fulfilled,
BATCH_PROMOTE launches the dm_bp_batch method. The method performs any action procedures
needed, as a single transaction. If an error occurs, the method exits and reports an error. Any objects
promoted before the action procedure error remain promoted; the object whose procedure caused
the error and any objects in the list after it are not promoted. For example, if the argument list
includes 6 objects and an error occurs on an action procedure for object 4 in the list, objects 1, 2,
and 3 are promoted. Objects 4, 5, and 6 are not.
There are two limitations on the use of BATCH_PROMOTE:
• BATCH_PROMOTE cannot run actions on behalf of the lifecycle_owner. If the value in the
a_bpaction_run_as property in the repository configuration (docbase config object) is set to
lifecycle_owner, BATCH_PROMOTE exits with an error.
• You can specify a maximum of 200 objects in each execution of BATCH_PROMOTE.
Related administration methods
None

CAN_FETCH
CAN_FETCH
Purpose
Determines if the server can fetch a specified content file.
Syntax
EXECUTE can_fetch FOR 'content_object_id'
Arguments
CAN_FETCH has no arguments.
Return value
CAN_FETCH returns a collection with one query result object. The object has one Boolean property
that is set to TRUE if the fetch is possible or FALSE if it is not.
Permissions
Anyone can use this method.
Description
If a repository has a distributed storage area, it is possible to configure the servers so that they can
fetch from all distributed storage area components, from a subset of the components, or only from the
local component. (For information about configuring this capability, refer to the EMC Documentum
Content Server Distributed Configuration Guide.) In such repositories, you can use the CAN_FETCH
method to determine whether a server can fetch from a particular distributed storage area component.
In a content server configuration, the CAN_FETCH method is executed by the content server.
GET_PATH, page 221
Examples
The following examples determine whether Content Server can fetch the content file associated with
the content object 06000002472185e1:
EXECUTE can_fetch FOR '06000002472185e1'

CHECK_CACHE_CONFIG
CHECK_CACHE_CONFIG
Purpose
Requests a consistency check on a particular cache config object.
Syntax
EXECUTE check_cache_config [FOR 'cache_config_id']
[WITH argument = value ][,argument = value]
Do not include the FOR clause if you include the CONFIG_NAME argument.
Arguments
Table 35. CHECK_CACHE_CONFIG arguments
CONFIG _NAME S name of cache config
object
Identifies the cache config
object on which to perform the
consistency check. Use the cache
config’s object name.
This is an optional argument. If
you do not include it, you must
include the object ID of the cache
config object.
FORCE _CHECK B T (TRUE) or
F (FALSE)
TRUE directs Content Server to
re-execute the queries regardless
of the server_check_interval
value. (Refer to the General Notes
for a description of the use of
the server_check_interval in the
context of CHECK_CACHE_
CONFIG.)
This is an optional argument. The
default is F (FALSE).
Return value
The method returns a collection with one query result object. The object has the properties listed
in Table 36, page 175.

CHECK_CACHE_CONFIG
Table 36. Properties in the CHECK_CACHE_CONFIG result object
Property Datatype Description
r_object_id ID Object ID of the cache config object
r_last_changed_date Date/time The date and time at which Content Server
last validated the query results and found that
something had changed.
r_last_checked_date Date/time The date and time at which Content Server
last ran the queries to determine whether the
results had changed.
server_check_interval integer How long the server waits between validations
of the query results. The time is expressed in
seconds.
client_check_interval integer The consistency check interval used by
clients when issuing fetch or query methods
that include this cache config object as an
argument.
server_time Date/time Current server time
server_time_secs integer Current server time in seconds
cache_config_exists Boolean F (FALSE) if the specified cache config object
is not found. T (TRUE) otherwise.
If an error occurs, the method returns an error rather than the collection.
Permissions
The cache config object must be owned by a Superuser.
You must have at least Browse permission on the cache config object to issue this method. To issue
the method with FORCE_CHECK set to TRUE, you must be a Superuser or have Execute Procedure
permission on the cache config object.
Description
CHECK_CACHE_CONFIG directs Content Server to check whether the data defined by a particular
cache config object is current. (Cache config objects define queries whose results are cached
persistently on the client.) To determine if the data is current, the server compares the current time
to the date and time in the object’s r_last_checked_date property. If the difference is less than
the interval defined in server_check_interval, indicating the interval has not expired, the data is
considered current and the server does not recompute the data. If the interval has expired, the data
is considered out of date. The server executes the dm_CheckCacheConfig method to recompute
the data. The method executes the queries defined in cache_element_queries, computes a hash of
the results, and stores the hash in the i_query_result_hash property. The method returns the new
computation date and time in the query result object’s r_last_checked_date property.
You can use the FORCE_CHECK argument to override the server check interval and force Content
Server to execute the queries.

CHECK_CACHE_CONFIG
For information about persistent client caching and cache config objects and their use, refer to EMC
Documentum Content Server Fundamentals Guide. You can execute the dm_CheckCacheConfig method
manually, using a DO_METHOD administration method. For an example of using DO_METHOD
to call dm_CheckCacheConfig, refer to the EMC Documentum Content Server Administration and
Configuration Guide. However, explicit calls to the method are not normally necessary. The calls occur
automatically, as a side effect of referencing cache config objects in fetch and query methods.
None
Examples
The following example identifies the cache config object by its object ID and forces the recomputation
of the data:
EXECUTE check_cache_config FOR '08000002723ae36b'
WITH force_check = true
This next example identifies the cache config object by its owner and name.
EXECUTE check_cache_config
WITH config_name='johndoe.report_app_config'

CHECK_RETENTION_EXPIRED
Purpose
Generates a list of objects whose content, stored in content-addressed storage, has an expired
retention period or a zero retention period.
Syntax
EXECUTE CHECK_RETENTION_EXPIRED
WITH QUERY='where_clause'
[,SELECT_LIST='property_list']
[,INCLUDE_ZERO_RETENTION_OBJECTS=TRUE|FALSE]
Arguments
Table 37. CHECK_RETENTION_EXPIRED arguments
QUERY S where_clause Defines which SysObjects are to
be checked for an expired or no
retention period. Consists of a
DQL SELECT where clause. Only
literal values and properties defined
for the dm_sysobject type may be
referenced. The string must be
enclosed in single quotes.
SELECT_LIST S property_list Identifies additional properties
whose values are to be returned in
the result object.
Only properties defined for the
dm_sysobject object type may be
included.
Separate multiple property names
with commas and enclose the entire
list in single quotes.
INCLUDE_ZERO_
RETENTION_
OBJECTS
B T (TRUE) or F
(FALSE)
If set to T, the method checks objects
stored in content-addressed storage
areas that allow but do not require a
retention period.
The default is F, meaning that only
objects stored in content-addressed
storage areas that require a retention
period are checked.

Return value
CHECK_RETENTION_EXPIRED returns a collection of result objects, each representing one
SysObject whose retention has expired or that had no retention period. The result objects have five
default properties plus any specified in the SELECT_LIST argument. The default properties of
the result object are:
• r_object_id
• object_name
• a_storage_type
• r_creation_date
• retention_date
The retention_date property is a computed property. The date value is the GMT equivalent of
the retention period as it is defined in the time zone of the Centera storage system. For example,
suppose the Centera system is in the Eastern time zone (EST) and the client on which user is
working is in the Pacific time zone (PST). If the user sets the retention value to March 15, 2004
11 a.m. PST, the retention value is stored as March 15, 2004 2 p.m.—the Eastern time zone
equivalent of March 15, 2004 11 a.m. The value returned in the computed property is the GMT
equivalent of March 15, 2004 2 p.m.
Permissions
You must have Sysadmin or Superuser privileges to execute this method.
Description
The CHECK_RETENTION_EXPIRED method is used by the RemoveExpiredRetnObjects
administration job to find the content in content-addressed storage that has an expired retention
period. The method returns a list of the objects. It does not remove the content from the storage area,
nor does it remove from the repository any of the objects that contain the content.
A content-addressed storage area can have three possible retention period configurations:
• The storage area may require a retention period.
In this case, the a_retention_attr name property is set and the a_retention_attr_req is set to T.
• The storage area may not allow a retention period.
In this case, the a_retention_attr name property is not set and the a_retention_attr_req is set to F.
• The storage area may allow but not require a retention period.
In this case, the a_retention_attr name property is set , but the a_retention_attr_req is set to F.
By default, the method operates only on objects that are stored in content-addressed
storage areas that require a retention period. The method never includes objects stored in
content-addressed storage areas that do not allow retention periods. If you want it to examine
objects stored in content-addressed storage areas that allow but do not require a retention
period, you must set the INCLUDE_ZERO_RETENTION_OBJECTS argument to true (refer to
INCLUDE_ZERO_RETENTION_OBJECTS argument, page 179, for more information).

QUERY argument
The QUERY argument’s where clause is a DQL where clause qualification. It is used to select objects
for possible inclusion in the results returned by the method. The objects that fulfill the QUERY
argument and are stored in an appropriate content-addressed storage area are then examined to
determine whether the retention period has expired. If so, the object is included in the results.
The QUERY argument can reference only literal values or properties defined for the dm_sysobject
object type. If the where clause includes single quotes, you must escape them with single quotes.
For example:
...QUERY,S,'a_storage_type=''castore_1''
and r_creation_date > DATE(01/01/2003)'
INCLUDE_ZERO_RETENTION_OBJECTS argument
By default, the method does not include objects whose content has a 0 retention period because the
assumption is that such content is meant to be kept forever. However, in a storage area that allows
but does not require a retention period, a 0 retention period can be result from two possible causes:
• The user deliberately set no retention period, and consequently, the server set the retention
period to 0
• The user specified a retention date that had already elapsed. When this occurs, the server sets the
retention period to 0.
Because the meaning of 0 is ambiguous in such storage areas, the method supports the
INCLUDE_ZERO_RETENTION_OBJECTS argument to allow you to include content with a zero
retention in storage areas that allow but do not require a retention period.
If you set INCLUDE_ZERO_RETENTION_OBJECTS to T, when the method examines objects in
storage areas that allow but do not require a retention period and it will include in the results any
object with an expired or zero retention period.
SELECT_LIST argument
The SELECT_LIST argument allows you to include additional SysObject properties in the result
objects. You must enclose the argument’s value in single quotes.
None
Examples
The following example checks SysObjects stored in the content-addressed storage area named
"castore_2” owned by the user named "John Arthur”. It also adds the title and subject property
values to the result objects.
EXECUTE check_retention_expired
WITH QUERY='a_storage_type=''castore_2'' and
owner_name=''John Author''',
SELECT_LIST='title,subject'
This example directs the method to include objects in a storage area that allows but doesn’t require a
retention period:

EXECUTE check_retention_expired
WITH QUERY='a_storage_type=''castore_3'' and
owner_name=''Mary Writer''',
SELECT_LIST='title,subject',
INCLUDE_ZERO_RETENTION_OBJECTS=true

CHECK_SECURITY
CHECK_SECURITY
Purpose
Checks a user’s or group’s access permissions on one or more objects or checks a user’s or group’s
permission level in one or more ACLs.
Syntax
EXECUTE check_security WITH user_name='name'|group_name='name',
level=security_level,object_list='list_of_objectids'
Arguments
Table 38. CHECK_SECURITY arguments
USER_NAME S name Name of the user for whom you
are checking permissions. If you
include this argument, do not
include GROUP_NAME.
GROUP_NAME S name Name of a group for which you
are checking permissions. If you
include this argument, do not
include USER_NAME.
LEVEL I security_ level Minimum access permission level
for which you are checking. Valid
values are:
1, for None
2, for Browse
3, for Read
4, for Relate
5, for Version
6, for Write
7, for Delete
OBJECT_LIST S list of object IDs The objects being checked. This
can be a list of SysObject object
IDs, a list of ACL object IDs, or
both. Use a space to delimit the
individual items.

CHECK_SECURITY
Return value
CHECK_SECURITY returns a collection of query result objects. The objects have one property,
r_object_id. The property contains the object IDs of those objects submitted in the OBJECT_LIST
argument for which the user or group has at least the permission level identified in the LEVEL
argument.
Permissions
You must have Superuser privileges to use this method.
Description
There are two uses for CHECK_SECURITY:
• You can use it to determine whether a user or group has a particular access permission or better
for a group of SysObjects
• You can use to determine whether a user or group has entries in one or more ACLs that give the
user or group a particular access permission (or better).
To determine whether a user or group has at least the permission identified in the LEVEL argument
for a particular document, include the document’s object ID in the OBJECT_LIST argument.
To determine whether a user or group has entries in an ACL that give at least the permission level
identified in the LEVEL argument or better to the user or group, include the object ID of the ACL in
the OBJECT_LIST argument.
The method ignores all object IDs that do not represent SysObjects or ACLs or object IDs that are
incorrectly identified (too few characters, too many characters, and so forth).
None
Examples
This example checks to determine whether the user LibbyLoo has at least Write permission on three
documents:
EXECUTE check_security WITH user_name='LibbyLoo',
level=5,object_list='09000001734912ac
0900000153813f2b 0900000116572af3'
This example determines whether the group Engineering has accessor entries that give the group
at least Read permission in the ACLs included in the object list:
EXECUTE check_security WITH group_name='Engineering',
level=3,object_list='4500000112562e4c 450000017351213c'

CLEAN_DELETED_OBJECTS
Purpose
Removes deleted lightweight and parent objects from the repository.
Syntax
To remove a deleted parent and all its children:
EXECUTE clean_deleted_objects [[FOR] deleted_parent_object_id]
To remove private parents that no longer have lightweight children:
EXECUTE clean_deleted_objects WITH execution_mode='remove_orphaned_parents'
Arguments
Table 39. CLEAN_DELETED_OBJECTS arguments
execution_mode S remove_
orphaned_parents
The only value allowed is
remove_orphaned_parent.
Return value
Returns a list of removed objects
Permissions
You must have Sysadmin or Superuser privileges to use this method.
Description
CLEAN_DELETED_OBJECTS removes deleted objects from the database. Since a shareable parent
object can have many lightweight children, for performance reasons, all the lightweight children are
not removed when a shareable parent is deleted. The dmClean job can be configured to remove these
objects as part of the usual repository cleanup, or this method can be used to start removing the
objects immediately.
When a lightweight object is reparented from its private parent to a shareable parent, the
private parent is not deleted until the CLEAN_DELETED_OBJECTS method is run with
remove_orphaned_parents set.

CLEAN_LINKS — Deprecated
Purpose
On Windows, this method removes unneeded dmi_linkrecord objects and resets security on file
store objects to the original state. On UNIX, this method removes unneeded dmi_linkrecord objects
and the auxiliary directories and symbolic links.
Syntax
EXECUTE clean_links [WITH force_active=true|false]
Arguments
Argument Name Datatype Value Description
FORCE _ACTIVE B T (TRUE) or F
(FALSE)
TRUE directs the server to clean
the links in all sessions. FALSE
directs the server to clean links only
in inactive sessions. The default is
FALSE.
Return value
CLEAN_LINKS returns a collection with one query result object. The object has one Boolean property
whose value indicates the success (TRUE) or failure (FALSE) of the operation.
Permissions
Description
Note: The CLEAN_LINKS method is deprecated. DFC Version 6 does not support linked storage
areas. Consequently, the CLEAN_LINKS method, which supports those storage areas, is deprecated.
Linked storage areas are handled differently on UNIX and Windows platforms. Consequently, the
behavior of CLEAN_LINKS is slightly different on each platform.
On UNIX, when a linked storage area is referenced, the server creates a dmi_linkedrecord object and
auxiliary directories and symbolic links. On Windows, the server creates a dmi_linkrecord object and
resets the security of the linked storage object. Generally, the server removes unneeded linkrecord
objects and, on UNIX, any unneeded auxiliary directories and symbolic links. On Windows, the
server typically resets the linked storage object’s security as needed. However, there are conditions
that do not allow the server to do this work.

You can use CLEAN_LINKS to perform this work manually. We recommend running CLEAN_LINKS
regularly. (Note that CLEAN_LINKS is run automatically whenever the server is restarted.)
To determine if you need to run CLEAN_LINKS, run LIST_SESSIONS and compare the reported
session IDs (in session[x]) to the values in the session_id properties of the dmi_linkrecord objects. If
the session_id property in a link record object references an inactive session (a session not reported
in LIST_SESSIONS), that link record object is not needed. Depending on how many unneeded
link record objects you find, you may want to run CLEAN_LINKS to remove them and perform
the other, associated clean up work.
LIST_SESSIONS, page 237
Examples
This example runs CLEAN_LINKS against only inactive sessions because the FORCE_ACTIVE
argument is defaulted to FALSE:
EXECUTE clean_links
The following example removes the unneeded link record objects for all repository sessions, active
and inactive:
EXECUTE clean_links WITH force_active=true

DB_STATS
DB_STATS
Purpose
Returns a set of statistics about database operations for the current session.
Syntax
EXECUTE db_stats [WITH clear = true|false]
Arguments
Table 40. DB_STATS arguments
clear Boolean true | false Indicates whether you want to
clear the counters. The default is
FALSE.
Return value
DB_STATS returns a collection with one result object, described in Table 41, page 186.
Table 41. Query result object properties for DB_STATS administration method
updates integer Number of SQL update operations performed
inserts integer Number of SQL insert operations performed
deletes integer Number of SQL delete operations performed
selects integer Number of SQL select operations performed
rpc_ops integer Number of RPC calls between Content Server
and the RDBMS
Note: This may not be implemented for all
databases.
ddl_ops integer Number of data definition operations performed
Data definition operations are operations such
as create table or drop table.
max_cursors integer Maximum number of concurrently open cursors

DB_STATS
Permissions
Description
DB_STATS returns a set of statistics about database operations for the current session. The statistics
are counts of the numbers of:
• Inserts, updates, deletes, and selects executed
• Data definition statements executed
• RPC calls to the database
• Maximum number of cursors opened concurrently during the session
None
Examples
The following example uses EXECUTE to invoke DB_STATS. It returns the statistics for the current
repository session and resets the counters to zero:
EXECUTE db_stats WITH clear=true

DELETE_REPLICA
DELETE_REPLICA
Purpose
Removes a content file from a component of a distributed storage area.
Syntax
EXECUTE delete_replica FOR 'content_object_id'
WITH STORE='storage_name'
Arguments
Table 42. DELETE_REPLICA arguments
store string storage_name Specifies the component storage
area that contains the file to
remove. This is a required
argument. Use the storage area’s
name as defined in its storage
object.
Return value
DELETE_REPLICA returns a collection with one query result object. The object has one Boolean
property that indicates the success (TRUE) or failure (FALSE) of the operation.
Permissions
Description
DELETE_REPLICA removes a content file from a storage area that is a component of a distributed
storage area. Using DELETE_REPLICA also modifies the replica record object associated with the
content. The replica record maintains the information that allows the server to fetch the content from
any of the component storage areas. When you remove a file from a component storage area using
DELETE_REPLICA, this record is updated also.
To use DELETE_REPLICA, you must be using one server for both data and content requests. If the
configuration is set up for content servers, you must issue a connection request that bypasses the
content server to use DELETE_REPLICA in the session.

DELETE_REPLICA
IMPORT_REPLICA, page 228
Examples
This example removes the content file associated with the content object identified by
06000001684537b1 from the distcomp_2 storage area:
EXECUTE delete_replica FOR '06000001684537b1'
WITH STORE='distcomp_2'

DESTROY_CONTENT
DESTROY_CONTENT
Purpose
Removes content objects from the repository and their associated content files from storage areas.
Syntax
EXECUTE destroy_content FOR 'content_obj_id'
Arguments
DESTROY_CONTENT has no arguments.
Return value
DESTROY_CONTENT returns a collection with one query result object. The object has one Boolean
Permissions
Description
DESTROY_CONTENT removes orphaned content objects. An orphaned content object is a content
object that has no values in its parent_id property. The method also removes the content file
referenced by the orphaned content object if there are no other content objects that reference that file.
The DESTROY_CONTENT method is the method invoked by dmclean.
To run DESTROY_CONTENT, you must be using one server for both data and content requests. If
the configuration is set up for content servers, you must issue a connection request that bypasses
the content server to use DESTROY_CONTENT in the session.
Note: Do not use this method to archive content. It removes the file’s content object, rather than
marking it off-line.
Destroying content in content-addressed storage
If the storage area defined in the content object is a content-addressed storage system (a ca store
storage area), Content Server first determines whether the storage area allows content to be
deleted. If not, DESTROY_CONTENT fails with an error. If the storage system allows deletions,
DESTROY_CONTENT checks the retention period specified for the content represented by the
content object. If there are multiple addresses recorded for the content, Content Server checks the
retention period stored with each address. The retention period for all addresses must be expired
before Content Server can destroy the content.

DESTROY_CONTENT
Note: Some operations, such as those that modify the metadata or object replication, result in the
generation of a new content address for a content file. These additional addresses are stored in
the i_contents property of a subcontent object.
If that period has not expired, DESTROY_CONTENT cannot remove the file from the storage area.
The method fails with an error. If the retention period has expired or there is none set, the method
removes the content.
PURGE_CONTENT, page 293
Examples
This example uses the EXECUTE statement to invoke DESTROY_CONTENT:
EXECUTE destroy_content FOR '06000001684537b1'

DO_METHOD
DO_METHOD
Purpose
Executes an external program, a Docbasic script, or a Java method. (Note that use of Docbasic
is deprecated.)
Syntax
EXECUTE do_method WITH METHOD='method_name'
{,arguments='value'}
Arguments
Table 43. DO_METHOD arguments
save_results Boolean true | false Indicates if you wish to save the
results into a document.
arguments string command-line
arguments
Specifies the command-line
arguments for the program.
If the method is to be executed
on the application server, UTF-8
characters are accepted for the
argument strings.
If the method is to be executed
on the method server or Content
Server, only characters from the
server OS code page are accepted
for the argument strings.
Refer to Specifying the arguments,
page 195, for more information.
time_out integer value Specifies the length in seconds of
the default time-out period for the
program that you are executing.
Refer to Defining a time out period
, page 196 for more information.
method string method_name Name of the method object
representing the script, program,
or Java method to execute.
This argument is required.

DO_METHOD
launch_direct Boolean true | false Indicates whether to execute the
program using the Windows
(UNIX) system API or the exec
API. Set this to TRUE to use
the system API. By default, it is
FALSE, which uses the system
call. Refer to Launching directly,
page 196 for more information.
This argument is ignored if the
method’s use_method_server
property is TRUE.
launch_async Boolean true | false Indicates whether to execute
the program asynchronously.
TRUE executes the method
asynchronously. The default is
FALSE.
Setting this argument to TRUE
is ignored if the SAVE_RESULTS
argument is also TRUE. Refer to
Launching asynchronously, page
196 for more information.
run_as_server Boolean true | false Indicates whether to run the
method under the server’s
account. If the argument is not set,
the server uses the value of the
method’s run_as_server property.
You must have Sysadmin or
Superuser privileges to set this
to TRUE on the command line
if the method identified in the
METHOD argument has its
run_as_server property set to
FALSE.
This argument must be TRUE
for methods that have their
use_method_server property set
to TRUE.
trace_launch Boolean true | false Indicates whether to generate
tracing information for the
method.
Recording tracing information
and program output, page 198

DO_METHOD
describes where the information
is stored.
Return value
A DO_METHOD function returns a collection that contains one query result object. Table 44, page
194 lists the properties of this object.
Table 44. Properties of the query result object returned by DO_METHOD
result Integer or ID If the program you launched was dmfilescan or
dmclean, this property holds the object ID of the
script run by the utility.
Otherwise, the property contains an integer value
indicating the success or failure of the program.
For methods executed on an application server, the
values are:
0, meaning a status of HTTP/1.1 2xx
1, meaning a status of HTTP/1.1 5xx
-1, for any other status
result_doc_id ID Contains the object ID of the document that contains
the output of the program. (Note that the output is
captured even if the program times out.)
This property is present only when the
SAVE_RESULTS argument is set to TRUE.
launch_failed Boolean Indicates whether the program was successfully
executed. T means that the program was not
launched and the method_return_val property
is meaningless. F means that the program was
launched and the return value is the exit code of the
process when the method terminated.
method_return_val Integer Contains the return value of the executing program.
For methods executed on the application server, the
values are:
0, for a status of HTTP/1.1 2xx
1, for any other status
For all other methods, if the program times out, this
value is 0.

DO_METHOD
If launch_async is T, then method_return_val is
always 0.
os_system_error String Contains an operating system error if one occurs.
This property can be empty.
timed_out Boolean Indicates whether the DO_METHOD was
terminated due to a timeout. T means the method
was terminated while waiting for the program to
complete. F means the method received a response.
time_out_length Integer The length of the time-out period.
Permissions
If the method you are executing has the run_as_server property set to FALSE in its method object, you
must have at least Sysadmin privileges to override that setting in the DO_METHOD command line.
Description
Use DO_METHOD to execute a Java method, a Docbasic script, or other executable program. (Note
that Docbasic is deprecated.) You can direct the method execution to the Java method server or
Content Server. Docbasic programs are executed by DFC, using an emulation package. Which you
choose depends on the language in which the program is written and how your site is configured.
For details about each of the execution agents and how to direct a method to the agents, refer to the
EMC Documentum Documentum Content Server Administration and Configuration Guide.
To use DO_METHOD, the invoked script or program must be defined in the repository by a method
object. A method object contains properties that tell the server the program’s command line and
arguments and provide parameters for executing the program. If the program is a Docbasic script,
the script is stored as the content of the method. (For information about creating method objects, refer
to the EMC Documentum Content Server Administration and Configuration Guide.)
Specifying the arguments
There are no restrictions on the format of the argument list for methods executed by Content Server.
If the method is to be executed using the Java method server (use_method_server is T and
method_type is java), you must pass both argument names and values in the ARGUMENTS value.
The format for value is:
-argument_name argument_value
Separate multiple arguments with a single space. For example:
EXECUTE do_method WITH METHOD='payroll_report',
ARGUMENTS='-docbase accounting
-user auditor
-ticket DM_TICKET=0000000222a02024.accounting@host01'

DO_METHOD
If you are directing the DO_METHOD to the Java method server, Content Server encodes the
arguments using application/x-www-form-urlencoded format. For example, suppose you issued
the following DO_METHOD:
EXECUTE do_method WITH METHOD='paymethod',
ARGUMENTS='-docbase accounting
-user paymaster
-ticket DM_TICKET=0000000222a02054.accounting@host01
-document "weekly record"
Content Server sends the arguments in the following format:
docbase=accounting&user=paymaster&ticket=DM_TIICKET%
3D0000000222a02054.accounting%4Dhost01&document=weekly+record
Defining a time out period
The TIME_OUT argument defines a time out period for the program or script you are executing.
Assigning a value to this argument overrides any value assigned to the timeout_default property in
the program’s method object.
The value that you specify cannot be greater than the value assigned to the method object’s
timeout_max property or less than the value assigned to the object’s timeout_min property. If the
maximum or minimum time that you specify on the DO_METHOD command line violates this rule,
the server ignores your specification and uses the timeout_max or timeout_min value specified
in the method object.
Launching directly
When you execute DO_METHOD using Content Server as the execution agent, the server calls the
Windows or UNIX (depending on the platform on which the method is running) system API to
execute it by default. If you set LAUNCH_DIRECT to TRUE, the server calls the exec API instead.
This API executes the program or script directly, instead of calling a shell script, which provides the
advantage of better error reporting.
To execute with LAUNCH_DIRECT set to TRUE, the method’s method_verb property must contain
a fully qualified pathname.
Launching asynchronously
There are two ways to launch a program or script asynchronously:
• Use the ARGUMENTS argument to DO_METHOD to append an ampersand (&) to the end of
the command line arguments. If there are multiple command-line arguments, the ampersand
must be the last argument in the list.
For example,
EXECUTE do_method
WITH method = 'validate',arguments = '&'
• Set the DO_METHOD’s LAUNCH_ASYNC argument to TRUE.
For example,
EXECUTE do_method
WITH method = 'validate',launch_async = TRUE
If you launch asynchronously and the method is executing on the application server, it is not possible
to capture the method’s output.

DO_METHOD
Saving results
For DO_METHOD methods that are executing on the application server, EMC Documentum provides
a simple, example interface that captures the program output so the output can be saved into the
repository if SAVE_RESULTS is TRUE. This interface is described fully in the EMC Documentum
Content Server Administration and Configuration Guide.
Running as the server account
By default, the program invoked by the DO_METHOD runs as the logged-in user. If you want the
program to execute under the Content Server’s account, you can:
• Set the run_as_server property in the associated method object to TRUE and set the
RUN_AS_SERVER argument in the command line to TRUE.
• Set only the RUN_AS_SERVER argument in the command line to TRUE.
By default, both the run_as_server property and the RUN_AS_SERVER argument are FALSE. To run
as the server account, either both must TRUE or you must override the property setting by setting
the RUN_AS_SERVER argument to TRUE. Overriding the property in the DO_METHOD command
line by setting the RUN_AS_SERVER argument to TRUE requires Sysadmin or Superuser privileges.
(Overriding the property if it is set to TRUE by setting the RUN_AS_SERVER argument to FALSE
requires no special privileges.)
If you execute DO_METHOD using either the emulation package or the Java method server as
the execution agent, you must set both the method’s run_as_server property to TRUE and the
RUN_AS_SERVER argument to TRUE.
Notes:
• If LAUNCH_DIRECT is set to TRUE, either on the command line or in the method’s property,
RUN_AS_SERVER must also be set to TRUE. If it is not, the method does not execute correctly.
• Content Server uses the assume user program to run procedures and print jobs as the logged-in
user. If you disable the assume user program (by setting the assume_user_location property in
the server config object to a blank), Content Server runs all procedures and all print jobs under
its account.
Ensuring security on the application server
There are two security issues to consider when using an application server to execute a DO_METHOD
that invokes a Java servlet or method:
• Determining the origin of the HTTP_POST request
• Login without passwords (this is only possible on Windows platforms)
Issuing a DO_METHOD to execute on the application server sends an internal HTTP_POST request to
the application server. The invoked servlet ensures that the generated request comes from a machine
that hosts a Content Server by checking the IP address of the sender against a list of repositories.
This list is set up when the application server is installed and configured. If the sender’s IP address
doesn’t match the IP address of a repository host, the request fails.
The application server runs as the Content Server installation owner. Consequently, the servlet it
invokes to execute DO_METHOD calls also runs as the installation owner. On Windows platforms,
the current operating system user is allowed to log in to the repository without providing a password.
Consequently, a servlet or an invoked Java method can log into the repository through the Content
Server with superuser privileges without providing a password.

DO_METHOD
If you write a method that uses that process to log in, you may want to ensure that the actual user
who issues the DO_METHOD to invoke the method has the appropriate privileges to execute the
program as a superuser. To do this, send the current user’s name and a login ticket to the method
in the arguments. The method can use these to log in and check the privileges of the user before
connecting as the current operating system user.
Recording tracing information and program output
Trace information and program output are two different sets of information. Trace information is
information about the DO_METHOD invocation and its success or failure. The program output is
the results returned by the program called by the DO_METHOD.
Setting TRACE_LAUNCH to TRUE logs tracing information, of up to 2047 characters, to the
repository log. Setting SAVE_RESULTS to TRUE saves the execution output of the invoked program.
EXEC_SQL, page 203
HTTP_POST, page 224
Examples
This example executes the update_accounts procedure, specifying a timeout period of 120 seconds (2
minutes):
EXECUTE do_method
WITH method_name='update_accounts',
time_out=120
The following example executes the user-defined procedure run_rpt_newaccts and saves the results
to a document:
dmAPIGet("apply,s0,NULL,DO_METHOD,
METHOD,S,run_rpt_newaccts,SAVE_RESULTS,B,T")
This example executes a method on the application server:
dmAPIGet("apply,s0,NULL,DO_METHOD,
METHOD,S,check_vacation,SAVE_RESULTS,B,T,
ARGUMENTS,S,-docbase HumanRSRC -user adminasst
-ticket DM_TICKET=0000000221c02052.HumanRSRC@hr025
-document "monthly payroll")

DROP_INDEX
DROP_INDEX
Purpose
Destroys a user-defined object type index.
Syntax
EXECUTE drop_index [[FOR] 'dmi_index_id']
[WITH name = 'index_name']
Do not include the FOR clause if you include the NAME argument.
Arguments
Table 45. DROP_INDEX arguments
name string index_name Identifies the index by the name of
its index (dmi_index) object.
Return value
DROP_INDEX returns a collection that contains one query result object. The object has one Boolean
property whose value indicates the success (TRUE) or failure (FALSE) of the operation.
Permissions
Description
You can obtain an object type index’s name or object ID from the dmi_index type table. Each row in
this table represents one index in the repository.
FINISH_INDEX_MOVES, page 207
MAKE_INDEX, page 245
MOVE_INDEX, page 275

DROP_INDEX
Examples
These examples illustrate using EXECUTE to drop a user-defined index on the dm_user object type.
The first example identifies the index by its name, user_index, and the second example identifies the
index by its object ID.
EXECUTE drop_index WITH name='user_index'
EXECUTE drop_index FOR '1f00000011231563a'

ESTIMATE_SEARCH
ESTIMATE_SEARCH
Purpose
Returns the number of results matching a particular SEARCH condition.
Syntax
EXECUTE estimate_search [[FOR] 'fulltext_index_obj_id']
WITH [name = 'index_name'] [,type = 'object_type']
[,query = 'value']
Arguments
Table 46. ESTIMATE_SEARCH arguments
name string index_name Identifies the index to be searched.
This is optional. You can specify
the object ID of the fulltext index
object instead.
type string object_type Identifies the type of objects to
be searched. All subtypes of the
specified type are included in the
search also.
If not included, the method
searches all object types in the
index.
query string value Defines the word or phrase for
which you are searching. You can
use a Boolean Plus expression for
the value or a particular word or
phrase.
If not included, the method’s
return value represents all objects
in the index.

ESTIMATE_SEARCH
Return value
ESTIMATE_SEARCH returns one of the following:
• The exact number of matches that satisfy the SEARCH condition if the user running the method is
a superuser or there are more than 25 matches.
• The number 25 if there are 0-25 matches and the user running the method is not a superuser.
• The number -1 if an error occurs during execution.
Errors are logged in the session log file.
Permissions
Any user can execute this method. However, the return value is affected by the user’s privilege level.
Refer to the General Notes for an explanation.
To execute this method, the server.ini key, use_estimate_search, must be set to TRUE. The key
defaults to TRUE.
Description
ESTIMATE_SEARCH is a useful tool for fine-tuning a SEARCH condition in a SELECT statement.
ESTIMATE_SEARCH provides an estimate of the number of results a query will return. Use it to
determine how selective or unselective a particular query is. Do not use it to determine the exact
number of results a particular query will return. It is intended only as a way to tune queries. Factors
such as security affect the actual number of results returned when the actual query is run.
If the user executing the method is a superuser, the method returns the exact number of matches
regardless of how few or how many matches are returned.
If the user is not a superuser, the method returns the exact number of matches if the number is greater
than 25. If the number of matches is 0-25, the method always returns the number 25.
None
Examples
EXECUTE estimate_search WITH name='filestore2_indx',
type='dm_document',query='Competitor Evaluation'

EXEC_SQL
EXEC_SQL
Purpose
Executes SQL statements.
Syntax
EXECUTE exec_sql WITH query='sql_query'
Arguments
Table 47. EXEC_SQL arguments
query string sql_query Defines the query that you want
to execute.
Return value
EXEC_SQL returns a collection that contains one query result object. The object has one Boolean
property whose value is TRUE if the query succeeded and FALSE if it was unsuccessful.
Permissions
You must have superuser privileges to use this method.
Description
EXEC_SQL executes any SQL statement with the exception of SQL SELECT statements.
If you use an IDfSession.apply method to execute the method and the query contains commas, you
must enclose the entire query in single quotes.
In the EXECUTE statement, character string literals must always be single-quoted:
EXECUTE exec_sql
with query='create table mytable (name char(32), address char(64))'
DO_METHOD, page 192

EXEC_SQL
Examples
Refer to the Description.

EXPORT_TICKET_KEY
EXPORT_TICKET_KEY
Purpose
Returns a login ticket key.
Syntax
EXECUTE export_ticket_key WITH PASSWORD='password'
Arguments
Table 48. EXPORT_TICKET_KEY arguments
password string password User-defined password used to
encrypt the returned login ticket
key.
Return value
If successful, the method returns the login ticket key used by the repository as an encrypted and
ASCII-encoded string. The returned key is encrypted using the password provided as an argument
and then encoded as an ASCII string.
If the method fails, the method returns NULL.
Permissions
You must have Superuser privileges to execute this method.
Description
Use EXPORT_TICKET_KEY when you want to copy a login ticket key from one repository to
another, to configure a trust relationship between the two repositories. (For a description of trusted
repositories, refer to EMC Documentum Content Server Fundamentals Guide.)
Related methods
IMPORT_TICKET_KEY, page 230
RESET_TICKET_KEY, page 305
IDfSession.exportTicketKey()

EXPORT_TICKET_KEY
Examples
EXECUTE export_ticket_key WITH PASSWORD='myword'

FINISH_INDEX_MOVES
FINISH_INDEX_MOVES
Purpose
Completes all unfinished object type index moves.
Syntax
EXECUTE finish_index_moves
Arguments
FINISH_INDEX_MOVES has no arguments.
Return value
FINISH_INDEX_MOVES returns a collection with one query result object. The object has one Boolean
Permissions
Description
FINISH_INDEX_MOVES completes an interrupted move operation for an object type index.
Moving an object type index is not an atomic operation. Consequently, if a move operation is
interrupted, the repository may be left with a dmi_index object that has no associated index. To
resolve this situation, use FINISH_INDEX_MOVES. This method scans the dmi_index table and
completes all unfinished index moves.
DROP_INDEX, page 199
Examples
Refer to the syntax description.

FIX_LINK_CNT
FIX_LINK_CNT
Purpose
Updates the r_link_cnt property for a folder object.
Syntax
EXECUTE fix_link_cnt FOR folder_object_id
Arguments
None
Return Value
FIX_LINK_CNT returns T (TRUE) if successful or F (FALSE) if unsuccessful.
Permissions
Executing this method requires either Superuser privileges or Write permission on the specified
folder.
Description
Use this method if the value of a folder’s r_link_cnt property is incorrect. The method determines
how many documents are linked to the specified folder and then sets that value in the folder’s
r_link_cnt property.
Examples
EXECUTE fix_link_cnt FOR 0b0000023ca4574f1

GENERATE_PARTITION_SCHEME_SQL — Deprecated
GENERATE_PARTITION_SCHEME_SQL —
Deprecated
Purpose
(This method is deprecated, beginning in release 6.6. See PARTITION_OPERATION, page 277, for its
replacement.) Creates an SQL script to control repository partitioning.
Syntax
To generate a script to partition a database:
EXECUTE generate_partition_scheme_sql WITH [operation='db_partition',]
[type_name='type_name',|table_name='regtable_name',[owner_name='owner_name',]]
[last_partition ='partition_name',last_tablespace='tablespace',]
partition_name='partition_name',range=integer,tablespace='tablespace'
{,partition_name='partition_name',range=integer,tablespace='tablespace'}
[,include_object_type={TRUE|FALSE}]
To generate a script to add partition(s) to a database:
EXECUTE generate_partition_scheme_sql WITH operation='add_partition',
[type_name='type_name',|table_name=’regtable_name',[owner_name='owner_name',]]
partition_name='partition_name',range=integer,tablespace='tablespace'
{,partition_name='partition_name',range=integer,tablespace='tablespace'}
[,include_object_type={TRUE|FALSE}]
To generate a script to exchange a partition:
EXECUTE generate_partition_scheme_sql WITH operation='exchange_partition',[temp_
table_suffix='temp_table_suffix'],
type_name='type_name',
partition_name='partition_name'
,include_object_type={TRUE|FALSE}
EXECUTE generate_partition_scheme_sql WITH operation='exchange_partition',[temp_
table_suffix='temp_table_suffix'],
table_name=’regtable_name',[owner_name='owner_name',]
partition_name='partition_name'
,include_object_type={TRUE|FALSE}
Arguments
Table 49. GENERATE_PARTITION_SCHEME_SQL arguments
operation string db_partition,
add_partition,
or exchange_
partition
Defines the operation that
you want to execute. The
only values allowed are:
db_partition, add_partition, and
exchange_partition. Db_partition
creates a script to partition the
database, add_partition creates
a script to add a partition to the

database, and exchange_partition
creates a script to exchange a
partition.
type_name string type_name Specifies a type to partition. Must
be a supertype.
table_name string regtable_name Specifies a registered table. The
table must already have an
i_partition column. Do not use
this for repository-created tables;
use the type_name argument for
those tables.
owner_name string owner_name The table owner name.
last_partition string partition_name The name of the partition for
objects whose i_partition value
is larger than the highest range
defined.
last_tablespace string tablespace The tablespace of the last partition.
partition_name string partition_name The name of the partition.
range integer integer Uppermost i_partition value of an
object placed in this partition.
tablespace string tablespace The tablespace for the partition.
temp_table_suffix string temp_table_suffix The suffix for the temporary table.
The default is "x”.
include_object_type Boolean TRUE | FALSE Whether to include the internal
table dmi_object_type in
partitioning. The table can
only be included if it has already
been partitioned.
Return Value
GENERATE_PARTITION_SCHEME_SQL returns an object id for the text file containing the SQL
script to generate the partitioning.
Permissions
In Oracle installations, you must be logged in as sysdba to run the script.

Description
Note: This method is deprecated, beginning in release 6.6. This method is replaced by
PARTITION_OPERATION.
Use this method to generate a database partitioning script. After the script is generated, it is run
against the underlying database, using the database’s SQL facility. For all versions of this method,
except for add_partition, stop Content Server before you run the script against the database, then
restart Content Server for the changes to take effect. For the add_partition method, you do not
need to stop or restart Content Server.
The first form of the command, in which operation=’db_partition’, allows you to specify a type or
a table to partition. If you do not specify a type or table, the generated script will partition all the
partitionable types in the repository if it is run. You would use this command to partition a repository
that was upgraded from an earlier version. The first range value means that objects with i_partition
values from 0 to the first value will be in the first partition. The second partition will contain objects
with i_partition values greater than first range value up to the second range value. Each subsequent
partition will contain all the objects with i_partition values greater than the previous partition and
up to its value. The last partition contains those objects with i_partition values greater than any
previously specified partition.
The second form of the command, in which operation=’add_partition’, allows you to add partitions.
This form is similar to the first form, but the first added partition begins with values greater than
previously defined partitions. If there happen to be objects in the last partition whose i_partition
values fall into the new partition’s range, they will be moved into the new partition.
If you create a new repository with partitioning enabled, there is only one partition, called the last
partition. You can then customize your repository by adding partitions. In this case, the first added
partition range goes from 0 to the value you specified.
The final two versions, in which operation=’exchange_partition’, allow you to exchange a partition
with a schema-identical offline table in the database tablespace. Commonly, you would use this
feature to load a large number of objects into a table and then swap the table into the partition.
Oracle installations
In order to use this method with an Oracle installation, you must run the script as SYSDBA. For
example, if dmadmin is the installation owner:
C:Documents and Settingsdmadmin>sqlplus "/as sysdba"
@ C:Documentumdatatestenvcontent_storage_010000005780000119.txt
Additionally, if you are partitioning a non-partitioned database, you may want to increase the number
of open database cursors, or the script may exit with the error:
ORA-0100 Max Opened Cursors Exceeded
To increase the number of cursors, consult your Oracle documentation. It may tell you to use a
command similar to:
ALTER SYSTEM SET OPEN_CURSORS=2000 SID='*' SCOPE=MEMORY;
or for Oracle versions earlier than Oracle 11:
ALTER SYSTEM SET OPEN_CURSORS=2000 SCOPE=MEMORY SID='*';
to alter the number of open cursors.

If you exit the script with an error (from inadvertently exceeding the number of open cursors, for
example), correct the error, and rerun the script, you may see an error message like:
ORA-12091: cannot online redefine table "TECHPUBS"."DMC_WFSD_ELEMENT_S" with
materialized views
caused by leftover temporary items from the previous script failure. One way to correct this error is
to run a command similar to:
execute DBMS_REDEFINITION.ABORT_REDEF_TABLE('<Schema Name>', '<Table Name>'
,'<Table Name>I');
Where <Table Name>I is the intermediate table name used for the redefinition. From the previous
error message, we would use this command:
execute DBMS_REDEFINITION.ABORT_REDEF_TABLE('TECHPUBS', 'DMC_WFSD_ELEMENT_S'
,'DMC_WFSD_ELEMENT_SI');
SQL Server installations
If you are using a SQL Server installation, use filegroup names as values for the tablespace attributes,
since SQL Server filegroups correspond to tablespaces. For example, use:
tablespace='filegroup'
and
last_tablespace='filegroup'
for SQL Server installations.
Partition Exchange
Partition exchange must be carefully planned. Typically, you will create a number of offline tables to
load with data, use whatever native database method is available to load the tables, create the table
indexes, and then swap the tables into the prepared repository partition. This technique can load
large amounts of data into a repository while causing a minimum of disruption to normal use of the
repository. This technique can also be used to remove large amounts of data from a repository by
swapping out a partition for small offline tables.
The typical steps you would take to do a partition exchange involve the following:
1. Identify the offline tables to create.
See EMC Documentum High-Volume Server Development Guide for an example of the tables to create.
2. Load the offline tables.
Load the tables with data using whatever methods are available to you with your database.
3. Create the offline index tables.
The offline tables must index the same properties as the online objects do. The schema must be
identical. Create the offline index tables in the same tablespace as the current online indexes.
4. Exchange the partition for the offline tables.
Run the GENERATE_PARTITION_SCHEME_SQL administration method to generate the
partitioning script, stop Content Server, run the script against the RDBMS, and restart Content
Server. After following these steps, the data that was previously in the offline tables is in the
online partition, and the previously online data is now in the offline table.
5. Validate the exchange.

Check the online data to be sure that the exchange was successful.
Examples
This example generates a script to partition all the partitionable types in a repository. After the script
is run, partitionable objects with an i_partition value of 0 to 10 will be stored in partition P1 of the
database. Partitionable objects with an i_partition value of 11 to 20 will be stored in partition P2
of the database.
EXECUTE generate_partition_scheme_sql WITH
"partition_name"='P1',"range"=10,"tablespace"='dm_techpubs_docbase',
"partition_name"='P2',"range"=20,
"tablespace"='dm_techpubs_docbase'
To get the script, you can issue the following DQL statement (assume that the result of the earlier
method returned the object ID 0855706080007c19):
EXECUTE get_file_url FOR '0855706080007c19' WITH "format"='text'

GET_FILE_URL
GET_FILE_URL
Purpose
Returns the URL to a particular content file.
Syntax
EXECUTE get_file_url FOR 'object_id'
WITH format='format_name'[,page=page_number,]
[page_modifier='value']
object_id is the object ID of the document that contains the content file.
Arguments
Table 50. GET_FILE_URL arguments
format string format_name Name of the content format, as
specified in the format object.
category integer value The kind of items to retrieve. Valid
values are:
1, for workflow tasks
2, for router tasks (obsolete)
4, for notifications
8, for completed router tasks
To retrieve multiple kinds of
items, use the sum of the integers
representing the items. For
example, to retrieve workflow
tasks and notifications, specify the
value as 5.
The default is 3, workflow tasks
and router tasks.
page_modifier S value Identifies the rendition. Refer to
the General Notes for a detailed
description of the purpose of the
page modifier.

GET_FILE_URL
Return value
GET_FILE_URL returns a collection consisting of one object with five properties. One property
indicates the success or failure of the method call. If the call is successful, the other property values
can be concatenated to compose the URL.
The properties are:
result Boolean property whose value indicates whether the method was
successful. T (TRUE) indicates successful completion and F (FALSE)
indicates failure.
base_url The value in the base_url property is the base URL used by the Web
server to retrieve the content. This value is retrieved from the base_url
property of the storage area that contains the file.
store The value in the store property is the name of the storage area that
contains the file.
path The value in the path property is a path to the file. The path is relative
to the storage area identified in the store property.
ticket Contains an encryption of the path value plus a time stamp.
Permissions
You must have at least Read permission on the object or Sysadmin privileges to use this method.
Description
Use GET_FILE_URL in an application when you want to access content using a Web server or a
streaming server.
None
Examples
EXECUTE get_file_url FOR '090000215400ac12'
WITH format='jpeg_th',page=0

GET_INBOX
GET_INBOX
Purpose
Returns items from an Inbox.
Syntax
EXECUTE get_inbox [WITH name='user_name'][,category=value]
[,batch=value]{,order_by='attr_name [asc|desc]'}]
Arguments
Table 51. GET_INBOX arguments
name string user_name User whose Inbox items you want
to retrieve. The default value is
the session user.
category integer value The kind of items to retrieve. Valid
values are:
1, for workflow tasks
2, for router tasks (obsolete)
4, for notifications
8, for completed router tasks
To retrieve multiple kinds of
items, use the sum of the integers
representing the items. For
example, to retrieve workflow
tasks and notifications, specify the
value as 5.
The default is 3, workflow tasks
and router tasks.

GET_INBOX
batch integer value Number of items returned by each
IDfCollection.next method issued
against the collection returned by
GET_INBOX.
The default value is 0, meaning
return all rows.
order_by string attr_name Property by which to order the
returned items.
The property must be a property
of the dmi_queue_item object
type.
Including asc sorts the returned
items in ascending order. desc
sorts the items in descending
order. The default is ascending
order.
The default ordering is by
r_object_id
Return value
GET_INBOX returns a collection containing the Inbox items in query result objects. The query result
objects have 49 properties. The first 41 are the properties of the dmi_queue_item object representing
the Inbox item. The property names in the query result object match the names of the properties of
the queue items. The remaining 8 properties identify the SysObject associated with the Inbox item, if
any. Generally, these properties have values if the Inbox item is a workflow task. Notification items
have no values in these 8 properties. (Refer to the General Notes for a more detailed explanation.)
Table 52, page 217, lists the SysObject-related properties and their corresponding SysObject property.
Table 52. SysObject-related property names for GET_INBOX query result objects
Query Result Property Name Associated SysObject Property
pkg_object_id r_object_id
pkg_object_name object_name
pkg_object_type r_object_type
pkg_content_type a_content_type
pkg_application_type a_application_type
pkg_link_cnt r_link_cnt
pkg_lock_owner r_lock_owner
pkg_is_virtual_doc r_is_virtual_doc

GET_INBOX
Permissions
Any user can issue this method to return either his or her own Inbox items or the Inbox items
of another user.
Description
With one exception, you can specify the arguments in any order. The exception is ORDER_BY. This
argument must appear last.
Generally, GET_INBOX returns one query result object for each item in user’s Inbox. However, if a
particular task has multiple objects associated with it, the method returns one query result object for
each object associated with the task. The queue item property values for the multiple objects will be
the same. Only the values of the eight SysObject-related properties will be different.
The eight SysObject-related properties contain null values in the following cases:
• The Inbox item is an event notification and therefore has no associated object.
• The Inbox item is a task, but its associated object has been deleted from the repository.
• The Inbox item is a task but the user who issues the method doesn’t have at least Browse
permission on the associated object.
• The Inbox item is a workflow task, such as a Beginning task, that has no package attached to it.
• The Inbox item represents an object in a remote repository.
None
Examples
This example returns all tasks and notifications for the user issuing the method:
EXECUTE get_inbox WITH category=7

GET_LAST_SQL
GET_LAST_SQL
Purpose
Retrieves the last SQL statement that was executed by the session in which this RPC is run.
Syntax
EXECUTE get_last_sql [WITH detail_info=true|false]
Arguments
Table 53. GET_LAST_SQL arguments
detail_info boolean TRUE | FALSE Whether to get all the recorded
SQL statements and weather to
enclose each statement within a
pair of XML tags showing the
index number.
When the detail_info
argument is not introduced,
GET_LAST_SQL only returns
the last SQL statement; when
you set detail_info to TRUE,
GET_LAST_SQL not only returns
all the recorded SQL statements,
but also enclose each statement
within a pair of XML tags showing
the current index number; for
example, the first statement is
enclosed within <0> and </0>.
Return value
GET_LAST_SQL returns a collection with one query result object by default. The result object has one
property whose value is the last SQL statement. To see the statement, issue a Next on the collection
and then dump the collection.
When the detail_info argument is not introduced, GET_LAST_SQL only returns the last SQL
statement; when you set detail_info to TRUE, GET_LAST_SQL not only returns all the recorded SQL
statements, but also enclose each statement within a pair of XML tags showing the current index
number; for example, the first statement is enclosed within <0> and </0>.
If the last SQL tracing option is turned off, this method returns the following error message:
No SQL Statement is available because Last SQL Trace is disabled.

GET_LAST_SQL
Permissions
Any one can issue this method.
Description
The last SQL tracing feature is turned on by default when a server starts. If the feature is turned
off, you can turn it on using the last_sql_trace option of the SET_OPTIONS method. (Refer to
SET_OPTIONS, page 317, for instructions.)
None
Examples
EXECUTE get_last_sql

GET_PATH
GET_PATH
Purpose
Returns the directory location of a content file stored in a distributed storage area.
Syntax
EXECUTE get_path [FOR] 'content_obj_id'
[WITH store = 'value']
Arguments
Table 54. GET_PATH arguments
store string storage_
component_name
Specifies a storage area that
contains the file whose full path
you want to determine. This is
an optional argument. Use the
storage area’s name as defined in
its storage object.
Return value
Returns the directory path of the specified content file.
Permissions
Description
In a content server configuration, the GET_PATH function is executed by the content server.
If you do not include the STORE argument, the method looks in the local component of the
distributed storage area. If the file isn’t found in the local area, the method attempts to create a replica
of the file in the local area and returns the path of the local replica.
None

GET_PATH
Examples
The following examples return the file path for the content file represented by the content object
whose object ID is 060000027435127c in the storage1 storage area:
EXECUTE get_path FOR '060000027435127c'
WITH store='storage1'

GET_SESSION_DD_LOCALE
Purpose
Returns the locale in use for the current session.
Syntax
EXECUTE get_session_dd_locale
Arguments
None
Return value
The method returns a collection with one result object. The result object has one property, named
dd_locale. The value of this property is the locale for the session.
Permissions
Description
None
None
Examples
EXECUTE get_session_dd_locale

HTTP_POST
HTTP_POST
Purpose
Sends an HTTP_POST request invoking a Java servlet.
Syntax
EXECUTE http_post WITH app_server_name='name'
[,arguments=argument_list][,save_response=value]
[,time_out=value][,launch_asynch=value][,trace_launch=value]
Arguments
Table 55. HTTP_POST arguments
app_server_name string name Name of the Java servlet.
This must match a servlet
identified in the app_server_name
property of the server config
object.
arguments string argument list Defines the command line
arguments passed to the servlet or
Java method.
UTF-8 characters are acceptable
for the argument strings.
save_response integer integer Indicates whether to save the
response in a document in the
repository. Value values are:
0, do not save
1, save the results
-1, save the results if there is an
error
The default is 0, do not save.
time_out integer integer The length of time, in seconds,
to wait for a response to the
HTTP_POST request.
The default is 60 seconds.

HTTP_POST
launch_async Boolean true | false Indicates whether to execute
the HTTP_POST request
asynchronously. TRUE executes
the HTTP_POST asynchronously.
The default is FALSE.
This argument is ignored if
SAVE_RESPONSE is TRUE.
trace_launch Boolean true | false Indicates whether to generate
tracing information for the
HTTP_POST request. If
TRACE_LAUNCH is TRUE,
the information is stored in the
server log. The default is FALSE.
Return value
HTTP_POST returns a collection with one query result object that has seven properties. Table 56,
page 225, lists the properties:
Table 56. Query result object properties for HTTP_POST
Property Name Description
result Indicates the status of the HTTP_POST. Possible values are:
0, indicating the status HTTP/1.1 2xx
1, indicating the status HTTP/1.1 5xx
-1, indicating any other status
http_response_status The response returned for the HTTP_POST.
Some example values are:
HTTP/1.1 2xx OK
HTTP/1.1 5xx Internal Server Error
For a complete list of responses, refer to the HTTP protocol
specification.
request_failed Indicates whether the method sent an HTTP request to the
application server. T (TRUE) means the method failed to
send a request. F (FALSE) means the request was successfully
sent to the application server.
response_doc_id Object ID of the document in which the response is stored.
This only has a value if SAVE_RESPONSE was set to TRUE.

HTTP_POST
Property Name Description
time_taken Length of time, in seconds, between when the request was
sent and when a response was received. This represents the
execution time of the Java method plus the time used for
communication between Content Server and the application
server.
timed_out Indicates whether the method timed out. T (TRUE) means
that the HTTP_POST method timed out while waiting for a
response. F (FALSE) means that a response was received.
time_out_length Time, in seconds, of the time out period.
Permissions
You must have Superuser privileges to execute this method. (Refer to Preserving security, page 226,
for more information about security when using this method.)
Description
Use HTTP_POST to invoke a servlet or Java method installed in an application server.
If you set LAUNCH_ASYNC to TRUE, Content Server closes the connection to the application server
immediately after sending the request. If LAUNCH_ASYNC is FALSE, the server waits for a response
until a response is received or the time out period is reached.
Setting TRACE_LAUNCH to TRUE logs tracing information about the invocation of the HTTP_POST
method and its success or failure.
The ARGUMENTS argument can contain only UTF-8 characters. Content Server encodes the
arguments using application/x-www-form-urlencoded format. For example, suppose you issued
the following HTTP_POST method:
EXECUTE http_post WITH app_server_name='payroll',
save_response=true,trace_launch=true,
arguments='-docbase accounting -user paymaster
-ticket DM_TICKET=0000000222a02054.accounting@host01
-document "weekly record"'
Content Server sends the arguments in the following format:
docbase=accounting&user=paymaster&ticket=DM_TIICKET%
3D0000000222a02054.accounting%4Dhost01&document=weekly+record
Preserving security
The application server, all servlets that it invokes, and the methods invoked by the servlets run as the
Content Server installation owner, which is an account with Superuser privileges. Consequently, it is
important that they are written and execute using adequate security precautions.
There are two primary security issues:
• Determining origin of HTTP_POST requests
• Login without passwords (this is only possible on Windows platforms)

HTTP_POST
The application server or the servlet has no inherent way to know whether the HTTP_POST request
was sent from a Documentum Server. When you write a servlet, you must include security checking
to ensure that the request comes from a Documentum Server. One recommended way is have the
servlet ensure that the generated request comes from a machine that hosts a Content Server by
checking the IP address of the sender against a list of valid IP addresses. (This is how the do_method
servlet checks the origin. Refer to Installing Content Server for information about how that is set up.)
On Windows platforms, the current operating system user is allowed to log in to the repository
through Content Server without providing a password. Consequently, a servlet or an invoked Java
method can log into the repository with Superuser privileges without providing a password.
If you write a servlet or method that uses this manner of login, you may want to ensure that the actual
user who issues the HTTP_POST to invoke the program has the appropriate privileges to execute the
program as a superuser. To do this, send the current user’s name and a login ticket to the method
in the arguments. The method can use these to log in and check the privileges of the user before
connecting as the current operating system user.
Sample servlet and method
Documentum provides a sample servlet for handling HTTP_POST method calls and a sample method.
The servlet is named DmSampleServlet.java and the method is named DmSampleMethod.java. The
sample servlet and method are located in the classes folder under the WEB-INF folder. The WEB-INF
folder was set up when you installed the application server.
Recording output
If you set SAVE_RESPONSE to TRUE, then anything that the invoked servlet writes to
HttpServerletResponse.OutputStream is saved to a document in the repository.
DO_METHOD, page 192
Examples
The following examples send an HTTP_POST request to the Java servlet called DevAppSrv. The
content of the request passes a repository name, user name, a login ticket, and a document name to
the Java servlet.
EXECUTE http_post WITH app_server_name='DevAppSrv',
save_response=1,trace_launch=T,time_out=60,
arguments='-docbase DevTest -user test_user
-ticket DM_TICKET=0000000221c02052.DevTest@dev012
-document "A Test"'

IMPORT_REPLICA
IMPORT_REPLICA
Purpose
Imports files from one distributed storage area into another distributed storage area.
Syntax
EXECUTE import_replica FOR 'content_object_id'
WITH store='storage_name',file='file_name'
[,other_file='other_file_name']
Arguments
Table 57. IMPORT_REPLICA arguments
store string storage_name Identifies the storage area in which
to place the imported content file.
This must be a component of a
distributed storage area. Use the
storage area’s name.
file string file_name Identifies the file to replicate.
other_file string other_file_name The OTHER_FILE argument is
optional. It directs the server
to copy the specified Macintosh
resource fork file in addition to the
data file. Specify the name of the
resource fork file.
Use this only when the file you
are importing was created on a
Macintosh.
Return value
IMPORT_REPLICA returns a collection with one query result object. The object has one property
whose value indicates success (TRUE) or failure (FALSE).
Permissions

IMPORT_REPLICA
Description
IMPORT_REPLICA imports files from one distributed storage area into another distributed storage
area. The files are considered replicas in the target storage area.
To use IMPORT_REPLICA, you must be using one server for both data and content requests. If the
content server to use IMPORT_REPLICA in the session.
CAN_FETCH, page 173
DELETE_REPLICA , page 188
GET_PATH, page 221
REPLICATE, page 303
Examples
The following examples import the file mydoc into the storage area named distcomp_2:
EXECUTE import_replica FOR '06000001402371e1'
WITH store='distcomp_2',FILE='mydoc'

IMPORT_TICKET_KEY
IMPORT_TICKET_KEY
Purpose
Imports a login ticket key into a repository.
Syntax
EXECUTE import_ticket_key
WITH KEY_STRING='string',
PASSWORD='password'
Arguments
Table 58. IMPORT_TICKET_KEY arguments
key_string string string The ASCII-encoded string
returned by an EXPORT_
TICKET_KEY method.
password string password Password used when the key was
exported.
Return value
This method returns a collection with one query result object. The object has one property, named
result, that contains T (TRUE) if the method was successful or F (FALSE) if unsuccessful.
Permissions
Description
Use IMPORT_TICKET_KEY to import a login ticket key into a repository. The key must have been
exported from a repository using the EXPORT_TICKET_KEY method. When you import the key,
the password argument in IMPORT_TICKET_KEY must be the same password that you used in
the EXPORT_TICKET_KEY method.
You must restart Content Server after importing a login ticket key to make the key take effect.
Keys are typically exported from one respository and imported into another as part of the
configuration process when setting up trusted respositories, to allow use of global login tickets or

IMPORT_TICKET_KEY
tokens. (For a description of trusted repositories, global login tickets, and global tokens, refer to EMC
Documentum Content Server Fundamentals Guide.)
EXPORT_TICKET_KEY, page 205
RESET_TICKET_KEY, page 305
IDfSession.importTicketKey()
Example
EXECUTE import_ticket_key
WITH KEY_STRING='ticket_string',
PASSWORD='myword'

LIST_AUTH_PLUGINS
LIST_AUTH_PLUGINS
Purpose
Lists the authentication plugins that the server has loaded.
Syntax
EXECUTE list_auth_plugins
Arguments
None
Return value
The method returns a collection with one query result object. The object has two repeating string
properties, plugin_id and plugin_filepath. The values in plugin_id are the unique plugin identifiers
and the values in plugin_filepath are the full paths of the plugins. The values at corresponding
index positions represent one plugin.
Permissions
You must have Sysadmin or Superuser privileges to execute this method.
Description
This method is useful only at sites that have a Trusted Content Services license because the ability to
use authentication plugins is a feature of Trusted Content Services.
None
Examples
Refer to the syntax.

LIST_RESOURCES
LIST_RESOURCES
Purpose
Lists a variety of information about the server’s operating system environment and the server.
Syntax
EXECUTE list_resources [WITH reset=true|false]
Arguments
Table 59. LIST_RESOURCES arguments
RESET B T (TRUE) or F
(FALSE)
TRUE reinitializes the file handle
and heap size counters. FALSE
keeps the present values of the
counters.
Return value
LIST_RESOURCES returns a collection with one query result object. On Windows platforms, the
query result object has twelve properties, described in Table 60, page 233.
Table 60. Collection properties for LIST_RESOURCES
file_handles_in_use integer The number of file handles in use by
the main thread and all session threads
(Windows) or the current child process
(UNIX).
file_handles_max integer The configured limit at the
operating-system level on the number of
file handles the process can open.
file_handles_new integer A counter that indicates how many
file handles have been created or
destroyed since the last LIST_RESOURCES
with RESET = T. If the number is
negative, it means that there are fewer
handles open than there were at the
last LIST_RESOURCES call. (Issuing
LIST_RESOURCES with RESET=T
reinitializes file_handles_new to zero.)

LIST_RESOURCES
session_heap_size_max integer Maximum size, in bytes, of a thread’s
session heap. When a session is started, its
maximum heap size corresponds to this
value. A value of 0 indicates that the size of
the session heap will be unconstrained (the
heap will grow to whatever size the server
machine resources will allow).
current_heap_size_max integer Maximum size of the thread’s session
heap. This reflects the value that was in
session_heap_size_max when the session
was started, and is the size of the heap
available to the session.
session_heap_size_in_use integer How much, in bytes, of the currently
allocated heap (virtual memory) is in use
by the session.
session_heap_size_new integer A count of the bytes that the heap
has grown or shrunk since the last
LIST_RESOURCES call. (Issuing
reinitializes session_heap_size_new to
zero.)
root_heap_size_in_use integer How much, in bytes, of the main server
thread’s heap is in use.
root_heap_size_new integer A count of the bytes that the heap has grown
or shrunk since the last LIST_RESOURCES
call.
Issuing LIST_RESOURCES with RESET=T
reinitializes session_heap_size_new to zero.
max_processes integer The maximum number of processes that
can be created by the account under which
the server is running.
server_init_file string(255) The full path to the server’s server.ini file.
initial_working_directory string(255) The full path to the directory containing the
server executable.
On UNIX, the result object has eight properties, described in Table 61, page 234.
Table 61. Collection properties for LIST_RESOURCES
file_handles_in_use integer The number of file handles in use by
the main thread and all session threads
(Windows) or the current child process
(UNIX).

LIST_RESOURCES
file_handles_max integer The configured limit at the
operating-system level on the number of
file handles the process can open.
file_handles_new integer A counter that indicates how many
file handles have been created or
destroyed since the last LIST_RESOURCES
with RESET = T. If the number is
negative, it means that there are fewer
handles open than there were at the
last LIST_RESOURCES call. (Issuing
reinitializes file_handles_new to zero.)
heap_size_in_use integer The size, in bytes, of the session heap.
heap_size_new integer A count of the bytes that the heap has grown
or shrunk since the last LIST_RESOURCES
call.
Issuing LIST_RESOURCES with RESET=T
reinitializes heap_size_new to zero.
max_processes integer The maximum number of processes that
can be created by the account under which
the server is running.
server_init_file string(255) The full path to the server’s server.ini file.
initial_working_directory string(255) The full path to the directory containing the
server executable.
Permissions
Description
LIST_RESOURCES lists a variety of information about the server operating system environment and
the server (the information is described in Return value, page 233).
DB_STATS, page 186
LIST_TARGETS, page 240
SHOW_SESSIONS, page 322

LIST_RESOURCES
Examples
The following examples return the resources information and reset the heap size counters:
EXECUTE list_resources WITH reset=true

LIST_SESSIONS
LIST_SESSIONS
Purpose
Returns information about all the currently active repository sessions.
Syntax
EXECUTE list_sessions[WITH brief_info=true|false]
Arguments
Table 62. LIST_SESSIONS arguments
BRIEF_INFO B T (TRUE) or F
(FALSE)
Although the method will accept
this argument, setting the value
has no effect. Regardless of the
value set, the properties listed in
the complete information table are
returned. The default is FALSE.
Return value
LIST_SESSIONS returns a collection with one result object that describes the current active session.
If BRIEF_INFO is FALSE
Table 63, page 237 lists the information returned for either value of BRIEF_INFO.
Table 63. Complete information returned by LIST_SESSIONS
Property Description
root_start The time when the main server thread (root
server process) was started.
root_pid On Windows, the process ID of the Content
Server process.
On UNIX, the process ID of the root Content
Server process.
shared_mem_id The ID of the shared memory segment used by
the servers.
This property is returned only on UNIX
platforms.

LIST_SESSIONS
semaphore_id The ID of the semaphore used by the servers.
This property is returned only on UNIX
platforms.
session The object ID of the session begun by user_name.
db_session_id The ID of the database session.
typelockdb_session_id The database session ID for type locking
tempdb_session_ids List of temporary database sessions
pid The ID of the session thread (process).
user_name The user name of the user who started the
session.
user_authentication The user authentication state for the session.
Possible values are password, ticket, trusted
client, change password, or in progress.
If the value is change password, the user logged
in for that session can only perform the change
password operation. No other operations are
allowed.
client_host The host name of the machine on which the
session was started.
client_lib_ver The version number of the DFC in use for the
session.
client_locale A verbose description of the client’s
environment, including the operating system
version, the character set in use, the language in
use, and the date format.
start The starting time of the session.
last_used The last time the client session contacted the
server.
session_status The session status. Active means that the session
is connected and has not timed out. Inactive
means that the session has timed out.
shutdown_flag Records the setting of the immediacy_level
argument in the kill method.
last_rpc The last RPC the server ran for the session

LIST_SESSIONS
current_rpc The current RPC being run by the server for the
session
Note that after the RPC is completed the server
will not clear this field. Therefore, if last_rpc
and current_rpc have the same value, either the
server has completed an RPC for the session and
has not executed another RPC, or the server is
running the same RPC again.
last_completed_rpc The last time the server completed an RPC for
the session
Permissions
Description
LIST_SESSIONS returns a collection of a single query result object whose properties contain
information about the current active session.
SHOW_SESSIONS, page 322
Examples
This example executes LIST_SESSIONS:
EXECUTE list_sessions

LIST_TARGETS
LIST_TARGETS
Purpose
Lists the connection brokers to which the server is currently projecting.
Syntax
EXECUTE list_targets
Arguments
LIST_TARGETS has no arguments.
Return value
LIST_TARGETS returns a collection with one query result object. The Table 2-9. The values across the
properties at one index position represent the information about one connection broker to which
the server is projecting.
Table 64. Query result object properties for LIST_TARGETS
projection_targets string(80) A repeating property that contains the names of
the hosts on which the target connection brokers
are running.
projection_proxval integer A repeating property that contains the proximity
value projected to the connection broker
identified in the corresponding index position in
projection_targets.
projection_notes string(80) A repeating property that contains any
user-defined note for the connection broker
defined at the corresponding index position in
projection_targets and projection_proxval.
docbroker_status string(64) A repeating property that records whether the
connection broker identified in projection_targets
at the corresponding index position is available.
Valid values are: Available and Unavailable.
comments string(64) A repeating property that records whether the
projection target entry was taken from the server
config object or the server.ini file.

LIST_TARGETS
Permissions
Description
A server’s connection broker targets are those connection brokers to which the server is sending
checkpoint messages. Target connection brokers are defined in the server’s server config object and
may also be defined in the server.ini file. This method returns a list of the targets defined in the
server config object.
None
Examples
The following examples list the current connection broker targets for the server:
EXECUTE list_targets

LOG_OFF
LOG_OFF
Purpose
Turns off the RPC logging.
Syntax
EXECUTE log_off
Arguments
LOG_OFF has no arguments.
Return value
LOG_OFF returns a collection with one query result object. The object has one Boolean property
Permissions
Description
None
LOG_ON, page 243
Examples
The following example turns off RPC logging:
EXECUTE log_off

LOG_ON
LOG_ON
Purpose
Turns on logging for RPC calls.
Syntax
EXECUTE log_on [WITH detail=true|false]
Arguments
Table 65. LOG_OFF arguments
DETAIL B T (TRUE) or F
(FALSE)
Indicates whether you want
information about the arguments
passed with the RPC call and the
results. The default is FALSE.
Return value
LOG_ON returns a collection with one query result object. The object has one Boolean property
Permissions
Description
The LOG_ON function turns on logging for RPC calls.
If you execute the LOG_ON function without including the DETAIL argument, the server logs
information about the start and stop names and time information for the operation. If you set
DETAIL to TRUE, the server also includes information about the arguments passed with each call
and the results of the call.
LOG_OFF, page 242

LOG_ON
Examples
These examples turn on detailed RPC logging:
EXECUTE log_on WITH detail=true

MAKE_INDEX
MAKE_INDEX
Purpose
Creates an index for a persistent object type.
Syntax
EXECUTE make_index
WITH type_name='object_type',attribute='property_name'
{,attribute='property_name'}
[,unique=true|false][,index_space='name'][,index_name='name']
[,use_id_col=true|false][,use_pos_col=true|false]
[,global_index=true|false]
Arguments
Table 66. MAKE_INDEX arguments
TYPE_NAME S object_type Identifies the object type for which
to create an index.
This is a required argument.
ATTRIBUTE S property_name Identifies the property or
properties on which to build the
index. You can specify multiple
properties, but you cannot mix
single-valued and repeating
properties in the same statement.
UNIQUE B TRUE or FALSE Indicates whether to create a
unique index. The default is
FALSE.
INDEX_SPACE S Name of
tablespace or
segment
Specifies the tablespace or segment
in which to store the index. If
unspecified, the default is the
tablespace or segment associated
with the repository.
INDEX_NAME S Name of index
table
Specifies the name of the index
table. If unspecified, the system
generates this name.
USE_ID_COL B TRUE or FALSE Indicates whether to include the
r_object_id column value in the
index. The default is FALSE.

MAKE_INDEX
USE_POS_COL B TRUE or FALSE Indicates whether to include the
i_position column value in the
index. The default is FALSE.
GLOBAL_INDEX B TRUE or FALSE Indicates whether the index being
created is a global index. This
parameter can only be used for
partitioned types. The default is
FALSE.
Return value
MAKE_INDEX returns a collection with one query result object. The object has one property, named
result, that contains the object ID of the dmi_index_object for the new index if successful.
If the method failed because the syntax was incorrect or the property specified did not exist, the
result property contains F (FALSE).
If the specified index already exists, the result property contains ’0000000000000000’.
Permissions
Description
MAKE_INDEX creates an index for a persistent object type. You can create an index for any
persistent object type.
You can specify one or more properties. However, if you specify multiple properties, you must
specify all single-valued properties or all repeating properties. You cannot mix single and repeating
valued properties in the same argument list. (This is because the underlying RDBMS stores an object’s
single and repeating properties in separate tables.)
If you specify multiple properties, the sort order within the index corresponds to the order in which
the properties are specified in the statement. To include the r_object_id column in the index, include
the USE_ID_COL argument.
The properties you specify must be defined for the type you specify. They cannot be inherited
properties. For example, if you want to create an index on the subject and title properties, you would
specify dm_sysobject in TYPE_NAME, as these properties are defined for dm_sysobject. (The EMC
Documentum Content Server System Object Reference manual lists the properties defined for each object
type.)
A unique index is an index in which the values in the columns to be indexed are unique within the
index. If you include the UNIQUE argument, the RDBMS server enforces the uniqueness.
GLOBAL_INDEX is a boolean parameter that indicates whether the index being created is a global
index. The default value is FALSE. This parameter can be used only when the type is a partitioned
type, and is only useful if UNIQUE=TRUE is also specified. If an index of a partitioned table is

MAKE_INDEX
a unique local index, it must include the i_partition column as part of the index. Otherwise, the
database server will return an error. This is a requirement of the database for unique local indexes in
a partitioned table. Since i_partition is an internal attribute, both single value tables and repeating
value tables have this column. The internal attribute i_partition will not be included for unique
indexes created for a partitioned type if GLOBAL_INDEX is TRUE.
Note: If you plan to use partition exchange, do not create global indexes on the types to use for the
exchange. See EMC Documentum High-Volume Server Development Guide for more details on partition
exchange.
Examples
This example creates an index for the dm_sysobject object type, indexing on the subject and title
properties:
EXECUTE make_index WITH type_name='dm_sysobject',
attribute='subject', attribute='title'
The next example creates an index for the dm_sysobject type on the property r_creator_name:
attribute='r_creator_name'
The next example creates an index for the dm_sysobject type on the properties r_creator_name
and r_creation_date:
attribute='r_creator_name',
attribute='r_creation_date'

MARK_AS_ARCHIVED
MARK_AS_ARCHIVED
Purpose
Sets the i_is_archived property of an audit trail entry to TRUE.
Syntax
EXECUTE mark_as_archived FOR audit_obj_id
audit_obj_id is the object ID of the audit trail entry. This is a dm_audittrail, dm_audittrail_acl, or
dm_audittrail_group object.
Arguments
This method has no arguments.
Return value
The method returns TRUE if successful or FALSE if unsuccessful.
Permissions
Description
Use this method to set an audit trail entry’s i_is_archived property to TRUE after you archive the entry.
PURGE_AUDIT, page 287
Examples
EXECUTE mark_as_archived FOR 5f000012ae6217ce

MARK_FOR_RETRY
MARK_FOR_RETRY
Purpose
Finds queue items representing events queued to the Index Agent that are in the acquired or failed
state and resets them to the pending state
Syntax
EXECUTE mark_for_retry
WITH NAME = 'index_name'
Arguments
Table 67. MARK_FOR_RETRY arguments
NAME S index_name Identifies the index that contains
the objects to mark for a retry.
Use the name associated with the
index’s fulltext index object.
Return value
MARK_FOR_RETRY returns a collection with one query result object. The object has one Boolean
Permissions
Description
Use MARK_FOR_RETRY in the recovery procedure if you encounter problems with the index agent.
The method scans the repository to find queue items representing events queued to the index agent
that are in the acquired or failed state and resets them to the pending state.
The index agent registers itself for certain events on all SysObjects. The events serve as notice to the
index agent that the objects need to be indexed or updated in the index. If there is a problem with the
index agent, for example, if it crashes, some of the queue items representing those events may be left
in the acquired or failed state. MARK_FOR_RETRY is used to reset those events to pending so that the
index agent will pick up those events and properly process the objects after the problem is resolved.

MARK_FOR_RETRY
None
Examples
These examples find all the content objects in the index that have an update_count value of -5 and
marks them for a retry:
EXECUTE mark_for_retry
WITH NAME = 'storage1_index'

MIGRATE_CONTENT
MIGRATE_CONTENT
Purpose
Moves content files from one storage area to another.
Syntax
To migrate a single object:
EXECUTE migrate_content [FOR]object_id
WITH target_store='target_storage_name'
[,renditions=value][,remove_original=TRUE|FALSE][,log_file='log_file_path']
[,log_level=value][,source_direct_access_ok=T[,direct_copy=T
[,update_only=T,command_file_name='command_file_name']]
|,source_direct_access_ok=T[,direct_move=T
]
To migrate all objects in a particular storage area:
EXECUTE migrate_content WITH source_store='source_storage_name',
target_store='target_storage_name',log_file='log_file_path'
[,log_level=value][,max_migrate_count=value][,batch_size=value]
[,remove_original=TRUE|FALSE][,parallel_degree=value]
[,source_direct_access_ok=T[,direct_copy=T
]
To migrate a set of objects identified by a DQL query:
EXECUTE migrate_content WITH target_store='target_storage_
name',query='DQL_predicate'[,sysobject_query=T [,type_to_query=
'type_name']],log_file='log_file_path'
[,log_level=value][,renditions=value][,max_migrate_count=integer][,batch_size=value]
[,remove_original=TRUE|FALSE][,parallel_degree=value]
[,source_direct_access_ok=T[,direct_copy=T
]
Arguments
Table 68. MIGRATE_CONTENT arguments
object_id ID content_object ID
or SysObject ID
When migrating a single content
file, you can identify the file using
its content object ID or the object
ID of the SysObject (or SysObject
subtype) that contains the content
file.

MIGRATE_CONTENT
Specifying a content object ID
requires Superuser privileges.
Specifying a SysObject object ID
requires Write permission on the
object.
SOURCE_STORE S source_storage_
name
Name of the storage area whose
content you wish to migrate. The
storage area may be a file store, a
ca store, a blobstore, a distributed
store, or an external store. Use the
name of the storage area’s object.
TARGET_STORE S target_storage_
name
Name of the storage area to which
you are migrating the content. The
storage area may be a file store,
a ca store, or a distributed store.
Use the name of the storage area’s
object.
(Blobstores and external stores
cannot be specified as a target
storage area, and ca stores cannot
be specified if DIRECT_MOVE is
specified when migrating from an
external store.)
RENDITIONS S all, primary, or
secondary
Indicates which content files
associated with the SysObject or
SysObjects you wish to move.
This argument is used only when
object_id references a SysObject or
when SYSOBJECT_QUERY is set
to T. Valid values are:
all, meaning move the first
primary content file (page 0) and
its renditions
primary, meaning move only the
first primary content file (page 0)
secondary, meaning move only
the renditions of the first primary
file (page 0)
The default is primary.

MIGRATE_CONTENT
REMOVE_ORIGINAL B T (TRUE) or F
(FALSE)
Whether to remove the content file
from the source storage area. T
directs Content Server to remove
the original content file if possible.
F directs Content Server not to
remove the content.
If set to F, you must include the
LOG_FILE argument also.
LOG_FILE S log_file_path Identifies where to log messages
generated by the method.
You must include this argument if
you are moving all content from
one storage area to another or if
you are using a DQL predicate to
select content for migration.
LOG_LEVEL I value Level of detail in the log file.
Valid values are:
• 1: full tracing (recommended).
• 2: print out warning messages
and above.
• 3: print out only error messages.
The default is 1.
MAX_MIGRATE_
COUNT
I value Defines the maximum number of
content files to migrate.
BATCH_SIZE I value Defines how many content files
to include in a single transaction
during the migration operation (or
how many files per worker session
if PARALLEL_DEGREE is greater
than 1).
The default is 500.

MIGRATE_CONTENT
PARALLEL_DEGREE I value Defines how many content
migration sessions are created to
do the migration.
The default value is 0, indicating
that a single session is created
and migration will not be done in
parallel.
The maximum value default is
10, but can be modified by setting
the max_cm_parallel_degree
parameter in the server.ini file.
The valid range for this parameter
is 1 to 128. PARALLEL_DEGREE
is set to a value greater than
max_cm_parallel_degree,
the server will use the
max_cm_parallel_degree value
instead.
QUERY S dql_predicate Specifies a DQL predicate to be
used to select content files for
migration.
This must be a valid WHERE
clause qualification.
The predicate is applied to either
content objects or SysObjects,
depending on the setting of the
SYSOBJECT_QUERY argument.
For more information, refer to the
Usage Notes.
SYSOBJECT_QUERY B T (TRUE) or F
(FALSE)
T directs the method to apply the
DQL predicate specified in the
QUERY argument to SysObjects
or the subtype identified in
TYPE_TO_QUERY. F directs the
method to apply the predicate to
content objects.
The default is F.

MIGRATE_CONTENT
TYPE_TO_QUERY S type_name Name of a SysObject subtype.
You can include this only if
SYSOBJECT_QUERY is set to T.
If specified, the DQL predicate
is applied to the subtype, rather
than dm_sysobject type. For
information about the uses of this
argument, refer to Including a
query, page 260.
SOURCE_DIRECT_
ACCESS_OK
B T (TRUE) or F
(FALSE)
Indicates whether the files in
this storage area can be accessed
directly using a path available in
Content Server.
Source store must be an external
store.
The default is F.
DIRECT_COPY B T (TRUE) or F
(FALSE)
Indicates whether to migrate the
content by using the underlying
OS file system command to
directly copy the content to a new
location.
store.
The default is F.
DIRECT_MOVE B T (TRUE) or F
(FALSE)
content by using the underlying
OS file system command to
directly move the content to a new
location.
store.
The default is F.

MIGRATE_CONTENT
UPDATE_ONLY B T (TRUE) or F
(FALSE)
content, or to instead save the OS
file system commands required
to migrate the content into a
command file for later execution.
If UPDATE_ONLY is T, the content
is not migrated, but the saved
commands in the command file
must be executed later.
store.
The default is F.
COMMAND_FILE_
NAME
S command_file_
name
Specifies the name of the file
used to store the OS file system
commands required to migrate
the content. Supply this argument
when UPDATE_ONLY is set to T.
store.
There is no default for this
parameter, so you must supply a
file name.
IGNORE_CONTENT_
VALIDATION_
FAILURE
B T (TRUE) or F
(FALSE)
Indicates if the content migration
occurs on failed content validation.
For example, if 1000 content
objects are migrated and the
second object fails the content
validation, then the following
action occurs based on the value
of the parameter:
• F: Migration is aborted and the
remaining 998 objects are not
migrated.
• T: Failed content objects are
ignored and the remaining 998
objects are migrated.
The default is F.
QUERY_BATCH_SIZE I value Specifies the number of objects
processed per query.
The default is 1000.

MIGRATE_CONTENT
Return value
The method returns a collection with one query result object. The object has one integer property,
named result, whose value is the number of contents successfully migrated.
Permissions
Superuser permissions are required to:
• Specify a content object ID when moving a single content file.
• Move all content from one storage area to another.
• Move content using a DQL predicate that operates on content objects (SYSOBJECT_QUERY=F)
Write permission on the object is required to:
• Specify a SysObject ID when moving the content of a single object.
• Move content using a DQL predicate that operates on a SysObject (SYSOBJECT_QUERY=T)
Description
General notes
MIGRATE_CONTENT is the preferred way to move content files from a file store, ca store, blobstore,
distributed store, or an external store storage area to a file store, ca store, or distributed store storage
area. You cannot move content to a blob store or external store storage area. Additionally, you cannot
use this method to move files to or from turbo storage. If content is stored in a retention-enabled ca
store storage area, you can only move the content to another retention-enabled ca store storage area.
If you move a file to a distributed storage area, the file is placed in the distributed component local to
the Content Server to which you are connected when executing the method.
The method moves content files associated with immutable objects as well as changeable objects. It
does not update a SysObject’s r_modify_date property when the object’s content is moved. You must
use this method if you want to move content files from an unencrypted storage area to an encrypted
storage area. The method allows you to move one file, all files in a particular storage area, or a set of
files selected by a DQL WHERE clause qualification.
The method operates on dmr_content objects, resetting their storage_id property to the new storage
area and resetting the data ticket property. It also updates the i_vstamp property. The i_vstamp
property is also incremented for any SysObject that contains the content file as the first primary
content (page 0). (Consequently, if the affected content objects are associated with replicated
SysObjects, the next replication job for those objects will refresh the objects.) Similarly, if the content
objects are associated with persistently cached objects, the next time the cached objects are accessed
by the client, they will be refreshed.
The method is aware of lightweight SysObjects (LWSOs), so you can move LWSOs using this method
without materializing the objects.

MIGRATE_CONTENT
Here are some general guidelines for using MIGRATE_CONTENT:
• If you want to move content to or from a distributed storage area, you must specify the
dm_distributedstore object as the target or source storage area—do not specify the actual
distributed component.
• Make sure none of the objects whose content you intend to migrate are checked out. If the method
attempts to migrate content associated with a checked out object, the method fails with an error.
• If the content to be migrated is stored in a content-addressed storage area with a retention date,
make sure the retention date has expired. If not, the method fails with an error.
• If the content’s current storage area and the target storage area are the same content-addressed
storage area and the storage area has a retention period, the method updates the content’s
metadata and retention period and creates a new address for the content. However, the content is
not physically moved.
• Back up the repository and the file store storage areas before using the method.
• Make sure that the target storage area has enough disk space to hold the migrated content.
• If you choose not to remove the original content, that content becomes an orphaned content file
and can be removed using dmfilescan.
• If the MIGRATE_CONTENT method fails with an error, the entire batch transaction is rolled back.
If the destination has content files that were created from the successful migrations within the
batch, you can clean up those files by running the Dmfilescan job.
Migrating from an external store
Content stored in an external store is accessed through a custom plugin that takes the request from
Content Server and returns the content. Typically, the content stored in an external filestore was
generated by some other application, or managed by a legacy content management system, and at
some point, management was transferred to Content Server. The storeage plugin is responsible for
translating the content storage location specified by Content Server into the location of the content in
the storage device and retrieving the content.
In many cases, the content in an external store is accessed through the underlying file system, but is
stored in a different file storeage scheme than a standard Content Server filestore. In these cases,
Content Server knows the path to content files in the external storeage area. Typically, migrating
content directly, using the underlying operating system facilities, is much more efficient than fetching
the file to Content Server and writing it out to the target storage location. In some cases, the target
store is on the same physical device, so that migrating content, in effect, can be done by restructuring
the paths to the content files without making copies. In most operating systems, when you move a
file from one location to another on the same physical device, the file does not move, but the directory
structure pointing to it changes.
To support this more efficient migration for external storage areas, some arguments are only
supported for content migrated from an external store:
• SOURCE_DIRECT_ACCESS_OK
• DIRECT_COPY
• DIRECT_MOVE
• UPDATE_ONLY
• COMMAND_FILE_NAME
The argument, SOURCE_DIRECT_ACCESS_OK, indicates that the path to access the content file is
available in Content Server and can be used to access the file. If you select this argument, you can

MIGRATE_CONTENT
optionally also select either DIRECT_MOVE or DIRECT_COPY. If you select DIRECT_MOVE or
DIRECT_COPY, Content Server will use the file path and the OS filesystem move or copy command
to migrate the specified content files to their new filestore. If you do not select DIRECT_MOVE or
DIRECT_COPY, Content Server will use the underlying OS system calls to read in the content files
and write them out to their new location. You cannot use DIRECT_MOVE or DIRECT_COPY if the
target store is configured for compression, encryption, or de-duplication.
Setting SOURCE_DIRECT_ACCESS_OK to TRUE and not setting DIRECT_MOVE or DIRECT_COPY
to TRUE is useful when migrating content from an external store to a CA store. In this case, the
content in external store is directly accessed by the Centera SDK to pump the content to Centera.
Setting SOURCE_DIRECT_ACCESS_OK to TRUE and setting DIRECT_COPY to TRUE is useful in
migrating content from an external store to a filestore, where efficiency of content copying performed
by the OS is leveraged, instead of chunking the content. Note that in this case, you will have to
provision double the storage capacity (the disk space for content in the original store plus the disk
space for content in the destination file store). Setting SOURCE_DIRECT_ACCESS_OK to TRUE and
setting DIRECT_MOVE to TRUE enables the server to move/rename the file instead of copying it,
useful in migrating the content from an external store to a filestore on the same volume, without
having to provision the double the storage.
If you select UPDATE_ONLY, the content is not migrated, but the operating system commands to do
so are saved in the file specified by COMMAND_FILE_NAME and must be executed later (you must
specify this file name if you select UPDATE_ONLY). After you have run the MIGRATE_CONTENT
method with UPDATE_ONLY selected, the content specified to be migrated will not be accessible
until the commands in the command file are executed.
Migrating XML content to an XML Store
XML content files can be moved into an XML store using the MIGRATE_CONTENT method. This
transactional method moves content files associated with immutable objects as well as changeable
objects. The method allows you to move one file, all files in a particular storage area, or a set of files
selected by a DQL WHERE clause qualification. You can move files from a file store, a ca store, a
blobstore, or a distributed store.
Note: The MIGRATE_CONTENT method can be used to migrate content between different XML
stores or from an XML store to a file store. However, it is not possible to move content that was
previously moved from an XML store to a file store to another file store.
XML content migration restrictions
The following restrictions apply for migrating XML content:
• The XML content must be well-formed XML.
If the content is not well-formed, XML store throws an error specifying the content parts that
are not well-formed.
• Migrating XML documents into the XML store only works with XML documents that are stored
in a single page. If you have legacy content that is stored in the two-page model, you need to
convert the content into the single-page model before you start the migration process, as described
in Converting XML content to single-page format, page 260.
• The argument SOURCE_DIRECT_ACCESS_OK is not supported for XML stores.

MIGRATE_CONTENT
Converting XML content to single-page format
XML content that is stored in two-page format must be converted into single-page format before
the content can be migrated into an XML store.
To convert XML content into single-page format:
1. Change the oldest_client_version attribute of the dm_docbase_config object of the repository to
"5.3”.
2. Check out the XML content that needs to be converted using DFC/Webtop. Repeat this step for
each XML object that needs to be converted.
3. Check in the XML content that was checked out in Step 2.
Migrating to content-addressed storage
If you move a file that has a retention policy to a content-addressed storage area, the retention set on
the object is calculated from the retention policy even if the default retention for the content-addressed
storage area is further in the future. If there are multiple retention policies on the object, the method
sets the retention on the file based on the policy with the retention date furthest in the future.
If you are moving content files to a content-addressed storage area that requires a retention date, you
must set the retention date for the content before migrating the content. Use either a setContentAttrs
method or the SET_CONTENT_ATTRS administration method to set the retention date.
Note: A content-addressed storage area requires a retention date if:
• The storage area’s a_retention_attr_name property is set and the a_default_retention_date is
not null, or
• The a_retention attr_name property is set and the a_retention_attr_required property is set to T
Including a query
You can include a DQL predicate in the method arguments. Depending on the value of the
SYSOBJECT_QUERY argument, the predicate is run against instances of a specified object type or
against content objects.
SYSOBJECT_QUERY is F (FALSE) by default, meaning that the predicate is applied to content objects.
You must have Superuser permissions to run the predicate against content objects. When you run the
query against content objects, it is recommended that you use a predicate that references a unique
value in the content object. For example, reference the storage area ID—there is only one storage area
in the repository with any given storeage ID. You cannot include the TYPE_TO_QUERY argument if
SYSOBJECT_QUERY is set to F.
If SYSOBJECT_QUERY is T and the TYPE_TO_QUERY argument is not included, the predicate
executes against instances of the dm_sysobject type. However, if SYSOBJECT_QUERY is T and the
TYPE_TO_QUERY argument is included, the predicate executes against instances of the object type
specified in the TYPE_TO_QUERY argument. When the query is executed, the method affects the
content of only those objects that match the query and for which you have at least Write permission.
When the query is run against instances of an object type, the properties specified in the DQL
predicate must be defined for the object type. Consequently, if you want to specify custom properties
of a SysObject subtype in the predicate, you must include the TYPE_TO_QUERY argument and
identify the subtype in that argument.

MIGRATE_CONTENT
For example, suppose you want to migrate all content of a subtype called "loan_documents”, and that
subtype has a property named customer_state. The following MIGRATE_CONTENT statement will
execute against all loan_document objects and moves the content for those belonging to customers in
CA (California) state:
EXECUTE migrate_content
WITH target_store='storage_005',
query='customer_state=''CA''',
sysobject_query=T,
type_to_query='loan_documents'
Note: Since there are additional arguments for migrating content from an external store, if your query
selects content in external stores and other stores, some undefined conditions can occur. Be sure that
your query selection is all external store content, or no external store content.
Affected content when referencing SysObjects or its subtypes
When you execute the method against SysObjects or subtypes of SysObject, the method acts only on
the first primary content page (page 0) of each object, the renditions of the primary pages, or both.
If a SysObject has multiple primary content pages, the primary pages other than page 0 and their
renditions are not affected by the method. (If you want to move all the content, use a method syntax
that references the content objects, rather than the SysObject.)
You can specify which of these files you want the method to act on by setting the RENDITIONS
argument. The default for RENDITIONS is primary, which means that the method will only operate
on the first primary content files of the SysObjects. If you set the argument to all, the method moves
both the primary content page and its renditions. If you set the argument to secondary, the method
moves only the renditions but not the primary content.
Controlling the size of the operation
If you are moving a large number of files, you can control the operation using the
MAX_MIGRATE_COUNT and BATCH_SIZE arguments. MAX_MIGRATE_COUNT controls how
many content files are moved with each execution of the method. Use this argument if you have a
large number of files to migrate and want to perform the migration in smaller steps rather than all
in one operation. BATCH_SIZE controls how many content files are moved in a single transaction
within the migration operation (or how many files per worker session if PARALLEL_DEGREE is
greater than 1). Setting BATCH_SIZE can help protect the operation from overrunning system and
database resource limits during the operation.
If you set MAX_MIGRATE_COUNT, make sure that each execution of the method does not attempt
to move content files that have already been migrated in previous executions of the method.
Parallel content migration
In previous releases, the MIGRATE_CONTENT method performed content migration sequentially, in
a single session. Migrating a large number of objects in a production environment could take a long
time. To achieve better performance, an administrator would execute the method in multiple sessions
concurrently (in parallel), with each session migrating a unique set of objects. This workaround can
be challenging to implement because of the difficulty of identifying individual sets of objects to
migrate. Additionally, this approach may lead to database deadlocks because of database update
collisions caused by multiple sessions attempting to migrate the same object.
The PARALLEL_DEGREE parameter controls how many content migration sessions are created to do
the migration. Using the built-in parallel migration facility avoids the difficulties caused by manually

MIGRATE_CONTENT
implementing concurrent sessions. Use of the PARALLEL_DEGREE parameter with a value greater
than zero requires a Content Storage Services license.
For any value of the PARALLEL_DEGREE parameter, a master session is created. If the value of the
parameter is 0, the master session does all the migration; the migration is not done in parallel. For
positive values of the parameter, the master session creates at least that number of worker sessions.
The master session validates the method arguments, finds all the qualified content objects, and
assigns individual content objects for migration to the individual worker sessions. The master session
waits until all the worker sessions have terminated. Then it aggregates the results and statistics,
cleans up worker sessions if necessary, destroys the task queue, and returns.
There are two special cases that will cause problems if migrated by different worker sessions:
1. Any parent of a migrating content object has more content objects to migrate.
2. The content object has more than one parent and the qualification predicate is specified against
dm_sysobject.
In case 1, the different worker sessions will attempt to migrate the different content objects. One
session will succeed, but the others will fail due to an i_vstamp mismatch.
In case 2, a content object shared by two or more sysobjects may get assigned to multiple worker
sessions. As a result, it is possible for the sessions to migrate the same content object simultaneously,
and thus cause a deadlock or some indefinite behaviors.
To avoid those difficulties, the master session will create a special worker session to migrate any
content that is in a special case. The special session will migrate all of the special case content objects.
If there is a special worker session, there will be total of PARALLEL_DEGREE + 1 worker sessions.
For example, if PARALLEL_DEGREE is 3, there will be total of 4 worker sessions. These worker
sessions are numbered as 0, 1, 2, and 3. Worker session 0 always acts as the special worker session.
The special session will always try to acquire tasks from the special task queue. If the queue is empty,
it will try to acquire a task from the regular task queue. In this way, the special session will share
the workload with the other sessions. However, if most of the migration objects are in one of the
special cases, the special session will perform most of the migration sequentially. If the content to be
migrated falls mostly into the special cases, the migration will be mostly sequential.
The BATCH_SIZE parameter applies to each worker session individually. For example, if the
BATCH_SIZE is 400, each worker session will migrate 400 content objects in a single transaction.
Removing the original content
When you move content, you can choose whether to remove the content from the current (source)
storage area or not. However, you cannot remove the original content when the source storage
area is an external store, so the REMOVE_ORIGINAL argument is set to FALSE when the source is
an external store.
If you choose not to remove the content from the current storage area and no other content objects
reference that file, Content Server writes a message to the log file to identify the content file as an
orphaned file that must be removed manually. (The file is now an orphan because it is no longer
referenced by any content objects.) If the source is an external store, the log file will not identify the
orphan content files.
If you choose to remove the content from the current storage area and the storage area is a file store
storage area, Content Server first checks to ensure that no other content objects reference that content.
If another content object is found that references the content, the content is not removed. (Multiple
content objects may reference one content file if content duplication checking and prevention

MIGRATE_CONTENT
is enabled for the file store storage area. For information about that feature, refer to the EMC
Log file use
You must identify a log file location if you are migrating an entire storage area or including a DQL
predicate or setting REMOVE_ORIGINAL to false in the method arguments. Specifying a log file
location is only optional if you are migrating just a single object. For all other cases, you must specify
a log file location. If the file you specify already exists, the method appends to that file.
Note: If you are migrating a single content file, you can retrieve messages using a
IDfSession.getMessage method.
The method logs the following messages:
• A success message for each content object successfully migrated without errors.
• A failure message for each content object that was not successfully migrated. The message
includes the reason for the failure.
• Messages for all database errors. (These are written to the server log and the session log file.)
• A success or failure message for each batch in the operation.
• Performance metrics about the amount of time spent in the RDBMS and in the storage area. The
information is logged for each object successfully migrated, and accummulated totals for ever 100
objects, and a summary total for all objects.
Here are samples of the performance messages. The first sample shows the column headers for
entries for individual objects and the information about one object:
Fri Jan 12 10:03:52 2007 953000
[DM_CONTENT_T_MIGRATE_CONTENT_PERF_REC]info: "Total Time (secs)
Total Storage Time (secs) Total Database Time (secs) Storage Time
for current file (secs) Database Time for current file (secs) Total
Objects Current file size Xput (bytes/sec) for current file Max
storage time (secs) Max database time (secs) Max file size Migration
Xput (docs/sec) Storage Xput (bytes/sec) Total KBytes"
Fri Jan 12 10:03:52 2007 953000
[DM_CONTENT_T_MIGRATE_CONTENT_PERF_REC]info: "6.765 6.703 0.062
6.703 0.062 1 8182 1209.46045824095 6.703 0.062
8182 0.147819660014782 1220.64747128152 7"
This sample shows the summarizing entry for 100 objects:
Fri Jan 12 10:05:19 2007 674000
[DM_CONTENT_T_MIGRATE_CONTENT_PERF_FREC]info: "100 Objects Migrated.
256.513 1.289 3974"
Finally, here is a sample of the final summarizing entry:
Fri Jan 12 10:07:02 2007 270000
[DM_CONTENT_T_MIGRATE_CONTENT_PERF_FINAL]info:
"250 Migrated. Storage Time(secs): 547.891, Total Database Time: 3.141,
Total Content Size (KBytes): 9415
None

MIGRATE_CONTENT
Examples
This example migrates a single content file, represented by the content object 0600000272ac100d:
EXECUTE migrate_content FOR 0600000272ac100d
WITH target_store='filestore_02'
The next example migrates all the content from filestore_01 to engr_filestore:
WITH source_store='filestore_01',
target_store='engr_filestore',batch_size=100,
log_file='C:tempmigration.log'
The next example uses a DQL predicate to select the content to migrate. The query uses content size
to select the content.
WITH target_store='engr_filestore',
query='content_size>1000',max_migrate_count=1000,
batch_size=100,log_file='C:tempmigration.log'
The final example migrates all the content from the external filestore legacy_filestore to filestore_03.
The external content files allow direct access, and the files are moved directly by issuing the
underlying OS filesystem commands:
WITH source_store='legacy_filestore', target_store='filestore_03',
source_direct_access_ok=TRUE, direct_move=TRUE

MIGRATE_TO_LITE
MIGRATE_TO_LITE
Purpose
Migrates standard objects to lightweight objects.
Syntax
To generate, and optionally run, a script to split standard objects into a shareable parent object and
a lightweight object:
EXECUTE migrate_to_lite WITH source_type='source_type'
,shared_parent_type='shared_parent_type'
,execution_mode='execution_mode_split'[,recovery_mode=TRUE|FALSE]
,parent_attributes='parent_properties'
[ft_lite_add='property_list' | ft_base_add='property_list']
To generate, and optionally run, a script to migrate standard objects to lightweight objects:
EXECUTE migrate_to_lite WITH
shared_parent_type='shared_parent_type'
[,lightweight_type='lightweight_type']
,execution_mode='execution_mode'[,recovery_mode=TRUE|FALSE]
{,parent_sql_predicate='parent_sql_predicate'}
{,parent_id='ID'}
[,default_parent_id='ID']
[ft_lite_add='property_list' | ft_base_add='property_list']
Arguments
Table 69. MIGRATE_TO_LITE arguments
source_type string source_type Specifies the name of the source
type which needs to be split.
This type must be a dm_sysobject
subtype, not the dm_sysobject
itself. After the conversion, this
type will become the lightweight
type whose shared parent type
is specified in the following
parameter, shared_parent_type.

MIGRATE_TO_LITE
shared_parent_type string shared_parent_type Defines the shareable type to
use. If the specified type is not
a shareable type, then it will be
converted into one. The specified
type needs to be a subtype of
dm_sysobject and cannot be
dm_sysobject. If the specified type
is already a shareable type, then
the lightweight_type parameter
must be specified.
execution_mode string execution_mode_
split
Specifies the operation to perform.
The allowable values for the first
form of the method are:
• Split and Finalize
• Split without Finalize
• Finalize
These values are described in
detail in Execution Mode Split
Values, page 269
execution_mode Specifies the operation to perform.
The allowable values for the
second form of the method are:
• Generate Script Only
• Run and Finalize
• Run without Finalize
• Cancel
• Finalize
These values are described in
detail in Execution Mode Values,
page 269

MIGRATE_TO_LITE
recovery_mode boolean TRUE|FALSE Use this flag when the
EXECUTION_MODE is either
Run and Finalize, Run without
Finalize, Split and Finalize, or Split
without Finalize. The default is
FALSE. If the flag is set to TRUE,
then the internally-created interim
types (and tables) will be dropped
before the process begins.
For Split and Finalize and Split
without Finalize the shared parent
will be reused but the interim child
types (and tables) will be dropped.
parent_attributes string parent_properties Specifies properties in the
source type which will be split
into the parent type specified
in shared_parent_type. The
parameter contains a list of
comma-separated property
names from the source type. The
remainder of the properties in the
source_type will remain there.
lightweight_type string lightweight_type If the specified type is a standard
type, it will be converted to a
lightweight type. The type needs
to be a direct subtype of the one
specified in shared_parent_type.
If the specified type is already a
lightweight type, then you must
also set recovery_mode to TRUE,
and specify which lightweight
objects go with which parents.
Use the parent_sql_predicate,
parent_id, and default_parent_id
arguments to specify lightweight
objects and parents.
parent_sql_predicate string parent_sql_
predicate
This repeating string parameter
specifies an SQL (not DQL),
predicate qualifying a set of
lightweight SysObjects whose
parent will be set to the following
parameter, parent_id. This
parameter is allowed only when
the execution_mode is either Run
and Finalize or Finalize.

MIGRATE_TO_LITE
parent_id ID ID This repeating ID parameter
corresponds to the above
parent_sql_predicate parameter.
The lightweight SysObjects that
match the SQL predicate will
be assigned to this parent. The
specified ID needs to be a legal ID.
This parameter is allowed only
when the execution_mode is either
Run and Finalize or Finalize.
default_parent_id ID ID This ID parameter is used for
the remaining materialized
lightweight SysObjects created by
this method that are not qualified
with any predicate specified in a
parent_sql_predicate parameter.
If this parameter is not specified,
then the i_sharing_parent
property of all the remaining
lightweight SysObjects will be
set to the ID of the lightweight
object itself. That is, they will be
considered as materialized. This
parameter is allowed only when
the execution_mode is either Run
and Finalize or Finalize.
ft_lite_add string property_list A comma separated list of the
properties to add to the fulltext
index for the lightweight type,
or the string ’ALL’ to add all the
lightweight properties.
ft_base_add string property_list A comma separated list of the
properties to add to the fulltext
index for the shareable parent
type, or the string ’ALL’ to add all
the shareable properties.
Return Value
MIGRATE_TO_LITE returns an object id for the text file containing the SQL script to do the migration.
The script will be generated regardless of the execution_mode specified.

MIGRATE_TO_LITE
Execution Mode Split Values
For the first form of the method, execution_mode can take on the following values:
• Split and Finalize—Generates the script and immediately runs it.
• Split without Finalize—Generates the script, but the original types are not changed. You can then
validate the changes before making the changes permanent.
• Finalize—Makes the changes. For this mode, only the source_type, shared_parent_type, and
execution_mode parameters are required.
When you use this command to split a type, the indexes on the type attributes are dropped. You will
need to evaluate which attributes to index, and create those indexes, after splitting the type into
a lightweight and a shareable type.
Execution Mode Values
For the second form of the method, execution_mode can take on the following values:
• Generate Script Only—Generates the script, but does not execute it.
• Run and Finalize—Generates the script and immediately runs it.
• Run without Finalize—Generates the script and immediately runs it, but the original types are not
changed. All the changes will be applied to corresponding internally-created types (and tables)
with the name dm_lw_id_of_my_parent_type and dm_lw_id_of_my_child_type created with the
changes. You can then validate the changes before making the changes permanent.
• Cancel—Allows you to remove the interim types and tables created by the Run without Finalize
mode. Typically you would do this if you decided not to finalize the changes. After removing
the interim types and tables, the method ends.
• Finalize—Makes the changes permanent by replacing the original types with the internal types
and dropped the internal types afterwards. If this option is provided when there is nothing to
swap, a warning will be reported back to client and recorded in server log.
Permissions
You must have Superuser privileges, or be the owner of the involved types, to use this method. But
also see the note for Oracle installation users following the description section.
Description
Use this method to migrate standard types and objects to shareable and lightweight types and objects.
This method can also be used to reparent lightweight objects.
The first form of the command is used to convert a standard type to a lightweight and a shareable
type. In this form of the command, the shareable type is formed by splitting off properties from the
original standard type. The remaining properties are used to create the lightweight type. Specify
the name of the standard type to split in the source_type parameter, the new shareable type in
the shared_parent_type parameter, and list the properties to split off of the original type in the
parent_attributes parameter. When you use this method to split standard objects, each lightweight
object has its own private parent. In other words, each lightweight object is materialized. You

MIGRATE_TO_LITE
can then reparent the lightweight objects by using the second form of the command, specifying
parent_sql_predicates and parent_ids to point the lightweight objects to shareable parents.
The second form of the command does not move properties from one type to another. You will use it
when all the non-inherited standard type properties will remain with the lightweight type. You can
also use this form to reparent lightweight objects.
If you use the second form of the command and do not specify a lightweight type, and you specify a
standard type in the shared_parent_type argument, the standard type is changed into a shareable
type.
The second form can convert standard objects to lightweight objects and parent them with specified
shareable parents. To do this, you specify a shared_parent_type and a lightweight_type. Conceptually,
you can imagine that all the standard objects specified by the lightweight_type argument are
converted to materialized lightweight objects pointing to private parents of the shared_parent_type.
Next, all the objects of the lightweight type that match the parent_sql_predicate argument are made
to point to the shareable parent with the associated parent_id. The method continues associating
each group of lightweight objects that matches each subsequent parent_sql_predicate and parenting
that group with the associated shareable parent. If any unconverted materialized lightweight objects
remain after converting each parent_sql_predicate group, the remaining objects are parented with
the default_parent_id shareable parent, if specified. Typically, you would create some shareable
parents ahead of time, and then use this method to associate that list of objects with the groups
defined by the parent_sql_predicate.
To use this method to reparent lightweight objects, specify the shared_parent_type and the
lightweight_type, the parent_sql_predicate to select which objects to reparent and the parent_id to
reparent to. You must also specify recovery_mode as TRUE, or the method won’t work with already
converted objects of lightweight_type.
Fulltext Support — Use the ft_lite_add and ft_base_add arguments to specify which properties
are included in the fulltext index. These arguments add the shareable parent type properties and
the lightweight type properties to fulltext indexing. The ft_lite_add variations add the properties
specified for the lightweight type, and the ft_base_add variations add the properties specified for
the shareable parent type.
Note: In order to use this method with an Oracle installation, the user account that Content Server
uses to access the database must have enhanced privileges. The following example assumes that
you have logged into SQLPLUS as the SYS user, in SYSDBA mode to grant these privileges to the
repository_1 user:
grant DBA to repository_1;
These are powerful privileges and should be revoked when the migration is completed.
Examples
This example generates a script to convert usertype_1 to a shareable type. You do not execute this
script against your database. It is created so you can examine the commands that will be issued
when you later execute this method using another form of the method, either Run and Finalize or
Run without Finalize:
EXECUTE migrate_to_lite WITH "shared_parent_type"='usertype_1'
,"execution_mode"='Generate Script Only'

MIGRATE_TO_LITE
To get the script, you can issue the following DQL statement (assume that the result of the earlier
method returned the object ID 0855706080007c19):
This example converts usertype_2 to a shareable type in one operation:
,"execution_mode"='Run and Finalize'
This example converts usertype_3 to a shareable type in two steps. After step one, you can examine
the script and the temporary types and tables to verify the repository changes. For step one, use the
Run without Finalize execution_mode to run the script and create the temporary tables:
,"execution_mode"='Run without Finalize'
Then, for step two, use the Finalize execution_mode to swap in the temporary tables and commit
the change:
,"execution_mode"='Finalize'
This example converts an existing standard type, emc_payment_check, and all its objects, into a
lightweight type and a shareable type, emc_payment_bank. Each lightweight object will have its own
private parent object. The parent type will take the attributes bank_code and routing, but the other
attributes will be part of the lightweight type.
In this example, there are already objects created of type emc_payment_check. You can create the
type using the following command:
CREATE TYPE "emc_payment_check" (
account integer,
check_number integer,
transaction_date date,
amount float,
bank_code integer,
routing integer
) WITH SUPERTYPE "dm_document" PUBLISH
You can also use the batch example program listed in EMC Documentum High-Volume Server
Development Guide to create a number of objects for this example.
Use the following command to split the emc_payment_check objects into lightweight objects of type
emc_payment_check, and private parents of type emc_payment_bank. The bank_code and routing
attribute values will be associated with the private parent, and the other attributes will be associated
with the lightweight objects. The attributes account and check_number of the (now lightweight) type
emc_payment_check will be fulltext indexed:
EXECUTE migrate_to_lite WITH "source_type"='emc_payment_check'
,"shared_parent_type"='emc_payment_bank'
,"execution_mode"='Split and Finalize'
,"parent_attributes"='bank_code,routing'
,"ft_lite_add"='account,check_number'
The following example takes the materialized lightweight objects converted by the last example and
reparents them. Objects with check_numbers less than five are reparented to a separately created
emc_payment_bank shareable parent, and the remaining materialized objects are reparented to
another separately created shareable parent:
EXECUTE "migrate_to_lite" WITH
"shared_parent_type"='emc_payment_bank',"lightweight_type"='emc_payment_check'
,"execution_mode"='Run and Finalize',"parent_sql_predicate"='check_number<5'

MIGRATE_TO_LITE
,"parent_id"='0925392d80001d5d'
,"default_parent_id"='0925392d80001d5e'

MODIFY_TRACE
MODIFY_TRACE
Purpose
Turns tracing on and off for full-text index query operations.
Syntax
EXECUTE modify_trace
WITH subsystem='fulltext',value='tracing_level'
Arguments
Table 70. MODIFY_TRACE arguments
SUBSYSTEM S fulltext The keyword fulltext indicates
that you want to turn on tracing for
the full-text querying operations.
VALUE S tracing_level The tracing_level must be one of:
none, to turn off tracing
all, to log both Content Server and
full-text querying messages
Return value
MODIFY_TRACE returns a collection with one query result object. The object has one Boolean
Permissions
Description
MODIFY_TRACE turns tracing on and off for full-text index operations. When tracing is on, tracing
information for all full-text queries is logged. The trace information is placed in the server’s log file.
The tracing information includes informational, verbose, and debug messages. All trace messages
include time stamps and process and thread ID information.

MODIFY_TRACE
If the method returns FALSE, the level of tracing remains unchanged. For example, if tracing is
currently set at all and you execute MODIFY_TRACE to reset the level to none, but the method
returned FALSE, the trace level remains at the all level.
LOG_OFF, page 242
LOG_ON, page 243
SET_OPTIONS, page 317
Examples
The following example turns on full tracing:
EXECUTE modify_trace
WITH subsystem = 'fulltext',value = 'all'

MOVE_INDEX
MOVE_INDEX
Purpose
Moves an existing object type index from one tablespace or segment to another. (This method is
not supported for servers running against DB2.)
Syntax
EXECUTE move_index FOR 'dmi_index_obj_id'
WITH name = 'new_home'
[,global_index=true|false]
Arguments
Table 71. MOVE_INDEX arguments
NAME S new_home Identifies the new tablespace or
segment for the index. Use the
name of the tablespace or segment.
GLOBAL_INDEX B TRUE or FALSE Indicates whether the index being
created is a global index. This
parameter can only be used for
partitioned types. The default is
FALSE.
Return value
MOVE_INDEX returns a collection with one query result object. The object contains one Boolean
Permissions
Description
MOVE_INDEX moves an existing object type index from one tablespace or segment to another. You
can obtain the index’s object ID from the index’s index object. Refer to the EMC Documentum Content
Server System Object Reference manual for a list of the properties defined for the index type.

MOVE_INDEX
GLOBAL_INDEX is a boolean parameter that indicates whether the index being moved is a global
index. The default value is FALSE. This parameter can be used only when the type is a partitioned
type, and is only useful if the index is a unique index.
MOVE_INDEX only operates on object type indexes that are represented in the repository by
dmi_index objects.
Note: Those indexes that do not have a dmi_index object (DMI_OBJECT_TPE_UNIQUE and
DM_FED_LOG_INDEX for example) may be moved manually. However, be sure to shut down
Content Server before moving them.
Examples
This example moves the index represented by 1f0000012643124c to the index2 tablespace:
EXECUTE move_index FOR '1f0000012643124c'
WITH name = 'index2'

PARTITION_OPERATION
PARTITION_OPERATION
Purpose
Creates partitioning scheme objects and SQL scripts to control repository partitioning.
Syntax
To create a partitioning scheme object:
EXECUTE partition_operation WITH operation='create_scheme',partition_
scheme='scheme_name'{,partition_name='partition_name',range=right_
limit,tablespace='storage_name'}
To generate a script to apply a partition scheme to types and/or registered tables:
EXECUTE partition_operation WITH operation='db_partition',partition_
scheme='scheme_name'{,type_name='type_name'}{,table_name=table_
name,table_own='table_owner'}
To add a partition (range must be greater than the upper bound of the scheme):
EXECUTE partition_operation WITH operation='add_partition',partition_
scheme='partition_scheme'{,partition_name='partition_name',range='right_
limit',tablespace='storage_name'}
To delete a partition (must be last partition in the scheme):
EXECUTE partition_operation WITH operation='delete_partition',partition_
scheme='partition_scheme',partition_name='partition_name'
To split a partition (range must be one of the values defined for the partition to be split, the source
partition):
EXECUTE partition_operation WITH operation='split_partition',partition_
scheme='partition_scheme',source_name='source_name',partition_name='partition_
name',range='right_limit',tablespace='storage_name'
To merge a partition with the one to its left (with a lower range):
EXECUTE partition_operation WITH operation='merge_partition',partition_scheme=
'partition_scheme',partition_name='partition_name'
To switch out a partition to a temporary table:
EXECUTE partition_operation WITH operation='switch_out',partition_scheme=
'partition_scheme',temp_table_suffix='temp_table_suffix,partition_name=
'partition_name',execute_now=TRUE|FALSE
To switch data into a partition from a temporary table:
EXECUTE partition_operation WITH operation='switch_in',partition_scheme=
'partition_scheme',temp_table_suffix='temp_table_suffix,partition_name=
'partition_name',execute_now=TRUE|FALSE

PARTITION_OPERATION
Arguments
Table 72. PARTITION_OPERATION arguments
operation string listed in
description
column
Defines the operation that you
want to execute. The only values
allowed are:
• create_scheme
• db_partition
• add_partition
• delete_partition
• split_partition
• merge_partition
• switch_out
• switch_in
partition_scheme string scheme_name
partition_name string partition_name The name of the partition. Some
databases may restrict the choices
for a partition name. For example,
choosing default as a partition
name causes an error when you
execute the partitioning script.
range integer right_limit Uppermost i_partition value of an
object placed in this partition is
equal to (right_limit - 1)
tablespace string storage_name The tablespace for the partition.
You can find the tablespace
name for your repository
by accessing the underlying
database. By default, the name is
dm_<repository_name>_docbase.
type_name string type_name Specifies a type to partition. Must
be a partitionable supertype.
Create the type with CREATE
PARTITIONABLE TYPE, or
modify a type to be partitionable
with ALTER TYPE ... ENABLE
PARTITION.

PARTITION_OPERATION
table_name string table_name Specifies a registered table. The
table must already have an
i_partition column. Do not use
this for repository-created tables;
use the type_name argument for
those tables.
owner_name string owner_name The table owner name.
source_name string source_name The name of the partition to split.
temp_table_suffix string temp_table_suffix The suffix for the temporary table.
When tables are switched in or
out, the table name must be the
name of the type or registered
table with the suffix applied.
Use a short, one or two character
suffix.
Return Value
For create_scheme, PARTITION_OPERATION returns a Boolean value and an object id. The Boolean
value indicates if the operation succeeded. The object ID is either the ID of the partition scheme object,
or the object ID of the SQL script that is run against the database to perform the partitioning operation.
Note: One way to retrieve the script is to issue the following DQL statement (assume that the result of
PARTITION_OPERATION returned the object ID 0855706080007c19):
That statement will return the location of the file.
Permissions
You must have Sysadmin or Superuser privileges to use all the operations of this method. If you have
type owner privileges, you can use the db_partition operation for that type.
In Oracle installations, you must be logged in as sysdba to run the script.
Description
Use this method to create a partitioning scheme object (dm_partition_scheme). This object will record
the types and tables that use the scheme, the name and range of each partition, and the tablespace (or
filegroup) where each partition is stored. When a partitioning scheme object is first created, there are
no types or registered tables associated with it, just the name, range and tablespace for the partitions
in the scheme. When you associate types or registered tables with a partitioning scheme, a script is
created that must be run against the underlying database to partition the database tables to match the
partitioning scheme. Content Server does not run this script since it may take a long time to perform
the partitioning on a large database, and you may want to examine the script before running it.

PARTITION_OPERATION
You also use this method to modify a partition scheme. Modifying the scheme also creates a database
partitioning script that Content Server runs against the database for you.
When you run the script-creating variations of this method, the method returns the r_object_id of the
partitioning script to run in your database. The partition scheme object information and the database
are not consistent until the script is run, so the is_validated property of the partitioning scheme object
is set to FALSE. After the script is created, you must run it against the underlying database, using the
database’s SQL facility. The scripts make the changes to the database that reflect the information in
the partition scheme object. As the script is successfully run against the database, the is_validated
property of the partitioning scheme object is set to TRUE.
If you did not enable partitioning when you created your repository, create a scheme, and then
use the db_partition operation without specifying any tables or types to partition the repository
intrinsic types.
The following are the intrinsic type hierarchies that are partitioned when you create a repository with
partitioning enabled, or you later enable partitioning:
• dm_acl
• dm_sysobject
• dmr_containment
• dm_assembly
• dm_relation
• dm_relation_type
• dmi_otherfile
• dmi_replica_record
• dmr_content
• dmi_subcontent
• dmi_queue_item
Originally partition-enabled repository compared to originally non-partitioned
repository
You can use partitioning even if you don’t create a partition-enabled repository. You can partition the
repository after it has been created. In either case, you will have a partitioned repository.
If you create a partition-enabled repository, the eleven intrinsic type hierarchies listed above will be
partitioned, each in its own partition scheme. If you want more than one of those hierarchies to be
partitioned identically, you will need to run PARTITION_OPERATION on each hierarchy you want
to modify separately. However, you can use separate partitioning schemes on those hierarchies if
you need to for your application. There is no way to merge the partition schemes so that multiple
hierarchies are configured with a single scheme. You can add custom type hierarchies (where the
topmost custom type specifies supertype NULL), to any of the intrinsic type schemes and manage
them together, if needed.
In contrast, if you partition-enable your repository after creation, all the intrinsic types are configured
with a single partitioning scheme. In this case, all the type hierarchies are partitioned identically.
There is no way to separate them into different partition schemes if you later decide you need
to do this.
You should plan your repository partitioning carefully before deciding which alternative meets
your needs.

PARTITION_OPERATION
Create_scheme
To create a partition scheme, use the create_scheme operation. You will specify a name for the
scheme. Each scheme has a unique name. In this manual we use the convention that the name of a
partition scheme ends with the characters _sch. In addition to the scheme, you will specify one or
more partition name, range, and tablespace. The name is used to refer to the partition. The range is a
single integer that is one larger than the largest value in the i_partition attribute for any objects in
that partition. The tablespace specifies where the partition is stored. The r_object_id of the partition
scheme object is returned.
The following command creates a partition scheme named my_partition_sch, with four partitions
named zero, first, second, and third. Partition zero contains objects with i_partition values from 0 to
9, partition first contains objects with i_partition values from 10 to 19, partition second has 20 to 29,
and partition third has 30 to 39. Notice that the range for partition zero is set to 10, but partition zero
contains objects with i_partition values from 0 to 9:
EXECUTE partition_operation with "operation"='create_scheme',
"partition_scheme"='my_partition_sch',
"partition_name"='zero',range=10,"tablespace"='dm_techpubsglobal_docbase',
"partition_name"='first',range=20,"tablespace"='dm_techpubsglobal_docbase',
"partition_name"='second',range=30,"tablespace"='dm_techpubsglobal_docbase',
"partition_name"='third',range=40,"tablespace"='dm_techpubsglobal_docbase'
Db_partition
To apply a partition scheme, use the db_partition operation. You specify the partition scheme to
apply, and the types and tables to partition. This operation adds the specified types and registered
tables to the partition scheme object, and creates a script that partitions the database tables. A type
or registered table can be associated with only one partition scheme, so the command will fail if a
type or registered table is listed in any other partition scheme.
Use this operation to partition-enable a repository that was not originally created as partition-enabled.
If you run this operation and do not specify any types or registered tables, the base types will be
partitioned. Also, see the note below about increasing the number of cursors for an Oracle installation.
After you have created a partition scheme, perform these steps to partition the repository:
1. Use the db_partition operation of the PARTITION_OPERATION administrative method to
create the partitioning script
2. Stop Content Server
3. Execute the partitioning script against the database
4. Restart Content Server
Note: Do not execute the script if there have been changes to the repository schema after it was
generated. Create a new script if the schema changes.
The following command creates a script to apply my_partition_sch to the type, my_type. The
command returns the r_object_id of the script, and sets the is_validated property of my_partition_sch
to FALSE:
EXECUTE partition_operation with "operation"='db_partition',
"partition_scheme"='my_partition_sch',"type_name"='my_type'

PARTITION_OPERATION
After running the script, the is_validated property is TRUE, and the tables in the underlying database
are partitioned as set out in the my_partition_sch object.
Add_partition
To add a partition, use the add_partition operation. You specify the partition name, range, and
tablespace for each partition to add. The range of the new partition must be greater than the upper
bound of the current partition scheme. Any types or registered tables already attached to this
partition scheme will be modified. This operation modifies the partition information in the partition
scheme object, and creates and runs a script to modify the partitioning in the database.
The following command adds a partition to the partition scheme my_partition_sch, and runs a script
to add the additional partition to the database tables. In this example assume that my_type was
partitioned in the earlier db_partition example:
EXECUTE partition_operation WITH "operation"='add_partition',
"partition_scheme"='my_partition_sch'
,"partition_name"='fourth',range=50,"tablespace"='dm_techpubsglobal_docbase'
Delete_partition
To delete a partition, use the delete_partition operation. The partition must be the last partition in
the partition scheme. Like add_partition, this operation modifies the partition information in the
partition scheme object, and runs a script to modify the partitioning in the database.
The following command removes the last partition (created in the add_partition example):
EXECUTE partition_operation WITH "operation"='delete_partition',
"partition_scheme"='my_partition_sch',"partition_name"='fourth'
Split_partition
To split a partition, use the split_partition operation. You specify a source partition to split, and the
name, range, and tablespace of the new partition. The range value must fall into the defined range
of the source partition. The new partition is to the left of the source partition (the new partition
range is lower than the modified source partition).
The following command splits the zero partition into partitions five and zero:
EXECUTE partition_operation WITH "operation"='split_partition',
"source_name"='zero',
"partition_name"='five',range=5,"tablespace"='dm_techpubsglobal_docbase'
If you used the previous examples, before executing this command, the partition names and ranges
are: zero (10), first (20), second (30), third (40). After executing this command, they are: five (5), zero
(10), first (20), second (30), third (40).
Merge_partition
To merge two partitions, use the merge_partition operation. This operation merges the partition
specified into the partition on its left (into the lower range numbered partition).

PARTITION_OPERATION
The following command merges the partition named five into the next lower range partition
(partition zero, if you’ve used the previous examples):
EXECUTE partition_operation WITH "operation"='merge_partition',
"partition_scheme"='my_partition_sch',"partition_name"='five'
Switch_out
To move data out of a partition into temporary tables, use the switch_out operation. For some
databases (SQL Server), the temporary tables (if they’ve already been created), must be empty. After
the switch_out operation, the temporary table contents will be in the partition, and the data previously
in the partition will reside in the temporary tables for the types and registered tables in that partition
scheme. If you specify execute_now as TRUE, Content Server will execute the SQL script and switch
out the data, otherwise you will need to run the script yourself. The temporary table names will
match the type or registered table name that you switch out with the temp_table_suffix added.
While switch_out is executing, the data in the partition is not consistent, so you must insure that there
is no repository access to that partition during the switch.
The following example switches out the data in the second partition into a temporary table named
my_type_x (my_partition_sch only has one type, my_type, in it). Since execute_now is specified, you
do not need to run the SQL script; it is done for you:
EXECUTE partition_operation WITH "operation"='switch_out',
"temp_table_suffix"='_x',"partition_name"='second',"execute_now"=TRUE
Switch_in
To move data into a partition from temporary tables, use the switch_in operation. The partition must
be empty for SQL Server. Create the temporary tables and fill them with the data to switch into the
partition. After the switch_in operation, the temporary table contents will be in the partition, and the
data previously in the partition will reside in the temporary tables. If you specify execute_now as
TRUE, Content Server will execute the SQL script and switch out the data, otherwise you will need to
run the script yourself. The temporary table names will match the type or registered table name that
you switch out with the temp_table_suffix added.
You must have created all the temporary tables to switch into the empty partition. One quick way to
create those tables, is to use the switch_out operation on an empty partition. When the operation
executes, it will create all the required tables. Since the partition was empty, all the tables will be, too.
You can then fill the tables with data.
While switch_in is executing, the data in the partition is not consistent, so you must insure that there
is no repository access to that partition during the switch.
The following example switches in the data in the from the temporary table named my_type_x
(my_partition_sch only has one type, my_type, in it) into the partition named second. Since
execute_now is specified, you do not need to run the SQL script; it is done for you:
EXECUTE partition_operation WITH "operation"='switch_in',
"temp_table_suffix"='_x',"partition_name"='second',"execute_now"=TRUE

PARTITION_OPERATION
Oracle installations
Note: If you experience difficulties with the following information, please consult the Oracle
documentation or Oracle support for assistance.
In order to use this method with an Oracle installation, you must run the script as SYSDBA.
To increase the number of cursors, consult your Oracle documentation. It may tell you to use a
command similar to:
ALTER SYSTEM SET OPEN_CURSORS=2000 SCOPE=MEMORY SID='*';
For example, if dmadmin is the installation owner:
C:Documents and Settingsdmadmin>sqlplus "/as sysdba"
@ C:Documentumdatatestenvcontent_storage_010000005780000119.txt
to alter the number of open cursors.
If you exit the script with an error (from inadvertently exceeding the number of open cursors, for
example), correct the error, and rerun the script, you may see an error message like:
ORA-12091: cannot online redefine table "TECHPUBS"."DMC_WFSD_ELEMENT_S" with
materialized views
or like:
ORA-23539: table "TECHPUBS"."DM_PLUGIN_R" currently being redefined
caused by leftover temporary items from the previous script failure. One way to correct this error is
to run a command similar to:
execute DBMS_REDEFINITION.ABORT_REDEF_TABLE('<Schema Name>','<Table Name>'
,'<Table Name>I');
Where <Table Name>I is the intermediate table name used for the redefinition. From the first error
message, we would use this command:
execute DBMS_REDEFINITION.ABORT_REDEF_TABLE('TECHPUBS','DMC_WFSD_ELEMENT_S'
,'DMC_WFSD_ELEMENT_SI');
or from the second:
execute DBMS_REDEFINITION.ABORT_REDEF_TABLE('TECHPUBS','DM_PLUGIN_R'
,'DM_PLUGIN_RI');
Switch in and switch out partitions
Switching data into a partition must be carefully planned. Typically, you will create a number of
offline tables to load with data, use whatever native database method is available to load the tables
into temporary tables in the database, create the table indexes, and then swap the tables into the

PARTITION_OPERATION
prepared repository partition. This technique can load large amounts of data into a repository while
causing a minimum of disruption to normal use of the repository. This technique can also be used to
remove large amounts of data from a repository by switching out a partition for empty offline tables.
The typical steps you would take to do a partition exchange involve the following:
1. Create the offline tables.
See EMC Documentum High-Volume Server Development Guide for an example of the tables to create.
2. Load the offline tables.
Load the tables with data using whatever methods are available to you with your database.
3. Create the offline index tables.
The offline tables must index the same properties as the online objects do. The schema must be
identical. Create the offline index tables in the same tablespace as the current online indexes.
4. Exchange the partition for the offline tables.
Run the PARTITION_OPERATION administration method to switch in or switch out the data.
After following these steps, the data that was previously in the offline tables is in the online
partition, and the previously online data is now in the offline table.
5. Validate the exchange.
Check the online data to be sure that the exchange was successful.
More details about switching data into a partition are covered in EMC Documentum High-Volume
Server Development Guide.

PING
PING
Purpose
Determines if the client still has an active server connection.
Syntax
dmAPIGet("apply,c,NULL,PING[,RETRY_ON_TIMEOUT,B,T|F]")
You cannot use the EXECUTE statement to invoke PING.
Arguments
Table 73. PING arguments
RETRY_ON_
TIMEOUT
B T (TRUE) or F
(FALSE)
T (TRUE) forces a connection
attempt between the client and the
server. The default is F.
Return value
PING returns a collection with one query result object. The object has one property, called result, that
is TRUE if the connection is alive. If the connection has timed out, a null collection is returned.
Permissions
Description
None
None
Examples
dmAPIGet("apply,c,NULL,PING,RETRY_ON_TIMEOUT,B,T")

PURGE_AUDIT
PURGE_AUDIT
Purpose
Removes an audit trail entry from the repository.
Syntax
EXECUTE purge_audit WITH delete_mode='mode'
[,date_start='start_date',date_end='end_date']
[,id_start='start_id',id_end='end_id']
[,object_id='object_id'][,dql_predicate='predicate']
[,purge_non_archived=TRUE|FALSE][,purge_signed=TRUE|FALSE][,commit_size=value]
Arguments
Table 74. PURGE_AUDIT arguments
DELETE_MODE S mode Defines how the entries to be
deleted are chosen. Valid values
are:
• DATE_RANGE
Deletes all audit trail entries
generated within a specified
date range. If you include this,
include the DATE_START and
DATE_END arguments.
• ID_RANGE
whose object IDs fall within a
specified range. If you include
this, include the ID_START and
ID_END arguments.
• ALL_VERSIONS
whose audited_obj_id value
corresponds to a specified
object ID. ALL_VERSIONS is
valid only for audit trail entries
whose audited_obj_id identifies
a SysObject or SysObject
subtype. If you include
ALL_VERSIONS, include the
OBJECT_ID argument.

PURGE_AUDIT
• SINGLE_VERSION
whose audited_obj_id value
corresponds to a specified
object ID. If you include
SINGLE_VERSION, include
the OBJECT_ID argument.
• AUDIT_RECORD
Deletes the audit trail entry
whose object ID corresponds
to a specified object ID. If you
include AUDIT_RECORD,
include the OBJECT_ID
argument.
• PREDICATE
that satisfy a DQL predicate.
If you include PREDICATE,
include the DQL_PREDICATE
argument.
DATE_START S start_date The starting date of the entries to
be deleted. The date is compared
to the value in the time_stamp
property in the audit trail entry.
(time_stamp records the local time
at which the entry was generated.)
The date format can be any
acceptable format that does not
require a pattern specification.
(Refer to Date literals, page 15, for
a list of valid formats.
If you include DATE_START, you
must include END_DATE also.
Include a start and end date
only when DELETE_MODE is
DATE_RANGE.

PURGE_AUDIT
DATE_END S end_date The ending date of the entries to
be deleted. The date is compared
to the value in the time_stamp
property in the audit trail entry.
(time_stamp records the local time
at which the entry was generated.)
The date format can be any
acceptable format that does not
require a pattern specification.
(Refer to Date literals, page 15, for
a list of valid formats.
If you include END_DATE, you
must also START_DATE. The end
date must be greater than the start
date.
Include a start and end date
only when DELETE_MODE is
DATE_RANGE.
ID_START S start_id The object ID of an audit trail
entry.
Include ID_START if the
DELETE_MODE is ID_RANGE.
You must also include ID_END.
ID_END S end_id The object ID of an audit trail
entry.
Include ID_END if the
DELETE_MODE is ID_RANGE.
You must also include ID_START.
The ID_END object ID must be
larger than the object ID identified
in ID_START
OBJECT_ID S object_id Include this argument if
DELETE_MODE is ALL_
VERSIONS, SINGLE_VERSION,
or AUDIT_RECORD.
For all modes except
AUDIT_RECORD, object_id is
the object ID recorded in the
audited_obj_id property of the
audit trail entries.

PURGE_AUDIT
For AUDIT_RECORD mode,
object_id is the object ID of an audit
trail entry.
PURGE_NON_
ARCHIVED
B T (TRUE) or F
(FALSE)
Determines whether unarchived
audit trial entries are deleted.
Setting this argument to T
(TRUE) directs Content Server to
delete audit trail entries whose
i_is_archived property is set to F
(FALSE).
The default for this argument is
F (FALSE), meaning that only
entries whose i_is_archived
property is set to T (TRUE) are
removed.
You cannot set this to T if the
DQL_PREDICATE argument is
included.
PURGE_SIGNED B T (TRUE) or F
(FALSE)
Determines whether audit trail
entries for dm_adddigsignature
and dm_addesignature events are
considered for deletion.
Setting this argument to T
(TRUE) means entries for
dm_adddigsignature and
dm_addesignature events are
The default is F (FALSE), meaning
the dm_adddigsignature and
dm_addesignature entries are not
You cannot set this to T if the
DQL_PREDICATE argument is
included.

PURGE_AUDIT
DQL_PREDICATE S predicate Defines a DQL predicate to choose
the audit trail entries to be deleted.
A valid predicate is that portion
of a DQL SELECT statement that
occurs after the FROM keyword.
If you include this argument,
you may not include
PURGE_NON_ARCHIVED
or PURGE_SIGNED.
COMMIT_SIZE I value Defines how many audit trail
entries to delete in each transaction
within the overall operation. The
default is 1000.
Setting this to 0 means that
all entries are deleted in one
transaction.
Return value
The PURGE_AUDIT method returns a collection with one query result object. The query result
object has two properties, result and deleted_objects. The result property is set to T if the method
completed successfully and F if the method did not complete successfully. deleted_objects records
the number of audit trail entries that were deleted.
Permissions
You must have Purge Audit privileges to execute this method.
Description
Executing the PURGE_AUDIT method always generates at least one audit trail entry with the event
name dm_purgeaudit. The operation generates one audit trail entry for each transaction within the
operation. For example, if you set COMMIT_SIZE to 100 and the operation deletes 700 audit trail
entries, the operation generates 7 audit trail entries, one for each transaction.
The entry for each transaction has the event name dm_purgeaudit. The optional properties record
the following information for the transaction:
• string_1 stores the time_stamp value of the first audit trail entry deleted in the transaction
• string_2 stores the time_stamp value of the last audit trail entry deleted in the transaction
• string_3 stores the actual number of audit trail entries deleted by the transaction
• string_5 stores the entire list of arguments from the method’s command line
• id_1 stores the object ID of the first audit trail object deleted in the transaction
• id_2 stores the object ID of the last audit trail object deleted in the transaction

PURGE_AUDIT
None
Examples
This example deletes all archived audit trail entries generated from January 1, 2003 to January 1, 2004:
EXECUTE purge_audit WITH DELETE_MODE='DATE_RANGE',
date_start='01/01/2003 00:00:00 AM',
date_end='01/01/2004 00:00:00 AM'
This example deletes all audit trail entries generated from January 1, 2003 to January 1, 2004,
including unarchived entries. The number of entries deleted in each transaction is set to 500:
EXECUTE purge_audit WITH delete_mode='DATE_RANGE',
date_start='01/01/2003 00:00:00 AM',
date_end='01/01/2004 00:00:00 AM',
purge_non_archived=TRUE,commit_size=500
This example deletes all archived audit trail entries that identify the document 09000003ac5794ef
as the audited object:
EXECUTE purge_audit WITH delete_mode='ALL_VERSIONS',
object_id='09000003ac5794ef'
This example deletes the single audit trail entry whose object ID is 5f0000021372ac6f:
EXECUTE purge_audit WITH delete_mode='AUDIT_RECORD',
object_id='5f0000021372ac6f'
This example deletes all audit trail entries whose object IDs range from 5f1e9a8b003a901 to
5f1e9a8b003a925, including unarchived entries:
EXECUTE purge_audit WITH delete_mode='ID_RANGE',
id_start='5f1e9a8b003a901',id_end='5f1e9a8b003a925',
purge_non_archived=TRUE
This example deletes all audit trail entries that satisfy the specified DQL predicate:
EXECUTE purge_audit WITH delete_mode='PREDICATE',
dql_predicate=
'dm_audittrail where event_name like ''dcm%''and
r_gen_source=0'
If you need to include single-quotes in the predicate string, (for example, ’dcm%’), escape the
single-quotes with single-quotes. This is illustrated in the example above.

PURGE_CONTENT
PURGE_CONTENT
Purpose
Sets a content file off line and deletes the file from its storage area.
Syntax
EXECUTE purge_content FOR 'content_object_id'
Arguments
PURGE_CONTENT has no arguments.
Return value
PURGE_CONTENT returns a collection with one query result object. The object has one Boolean
Permissions
You must have Sysadmin or Superuser user privileges to use this method.
Description
PURGE_CONTENT marks a content file as off-line. Additionally, if the file is referenced by only
one content object, it deletes the file from the storage area. If the file is referenced by other content
objects, the file is not deleted. (A file may be referenced by multiple content objects if the storage
area is a file store storage area configured to use content duplication checking and prevention. For
information about that feature, refer to the EMC Documentum Content Server Administration and
Configuration Guide.)
Do not use this method unless you have previously moved the file to an archival or back up storage
area. PURGE_CONTENT does not back up the file. The method only deletes the file from the storage
area and sets the is_offline property of the file’s content object to TRUE.
To use PURGE_CONTENT, you must be using one server for both data and content requests. If the
content server to use PURGE_CONTENT in the session.
Examples
This example deletes the content file associated with the content object represented by
060000018745231c and set the is_offline property in the associated content object to TRUE:
EXECUTE purge_content FOR '060000018745231c'

PUSH_CONTENT_ATTRS
PUSH_CONTENT_ATTRS
Purpose
Updates content metatdata in a content addressable systems.
Syntax
EXECUTE push_content_attrs FOR object_id
WITH format=format_name [,page=number]
[,page_modifier=value]
object_id is the ID of a document or other SysObject.
Arguments
Table 75. PUSH_CONTENT_ATTRS arguments
FORMAT S name The file format of the content file.
Use the name of the format object
that defines the format.
PAGE I number The page number of the content
file in the document. If the
content file is a rendition, this is
the page number of the primary
content with which the rendition
is associated.
The default value is 0.
PAGE_MODIFIER S value This identifies a rendition. It is
the page modifier defined for the
rendition when the rendition was
added to the document. The value
is stored in the page_modifier
property in the content object.
Return value
PUSH_CONTENT_ATTRS returns a collection with one query result object. The object has one
Boolean property that contains TRUE if the method was successful and FALSE if unsuccessful.

PUSH_CONTENT_ATTRS
Permissions
To execute this method you must be a superuser and you must have at least Write permission on the
object identified in object_id.
Description
PUSH_CONTENT_ATTRS updates the metadata fields in a content-addressed storage system for a
particular content file. First, the method reads the values in the following content-related properties
in the file’s associated content object:
• content_attr_name
• content_attr_value
• content_attr_data_type
• content_attr_num_value
• content_attr_date_value
content_attr_name identifies the metadata fields to be set in the storage system. The
content_attr_data_type property identifies the data type of the value users must provide for each
metadata field in the corresponding index position in content_attr_name. The actual value is defined
in one of the remaining properties, depending on the field’s datatype. For example, if the field name
is "title” and its datatype is character string, the content_attr_value property contains the actual title
value. (The remaining properties, content_attr_num_value and content_attr_date_value are set to
the default NULLINT and NULLDATE values.)
After the method reads the content object properties, it invokes the content-addressed plugin library
to update the storage system metadata fields.
Note: To be updated, a field must be defined in both the content object’s content_attr_name property
and in the storage object’s a_content_attr_name property. If a field is named in the content object’s
content_attr_name property but not defined in the storage object’s a_content_attr_name object, it is
not set in the storage area. Similarly, if a field is named in the storage object’s a_content_attr_name
but not in the content object’s content_attr_name property, it is not set in the storage area.
If PUSH_CONTENT_ATTRS completes successfully, it generates a new content address for the
content. The new address is appended to the i_contents property of the subcontent object that records
content addresses for the content file.
For complete information about how to save a document to content-addressed storage, refer to EMC
Documentum Content Server Fundamentals Guide.
SET_CONTENT_ATTRS, page 313
Examples
EXECUTE push_content_attrs FOR 090000026158a4fc
WITH format=txt,page=1

RECOVER_AUTO_TASKS
RECOVER_AUTO_TASKS
Purpose
Recovers work items that have been claimed, but not yet processed, by a workflow agent associated
with a failed Content Server.
Syntax
EXECUTE recover_auto_tasks
WITH server_config_name=name
Arguments
Table 76. RECOVER_AUTO_TASKS arguments
SERVER_CONFIG_
NAME
I name of the server
config object
Use the name of the server config
object representing the Content
Server that crashed or failed.
Return value
The method returns T (TRUE) if it completes successfully and F (FALSE) if not.
Permissions
This method must be executed by a user with Sysadmin or Superuser privileges.
Description
If a Content Server fails, its workflow agent is stopped also. When the server is restarted, the
workflow agent will recognize and process any work items it had claimed but not processed before
the failure. However, if you cannot restart the Content Server, you must recover those work items
already claimed by its associated workflow agent so that another workflow agent can process them.
RECOVER_AUTO_TASKS performs that recovery.
Running RECOVER_AUTO_TASKS executes a query to update all work items claimed but
unprocessed by the failed server’s workflow agent. The query resets the a_wq_name property in
the work items to empty. This allows other workflow agents running against the repository to
claim those work items for processing.
Before executing this method, make sure that the specified Content Server is not running.

RECOVER_AUTO_TASKS
None
Examples
EXECUTE recover_auto_tasks
WITH server_config_name='DevRepository_1'

REGISTER_ASSET
REGISTER_ASSET
Purpose
Queues a request to the Media Server to generate a thumbnail, proxies, and metadata for a rich
media content file.
Syntax
EXECUTE register_asset FOR object_id
WITH [page=page_number][,priority=priority_level]
Arguments
Table 77. REGISTER_ASSET arguments
PAGE I page_number Page number of the content file. If
unspecified, the default is 0.
PRIORITY I priority_level Identifies the priority of the
request. A value of zero indicates
no priority. Priority rises as the
value rises. The Media Server
processes higher priority requests
before lower priority requests.
If unspecified, the default is 5.
Return value
REGISTER_ASSET returns a collection with one query result object. The object has properties, result
and result_id. The result property is a Boolean property whose value indicates success (TRUE) or
failure (FALSE) of the operation. The result_id property records the object ID of the newly created
queue item object.
Permissions
You must have at least Version permission on the object or Sysadmin privileges to use this method.
Description
REGISTER_ASSET is called by Content Server whenever an object with rich media content is saved
or checked in to the repository. The method generates an event, dm_register_event, that is queued

REGISTER_ASSET
to the Media Server. When the Media Server processes the event, the server creates a thumbnail,
proxies, and metadata for the content.
The dmi_queue_item that represents the event is queued to the dm_mediaserver user, who represents
the Media Server. (dm_mediaserver is created when Documentum Media Transformation Services is
installed.) REGISTER_ASSET sets the following properties in the queue item:
Table 78. Queue item property values set by REGISTER_ASSET
Property Set to
event dm_register_event
item_id object ID of the SysObject that triggered the request
instruction_page page number of the content in the SysObject
date_sent date the request was made
name dm_mediaserver
sent_by session user
priority priority level identified in the REGISTER_ASSET call
SET_CONTENT_ATTRS, page 313
TRANSCODE_CONTENT, page 324
Example
EXECUTE register_asset FOR 090002410063ec21
WITH page=0,priority=8

REINDEX_PARTITIONABLE_TYPE
Purpose
Reindex a partitionable type to increase performance.
Syntax
EXECUTE reindex_partitionable_type [WITH type_name='type_name']
Arguments
Table 79. REGISTER_ASSET arguments
type_name S type_name Type to reindex.
Return value
The method returns T (TRUE) if it completes successfully and F (FALSE) if not
Permissions
This method must be executed by a user with Sysadmin or Superuser privileges.
Description
REINDEX_PARTITIONABLE_TYPE is used for databases upgraded from release 6.0 or 6.0 SP1 to
6.5. In 6.0, indexes for types that have the i_partition attribute included a column for that property
in indexes that included r_object_id. This can cause a performance problem for non-partitioned
repositories. In 6.5 databases, these indexes do not include the i_partition value unless the repository
is partitioned. Only use this method for repositories upgraded to 6.5 that are showing performance
issues.
You can reindex all the partitionable types, or any specific partitionable type.
None.
Example
EXECUTE reindex_partitionable_type WITH "type_name"='my_type'

REORGANIZE_TABLE
REORGANIZE_TABLE
Purpose
Reorganizes a database table for query performance.
Syntax
EXECUTE reorganize_table WITH table_name='name',
[,index='index_name']
Arguments
Table 80. REORGANZE_TABLE arguments
TABLE_NAME S name Name of the RDBMS table to be
reorganized.
INDEX S index_name Name of an index on the table
identified in TABLE_NAME.
This argument is required for
an Oracle database table. It is
optional for a SQL Server or DB2
database table.
The argument is used differently
on each database. Refer to the
General Notes for details.
Return value
REORGANIZE_TABLE returns a collection with one query result object. The object has one property,
named result, that contains T if the method was successful or F if the method was unsuccessful.
Permissions
You must be a superuser to execute this method.
Caution: Do not use this method unless directed to do so by Technical Support.

REORGANIZE_TABLE
Description
This method is used by the UpdateStatistics administration tool, in conjunction with
UPDATE_STATISTICS, when the tool is run on a DB2 database. However, the method can be used
on any supported database.
The INDEX argument
On Oracle, you must include the INDEX argument. The method rebuilds the specified index. The
index must be an index on the table identified in TABLE_NAME.
For SQL Server, the argument is optional. If you include it, you must specify an index on the table
identified in TABLE_NAME. The method rebuilds that index. If you do not include the argument,
the method rebuilds all indexes on the specified database table.
For DB2, the argument is optional. If you include it, you must specify an index on the table identified
in TABLE_NAME. The method rebuilds the database table based on the ordering defined in the index.
Caution: Using the INDEX argument on a DB2 database table is a powerful feature. The choice
of index to use as a basis for re-ordering the table can greatly affect query performance against
that table— for better or worse, depending on the index choice. It is recommended that if you
use this argument, specify the index on the r_object_id property.
Note: Indexes created by Content Server are named using the following format: D_index_obj_id,
where index_obj_id is the object ID of the dmi_index object for the index.
UPDATE_STATISTICS, page 327
Examples
This example reorganizes the dm_sysobject_s table. (Note that these two examples do not work on
Oracle because the required INDEX argument is not included.)
EXECUTE reorganize_table
WITH table_name='dm_sysobject_s'
This example rebuilds an index associated with the dm_sysobject_s table.
EXECUTE reorganize_table
WITH table_name='dm_sysobject_s,
index='D_1f0C9fda80000108'

REPLICATE
REPLICATE
Purpose
Copies content files in distributed storage areas.
Syntax
EXECUTE replicate
WITH query='value',store='value'
[,type='value']
Arguments
Table 81. REPLICATE arguments
QUERY S dql_predicate
expression
This argument is required. The
predicate expression is used to
build a DQL query that selects the
objects whose content you want to
copy. The expression can be any
expression that would be a valid
WHERE clause qualification (refer
to The WHERE clause, page 142).
STORE S name This argument is required. It
identifies where to put the new
content copy or copies. The name
must be the name of a component
of a distributed storage area.
TYPE S type_name This argument is optional. It
identifies the type of objects
whose content you are replicating.
type_name must be a direct or
indirect subtype of dm_sysobject.
The default is dm_sysobject.
Return value
REPLICATE returns a collection with one query result object. The object has one Boolean property

REPLICATE
Permissions
You must have Sysadmin or Superuser privileges to use REPLICATE.
Description
REPLICATE copies content files from one component of a distributed storage area to another.
Typically, replication of content files is performed by the ContentReplication or surrogate get. Use
REPLICATE as a manual backup for these tools. (Refer to the EMC Documentum Content Server
Administration and Configuration Guide for information about the ContentReplication and to the EMC
Documentum Content Server Distributed Configuration Guide for information about surrogate get.)
REPLICATE checks the contents stored in the storage area at a specified site and copies back to the
local storage area any contents that meet the conditions defined in the function. (Refer to the EMC
Documentum Content Server Distributed Configuration Guide for more information.)
To use REPLICATE, you must be using one server for both data and content requests. If the
content server to use REPLICATE in the session.
DELETE_REPLICA , page 188
IMPORT_REPLICA, page 228
Examples
This example replicates all content for documents of the subtype proposals owned by jennyk:
EXECUTE replicate WITH query='owner_name=jennyk',
store='distcomp_1',type='proposals'
The next example replicates the content of all objects whose content is authored by Jane:
EXECUTE replicate WITH query='any authors in (''Jane'')',
store='diststorage3'

RESET_TICKET_KEY
RESET_TICKET_KEY
Purpose
Generates a login ticket key and stores it in the repository.
Syntax
EXECUTE reset_ticket_key
Arguments
None
Return value
The method returns a collection with one query result object. The object has one property, result, that
contains T (TRUE) if the method was successful or F (FALSE) if the method was unsuccessful.
Permissions
Description
Use this method when you need to replace a login ticket key in a repository with a new key. Any login
tickets generated by the repository before the LTK was reset cannot be used within the repository.
You must restart Content Server after resetting the login ticket key.
EXPORT_TICKET_KEY, page 205
IMPORT_TICKET_KEY, page 230
IDfSession.resetTicketKey()
Examples
EXECUTE reset_ticket_key

RESTORE_CONTENT
RESTORE_CONTENT
Purpose
Restores an off-line content file to its original storage area.
Syntax
EXECUTE restore_content FOR 'content_object_id'
WITH file = 'path_name' [,other file = 'other_path_name']
content_obj_id is the object ID of the content object associated with the specified file.
Arguments
Table 82. RESTORE_CONTENT arguments
FILE S file_path Specifies the current location of
the content file. Use a full path
specification.
OTHER_FILE S other_path_name Specifies the current location of
the resource fork for the content
file. Use a full path specification.
Include this argument only if the
file is a Macintosh file.
Return value
RESTORE_CONTENT returns a collection with one query result object. The object has one Boolean
Permissions
You must have Sysadmin or Superuser user privileges to use this method.
General notes
RESTORE_CONTENT restores an off-line content file to its original storage area. This method only
operates on one file at a time. To restore multiple files, use an IDfSession.restore method.
The method places the file in the storage area indicated by the storage_id property of the file’s content
object and sets the content object’s is_offline property to FALSE.

RESTORE_CONTENT
To use RESTORE_CONTENT, you must be using one server for both data and content requests. If
the content server to use RESTORE_CONTENT in the session.
Examples
The following examples restore the file named Myproposal using an EXECUTE statement:
EXECUTE restore_content
WITH file='c:archive1Myproposal'
or, on UNIX:
EXECUTE restore_content
WITH file='u02/archive1/Myproposal'

ROLES_FOR_USER
ROLES_FOR_USER
Purpose
Retrieves the roles assigned to the user in a particular client domain.
Syntax
EXECUTE roles_for_user [[FOR] 'dm_user_id']
WITH [USER_NAME=value][,DOMAIN=domain_name]
Arguments
Table 83. ROLES_FOR_USER arguments
USER_NAME S user_name User whose roles you are
retrieving. Identify the user by the
user’s user name.
DOMAIN S domain_name Domain of a client application.
If a domain is included, the
method returns the user’s roles
within the specified domain.
If a domain is not included, the
method returns all roles for the
user, regardless of domain.
Return value
This returns a collection of one query result object that has three properties: user_name, domain, and
roles. user_name contains the name of the user being queried. domain lists the domains searched for
user roles. roles is a repeating property that contains the roles for the user.
Permissions
Any one can use this method.
General notes
This method is used by client applications to determine the roles assigned to users who use the
applications. For example, when a user starts a session with Desktop Client, DTC executes this

ROLES_FOR_USER
method to query the domain group associated with Desktop Client, to determine what roles the
user has in Desktop Client.
(For more information about groups, roles, and domains, refer to EMC Documentum Content Server
Fundamentals Guide.)
None
Examples
EXECUTE roles_for_user
WITH user_name=MaryJean,domain="HR_app"

SET_APIDEADLOCK
SET_APIDEADLOCK
Purpose
Sets a deadlock trigger on a particular operation.
Syntax
EXECUTE set_apideadlock
WITH API=api_name,VALUE=TRUE|FALSE{,API=operation_name,VALUE=TRUE|FALSE}
Arguments
Table 84. SET_APIDEADLOCK arguments
API S api_name Name of the operation on which
to set the deadlock trigger.Table
85, page 311 lists the operations
on which you can set a trigger.
VALUE BOOLEAN T (TRUE) or F
(FALSE)
TRUE sets the trigger. FALSE
removes the trigger.
Return value
SET_APIDEADLOCK returns a collection with one query result object. The object has one Boolean
Permissions
Anyone can use this administration method.
General notes
Use SET_APIDEADLOCK to test deadlock retry code in client applications that execute in explicit
transactions. Operations that occur in explicit transactions are not protected by Content Server’s
internal deadlock retry functionality. Consequently, applications that execute in an explicit
transaction may include their own deadlock retry functionality.
To test the deadlock retry functionality, use SET_APIDEADLOCK to place a trigger on one or more
methods or operations executed in the application. When the application is tested, Content Server
simulates a deadlock when one of the methods or operations executes, allowing you to test the
deadlock retry code in the application. The simulated deadlock sets the _isdeadlocked computed
property and issues a rollback to the database.

SET_APIDEADLOCK
The deadlock trigger is removed automatically from the method or operation which triggered the
simulated deadlock.
Table 85, page 311, lists the operations and methods on which you can place a deadlock trigger.
Table 85. Valid operation names for SET_APIDEADLOCK
Operation Representing
acquire Acquisition of a work item
bp_transition Completion of the bp_transition method
branch Branching of a version tree
checkin A checkin operation
complete Completion of a work item
demote Demotion of an object
dequeue Dequeuing an object
destroy Destruction of an object
exec Any RPC EXEC call (on behalf of a Query, Readquery,
Execquery, or Cachequery operation)
next A next method
promote Promotion of an object
queue A queue method
resume Resumption of an object
revert A revert method
save Save on any object
Use this operation to put a deadlock trigger on a
Saveasnew method for a SysObject.
save_content A content save operation during a save, checkin,
saveasnew, or branch operation
save_parts A containment save operation during a save, checkin,
saveasnew, or branch operation
suspend Suspension of an object
None
Examples
This example sets a deadlock trigger on the Checkin method:
WITH api='checkin',value=T

SET_APIDEADLOCK
This example sets a deadlock trigger on the Promote and Revert methods:
WITH api='promote',value=T,
api='revert',value=T
This example removes the deadlock trigger from the Checkin method:
WITH api='checkin',value=F

SET_CONTENT_ATTRS
SET_CONTENT_ATTRS
Purpose
Sets the content-related properties of a content object.
Syntax
EXECUTE set_content_attrs FOR object_id
WITH format='format_name',[page=page_number,]
[page_modifier='value',]
parameters='name="value"{,name="value"}'[,truncate=TRUE|FALSE]
Arguments
Table 86. SET_CONTENT_ATTRS arguments
FORMAT S format_name Name of the content format. This
is the name of the format object.
PAGE I page_number Page number of the content in the
document’s set of content files. If
PAGE_ MODIFIER S value Identifies the rendition. Refer to
the General Notes for a detailed
description of the purpose of the
page modifier.
PARAMETERS S name = value pairs Comma separated list of property
names and values. value is one of:
"character_string"
FLOAT(number)
DATE(date_value)
Refer to the General Notes for
examples.
TRUNCATE B T (TRUE) or F
(FALSE)
Whether to remove existing values
in the content properties before
setting the new values. The
default value is F.

SET_CONTENT_ATTRS
Return value
SET_CONTENT_ATTRS returns a collection with one query result object. The object has one Boolean
Permissions
You must have at least Write permission on the object or Sysadmin privileges to use this method.
General notes
SET_CONTENT_ATTRS sets five properties of a content object:
• content_attr_name
• content_attr_value
• content_attr_num_value
• content_attr_date_value
• content_attr_data_type
The Media Server uses this method to store the metadata it generates for a content file in the file’s
content object.
Using the PARAMETERS argument
The metadata is specified in SET_CONTENT_ATTRS as a comma-separated list of name and value
pairs in the PARAMETERS argument. The format is:
'name=value{,name=value}'
Because the PARAMETERS argument is a string argument, enclose the full string in single quotes.
Additionally, if value is a character string, enclose the individual value in double quotes. For example:
EXECUTE set_content_attrs FOR 0900038000ab4d13
WITH format='jpeg',page=0,
PARAMETERS='name="photo_closeup"'
If the metadata value is a numeric value, use the FLOAT function to specify the value:
PARAMETERS='width=FLOAT(12.5)'
If the metadata value is a date, use the DATE function to specify the value:
PARAMETERS='take_down_date=DATE(09/30/2002)'
The method sets the content object properties with values specified in the PARAMETERS argument
beginning at the zero index position. The values are set in the order in which they are included in the
PARAMETERS argument.
The content_attr_name and content_attr_data_type values are set to the name of the property
and its datatype. If the property’s datatype is string, the value is stored in content_attr_value and
the remaining two properties (content_attr_date_value and content_attr_num_value) are set to the
default NULL values (NULLDATE and NULLINT). If the property’s datatype is numeric, the value
is stored in content_attr_num_value and the remaining two properties (content_attr_value and

SET_CONTENT_ATTRS
content_attr_date_value) are set to the default NULL values (NULLSTRING and NULLDATE). If the
property’s datatype is date, the value is stored in content_attr_date_value and the remaining two
properties (content_attr_num_value and content_attr_value) are set to the default NULL values
(NULLINT and NULLSTRING).
For example, suppose an application executes the following statement:
PARAMETERS='name="photo_closeup",width=FLOAT(12.5),take_down_date=DATE(09/30/2002)'
Table 87, page 315, shows the resulting values in the properties in the content object.
Table 87. Example settings for content metadata properties in content objects
Property Index position
[0] [1] [2]
content_attr_name name width take_down_date
content_attr_data_
type
2 4 5
content_attr_value photo_closeup NULLSTRING NULLSTRING
content_attr_num_
value
NULLINT 12.5 NULLINT
content_attr_date_
value
NULLDATE NULLDATE 09/30/2002
The TRUNCATE argument controls how existing values in the content properties are handled. If
TRUNCATE is set to T (TRUE), existing values in the properties are removed before the properties
are set to the new names and values. If TRUNCATE is F (FALSE), existing names and values are not
removed. If the PARAMETERS argument includes a name that already present in the properties, the
name’s value is overwritten. Names specified in the PARAMETERS argument that are not currently
present in the properties are appended to the properties. The default for TRUNCATE is F.
Using PAGE_MODIFIER
Use the PAGE_MODIFIER argument to define an identifier that distinguishes a rendition from any
other rendition in the same format associated with a particular content page.
There are no constraints on the number of renditions that you can create for a document. Additionally,
you can create multiple renditions in the same format for a particular document. To allow users or
applications to distinguish between multiple renditions in the same format for a particular document,
define a page modifier.
Including the PAGE_MODIFIER argument in SET_CONTENT_ATTRS sets the page_modifier
property of the content object. This property, along with three others (parent_id, page, i_format)
uniquely identifies a rendition. Applications that query renditions can use the modifier to ensure
that they return the correct renditions.
PUSH_CONTENT_ATTRS, page 294

SET_CONTENT_ATTRS
Examples
EXECUTE set_content_attrs FOR 090002134529cb2e
parameters='name="garage_band",sales=FLOAT(100.2),
release_date=DATE(01/02/2002)'

SET_OPTIONS
SET_OPTIONS
Purpose
Turns tracing options off or on.
Syntax
EXECUTE set_options
WITH option='option_name',"value"=true|false
Arguments
Table 88. SET_OPTIONS arguments
OPTION S option_name Identifies the tracing option you
want to turn on or off. Refer to
Table 89, page 318
VALUE B T (TRUE) or F
(FALSE)
TRUE turns tracing on. FALSE
turns tracing off.
Return value
SET_OPTIONS returns a collection with on result object. The object has one Boolean property whose
value indicates the success (TRUE) or failure (FALSE) of the operation.
Permissions
General notes
The tracing results for the options representing server operations are printed to the server log file.
The trace_method_server option traces method server operations. The tracing results generated by
setting trace_method_server are printed to the method server log file.
To turn off the nettrace option, disconnect all client sessions. After about one minute, the tracing
turns off.
Table 89, page 318, lists the values for the OPTION argument:

SET_OPTIONS
Table 89. Trace options for SET_OPTIONS
Value for the OPTION argument Meaning
ca_store_trace Enables tracing for operations in a content-addressed storage
system. The trace messages, up to 2047 characters, are placed
in the server log file.
Tracing content-addressed storage operations is only
recommended when needed to troubleshoot the system.
clean Removes the files from the server common area during startup
mode.
debug Traces session shutdown, change check, launch and fork
information
docbroker_trace Traces connection broker information
file_store_trace Traces digital shredding of content files. Also traces content
checking and duplication prevention operations.
i18n_trace Traces client session locale and code page.
An entry is logged identifying the session locale and client
code page whenever a session is started. An entry is also
logged if the locale or code page is changed during the session.
last_sql_trace Traces the SQL translation of the last DQL statement issued
before access violation and exception errors.
If an error occurs, the last_sql_trace option causes the server
to log the last SQL statement that was issued prior to the error.
This tracing option is enabled by default.
It is strongly recommended that you do not turn off this
option. It provides valuable information to Technical Support
if it ever necessary to contact them.
lock_trace Traces Windows locking information
net_ip_addr Traces the IP addresses of client and server for authentication
nettrace Turns on RPC tracing. Traces Netwise calls, connection ID,
client host address, and client host name
retention_trace Turns on tracing for the following retention-related operations:
• Assigning a retention policy to a document
• Recording a retention period applied by a retention policy
at the storage area level
• Removing a retention policy from a document
rpctrace Traces RPC calls.
sqltrace Traces SQL commands sent to the underlying RDBMS for
subsequent sessions, including the repository session ID and
the database connection ID for each SQL statement.

SET_OPTIONS
Value for the OPTION argument Meaning
ticket_trace Traces import and export operations for the login ticket key
and operations the use a single-use login ticket.
trace_authentication Traces detailed authentication information
trace_complete_launch Traces UNIX process launch information
trace_method_server Traces the operations of the method server.
trace_workflow_agent Traces operations of the workflow agent. Messages are
recorded in the server log file.
LOG_OFF, page 242
LOG_ON, page 243
MODIFY_TRACE, page 273
Examples
This example turns on SQL tracing:
EXECUTE set_options
WITH option='sqltrace',"value"=true
The following example turns on logon authentication tracing:
EXECUTE set_options
WITH option='trace_authentication',
"value"=true
This example turns off tracing for logon authentication:
EXECUTE set_options
WITH option='trace_authentication",
"value"=false

SET_STORAGE_STATE
SET_STORAGE_STATE
Purpose
Takes a storage area offline, places an off-line storage area back online, or makes a storage area
read-only.
Syntax
EXECUTE set_storage_state [[FOR] 'storage_area_obj_id']
[WITH store=storage_area_name{,argument=value}]
If you include the STORE argument, do not include the FOR clause.
Arguments
Table 90. SET_STORAGE_STATE arguments
OFFLINE B T (TRUE) or F
(FALSE)
If set to TRUE, the storage area is
taken off-line.
READONLY B T (TRUE) or F
(FALSE)
If set to TRUE, the storage area is
read-only.
STORE S storage_area_name Identifies the storage area whose
state you are changing. Use the
name of the area’s storage area
object.
Return value
SET_STORAGE_STATE returns a collection with one result object. The object has one Boolean
Permissions
General notes
A storage area has three possible states:
• Online
Users can read from and write to an online storage area.
• Offline

SET_STORAGE_STATE
Users can neither read from nor write to an offline storage area.
• Readonly
Users can only read from a readonly storage area. Users cannot write to a readonly storeage area.
To set a storage area’s state to offline, issue SET_STORAGE_STATE with the OFFLINE argument
set to T (TRUE); for example:
EXECUTE set_storage_state WITH store=filestore_33,OFFLINE=true
To set a storage area’s state to readonly, issue SET_STORAGE_STATE with the READONLY argument
set to T (TRUE); for example:
EXECUTE set_storage_state WITH store=filestore_33,READONLY=true
To change either an offline or readonly storage area back to the online state (users can read and write
the storage area), issue the SET_STORAGE_STATE method without specifying a state argument;
for example:
EXECUTE set_storage_state WITH store=filestore_33
Setting either OFFLINE or READONLY to FALSE directly has no effect.
To use SET_STORAGE_STATE, you must be using one server for both data and content requests. If
the content server to use SET_STORAGE_STATE in the session.
None
Example
The following example moves the storage area called manfred offline:
EXECUTE set_storage_state
WITH store = 'manfred', offline = T

SHOW_SESSIONS
SHOW_SESSIONS
Purpose
Lists information about all the currently active sessions and a user-specified number of historical
sessions.
Syntax
EXECUTE show_sessions
Arguments
SHOW_SESSIONS has no arguments.
Return value
SHOW_SESSIONS returns a collection. Each query result object in the collection represents one
currently active or inactive repository session. The properties of the objects contain information about
the sessions. The properties returned by SHOW_SESSIONS are the same as those for LIST_SESSIONS.
For a description of the properties, refer to Table 63, page 237.
Permissions
General notes
SHOW_SESSIONS returns information about currently active sessions and historical (timed out)
sessions. The information for each session is identical to the information returned by LIST_SESSIONS.
The maximum number of historical sessions returned by SHOW_SESSIONS is determined by the
parameter history_sessions in the server’s startup file (the server.ini file). There is also a startup
parameter, called history_cutoff, to define a cutoff time for the history sessions. For example, if
history_cutoff is set to 15 minutes, then SHOW_SESSIONS will not return any historical sessions
older than 15 minutes.
For a description of the server.ini file and the parameters you can set in it, refer to the EMC
Documentum Content Server Administration and Configuration Guide.
Note: You can use SHOW_SESSIONS to obtain the process ID if you need to shutdown or kill a
session or server.

SHOW_SESSIONS
Examples
Refer to the syntax description for examples.

TRANSCODE_CONTENT
TRANSCODE_CONTENT
Purpose
Queues a transformation request for a content file to the Media Server.
Syntax
EXECUTE transcode_content FOR 'object_id'
WITH [page=page_number,] [priority=priority_level,]
message='message',source_format='format_name',
target_format='format_name'
Arguments
Table 91. TRANSCODE_CONTENT arguments
PAGE I page_number Page number of the content file. If
PRIORITY I priority_level Identifies the priority of the
request. A value of zero indicates
no priority. Priority rises as the
value rises. The Media Server
processes higher priority requests
before lower priority requests.
If unspecified, the default is 5.
MESSAGE S ’-profile_id=
"object_id”’
Identifies the profile that
contains the definition of the
transformation to be performed
on the content. The profile is
identified by its object ID. You
must enclose the object ID in
double quotes, and the entire
value in single quotes. Refer to the
usage notes for an example.
Refer to the General Notes for a
more detailed description.

TRANSCODE_CONTENT
TARGET_FORMAT S format_name Identifies the format to which to
transform the content. Use the
name of the format as it is defined
in the format’s format object.
SOURCE_FORMAT S format_name Identifies the content file’s starting
format. Use the name of the
format as it is defined in the
format’s format object.
Return value
TRANSCODE_CONTENT returns a collection with one query result object. The object has two
properties, result and result_id. The result property is a Boolean property whose value indicates the
success (TRUE) or failure (FALSE) of the operation. The result_id property records the object ID of
the queue item that is created as by the method.
Permissions
You must have at least Write permission on the object or Sysadmin privileges to use this method.
General notes
TRANSCODE_CONTENT is issued by WebPublisher™ to request a transformation operation on
a content file by the Media Server. The method generates an event, dm_transcode_content, that is
queued to the Media Server. When the Media Server processes the event, the server performs the
transformation on the content as defined in the specified profile.
The dm_queue_item that represents the event is queued to the dm_mediaserver user, who represents
the Media Server. (dm_mediaserver is created when Documentum Media Transformation Services is
installed.) TRANSCODE_CONTENT sets the following properties in the queue item:
Table 92. Queue item properties set by TRANSCODE_CONTENT
Property Set to
event dm_transcode_content
item_id object ID of the SysObject that triggered the request
instruction_page page number of the content in the SysObject
content_type starting format of the SysObject
message value specified in the MESSAGE argument
item_type format to which to transform the content
date_sent date the request was made
name dm_mediaserver

TRANSCODE_CONTENT
Property Set to
sent_by session user
priority priority level identified in the REGISTER_ASSET call
The MESSAGE argument specifies the transformation operation to perform by specifying a profile. A
profile is an XML document, stored in the repository, that defines transformation operations. The
value for a MESSAGE argument has the following format:
'-profile_id="object_id"'
object_id is the object ID of the profile document.
For example, suppose that 090000132400c2e198 is the object ID of a document that you want to
transform and that 09000013240053ae1b is the object ID of a profile containing the definition of the
desired transformation. The following DQL EXECUTE statement generates a request to the Media
Server to perform the transformation:
EXECUTE transcode_content FOR '090000132400c2e198'
WITH page=0,message='-profile_id="09000013240053ae1b"',
source_format='jpeg',target_format='gif'
REGISTER_ASSET, page 298
Example
EXECUTE transcode_content FOR '090000017456ae1c'
WITH page=0,priority=5,
message='-profile_id="090000018125ec2b"',
source_format='jpeg',
target_format='jpeg_lres'

UPDATE_STATISTICS
UPDATE_STATISTICS
Purpose
Updates statistics for RDBMS tables in the repository.
Syntax
EXECUTE update_statistics WITH table_name='name'
[,count=integer][,extra_data=value]
Arguments
Table 93. UPDATE_STATISTICS arguments
TABLE_NAME S name Name of the database table whose
statistics you want to update.
COUNT I integer This argument is available only
if the database is Oracle, or SQL
Server. It is not available on DB2.
The argument is used differently
for each database. Refer to the
EXTRA_DATA S value The argument is used differently
for each database. Refer to the
Return value
UPDATE_STATISTICS returns a collection with one query result object. The object has one property,
called result, that contains T if the method was successful or F if the method was unsuccessful.
Permissions
You must be a Superuser to execute this method.
Caution: Do not use this method unless directed to do so by Technical Support.

UPDATE_STATISTICS
General notes
UPDATE_STATISTICS is used by the UpdateStatistics administration tool when the tool is run on a
DB2 database. However, the method can be used on any supported database.
The COUNT argument
This is an optional argument. It is used differently in Oracle, and SQL Server. It is not used if the
database is a DB2 database.
For Oracle, it defines the number of buckets to use when calculating the histogram statistics. The
valid range for COUNT on an Oracle database is 1 to 254. The default is 75. Setting COUNT to
0 deletes all histogram statistics for the table.
For SQL Server, this argument defines what percentage of table rows to use when calculating the
statistics. For example, if you set this to 50, the method uses one half (50%) of the table rows to
calculate the statistics. The default is 100, meaning all rows are used to calculate the statistics.
The EXTRA_DATA argument
The EXTRA_DATA argument is an optional argument. Its use differs depending on the database.
For Oracle, this identifies a specific set of table columns to be analyzed for statistics. The columns
must exist in the table. The columns are specified as a comma-separated list of column names.
Enclose the entire list in single quotes. For example:
EXECUTE update_statistics
WITH table_name='dm_sysobject_s',
extra_data='keywords,authors'
If you do not include EXTRA_DATA, all columns in the table are used.
For SQL Server and DB2, the argument identifies an index for which you want to generate statistics.
The index must be an index on the table identified in TABLE_NAME. If you do not include the
argument, the database table identified in TABLE_NAME and all of its indexes are analyzed.
REORGANIZE_TABLE, page 301
Examples
This example updates statistics for the dm_sysobject_s table:
WITH table_name='dm_sysobject_s'
This example updates the statistics for the dm_user_s table, specifying that only the user_name,
user_os_name and user_address columns be used for the analysis:
WITH table_name='dm_user_s',
extra_data='user_name,user_os_name,user_address'

WEBCACHE_PUBLISH
WEBCACHE_PUBLISH
Purpose
Invokes the dm_webcache_publish method to publish documents to a Web site. (This administration
method cannot be run using the EXECUTE statement.)
Syntax
apply,session,webc_config_obj_id,
WEBCACHE_PUBLISH[,ARGUMENTS,S,argument_list]"
Arguments
Table 94. WEBCACHE_PUBLISH arguments
ARGUMENTS S argument _list Identifies the arguments
that you want to pass to the
dm_webcache_publish method.
The valid entries for argument_list
are described in Table 95, page
329.
Table 95, page 329, lists the valid entries for the ARGUMENTS argument.
Table 95. Valid arguments for ARGUMENTS
Argument Value Description
-source_object_id object_id Identifies the object or source
folder to be refreshed. If this
identifies an object, the object
must reside under the source
folder identified in the webc
config object specified in the
command line. If this identifies
a folder, the folder must be the
source folder or reside under
the source folder.
If you specify a folder, all
objects in and underneath the
folder are refreshed.
If you don’t include this
argument, the entire source
data set is refreshed.

WEBCACHE_PUBLISH
This option is ignored if
-full_refresh is set to TRUE.
-full_refresh TRUE or FALSE When this is set to TRUE, the
content and property data
at the target WebCache are
deleted and republished.
You must have Superuser
privileges to set this option to
TRUE. The default is FALSE.
-force_refresh TRUE or FALSE When this is set to TRUE,
documents are refreshed even
if their modification dates do
not indicate a need for a refresh.
The default for this option is
FALSE.
-method_trace_level trace_level Turns on tracing for the method
at the indicated level. Valid
trace levels correspond to
the levels available for the
IDfSession.setServerTraceLevel
method.
-recreate_property_schema TRUE or FALSE Performs a full refresh and
destroys and recreates the
propdb tables.
You must have Superuser
privileges to set this option to
TRUE. The default is FALSE.
—resync_state_table TRUE or FALSE TRUE deletes state information
in the probdb_tablename_m
table and recreates the
information based on the
current configuration object.
The default for this flag is
FALSE.
-store_log TRUE or FALSE Creates a log file object in the
repository for each publish
operation.
By default, this flag is TRUE for
all full refresh and incremental
publishing operations and

WEBCACHE_PUBLISH
FALSE when the method is
issued for a single item.
Return value
The WEBCACHE_PUBLISH administration method returns a collection identifier for a collection
with one property. The property, method_return_val, contains 0 if the operation was successful.
Permissions
You must have Sysadmin or Superuser privileges to use this administration method.
Generating a log file
When the operations generate a log file, the document is stored in the repository in
/System/Sysadmin/Reports/Webcache. By default, the method does not generate a log file if you are
publishing only a single item. To force it to generate a log file for single-item operations you can
include the -store_log argument, set to TRUE
Example
dmAPIGet("apply,s0,080015b38001a43c,WEBCACHE_PUBLISH,
ARGUMENTS,S,-source_object 090015b38007bf19")

WEBCACHE_PUBLISH

Chapter 4
Using DQL
This chapter introduces DQL and then provides usability information for DQL. The chapter includes
the following topics:
• Introducing DQL, page 333, introduces the Document Query Language and its basic query
statements.
• Quoting object type and property names, page 335, describes a recommended best practice for
referencing object type and property names in queries.
• NULLs, default values, and DQL, page 335, describes how DQL handles NULLs and default
values.
• Repeating properties in queries, page 339, describes how to reference repeating properties
in queries.
• Querying virtual documents, page 344, describes how to query virtual documents.
• Full-text searching and virtual documents, page 344, describes how you can conduct a fulltext
search on an assembled virtual document.
• Querying registered tables, page 345, describes how to reference registered tables in queries.
• Caching queries, page 346, describes how to cache query results.
• Privileges, permissions, and queries, page 346, describes how privileges and permissions affect
query results.
For information about using DQL and XDQL to query XML documents, refer to the EMC Documentum
XML Application Development Guide.
Introducing DQL
DQL is the acronym for Document Query Language, the SQL-like language that you use to query
the objects in a repository. Using DQL, you can retrieve, update, and delete objects and create new
objects. DQL also allows you to search indexed content and metadata (property values).
You can also use DQL to access registered tables—tables in the underlying RDBMS that are known to
Content Server but that are not part of the repository. (EMC Documentum Content Server Fundamentals
Guide describes the database tables that make up a repository.)
The basic DQL query statements retrieve information about the objects in a repository and manipulate
those objects. Table 96, page 334, describes the basic query statements.

Using DQL
Table 96. DQL basic query statements
Statement Description
SELECT The SELECT statement is the primary query
statement. SELECT statements return a wide
variety of information from the repository.
SELECT statements are not only stand-alone
statements but are also used in other DQL
statements.
There are two variants of the SELECT statement.
One is the standard SELECT statement, and the
other is called an FTDQL SELECT statement.
FTDQL queries are run entirely against the
fulltext index, which provide performance
benefits. However, the syntax for an FTDQL
query is a subset of that for a standard query.
For more information, refer to Select, page 116
or to the summary of the syntax in Appendix E,
DQL Quick Reference
UPDATE The UPDATE statement changes the RDBMS
table that underlies a registered table.
UPDATE...OBJECT The UPDATE...OBJECT[S] statement modifies
objects in the repository. Using this statement,
you can:
• Set the value of a single-valued property
• Append or insert values in a repeating
property
• Remove a value from a repeating property
• Truncate a repeating property (remove all
values)
• Link an object to a folder or cabinet or unlink
an object
• Move an object to a new folder or cabinet
CHANGE...OBJECT The CHANGE...OBJECT[S] statement moves
one or more objects from one object type to
another. With some constraints, you can change
an object of a particular type to any type that
is a subtype or supertype of the current type.
For example, you can change an object of type
dm_document to a user-defined document
subtype.

Using DQL
Statement Description
INSERT The INSERT statement adds a row to the
RDBMS table that underlies a registered table.
DELETE The DELETE statement destroys rows in the
RDBMS table that underlies a registered table.
DELETE...OBJECT The DELETE...OBJECT[S] statement deletes
objects from the repository.
When you issue one of these statements, only those objects for which you have the appropriate
permissions are affected. For example, if you issue an UPDATE...OBJECT statement to update
document objects, only those documents for which you have at least Write permission are considered
for updating. In addition, the statements have optional clauses that let you identify specifically
which objects are the target of the operation.
For a complete description of query statement syntax and usage, refer to the statement descriptions in
Chapter 2, DQL Statements, in the EMC Documentum Content Server DQL Reference guide.
Quoting object type and property names
Place double quotes around all references to object type names and property names in DQL queries.
While not required, this best practice ensures that the names will not conflict with words reserved
by DQL or the underlying RDBMS.
To encourage this practice, object type and property names are double quoted in all Content Server
documentation.
NULLs, default values, and DQL
In releases before release 6.6, Documentum did not allow you to assign NULLs as a property value
because NULLs are used to terminate the lists of values in repeating properties. In 6.6, support for
assigning NULLs as a property value was added. If you specify the SPACEOPTIMIZE keyword when
creating or modifying a property, NULL values can be assigned to that property.
When you create an object, any properties that you do not set are assigned default values by the
server. If a default value is defined in the data dictionary, the server assigns that value. Otherwise,
the default value depends on the data type of the property. If SPACEOPTIMIZE was not specified,
the default value from the table below is stored.
If SPACEOPTIMIZE was specified, then a property that you do not set is stored as NULL, and if you
explicitly set a property to the default value in the table below, it is also stored as NULL. For strings, a
single blank (’ ’) and the empty string (”) are both treated as the default value and are stored as NULL.
This is different from previous behavior where the empty string is distinct from the default value.
For Oracle installations, SPACEOPTIMIZE can modify all properties. For SQL Server and DB2, only
the character and string properties and the ID properties can be set to SPACEOPTIMIZE and only
save space for NULL on VARCHAR.

Using DQL
Table 97, page 336, lists the default values by data type.
Table 97. Default property values by datatype
Datatype Default value
String A single blank
Numeric data type 0
Date The NULLDATE value:
• For Oracle and DB2, it is 01/01/0001.
• For MS SQL Server, it is 1/1/1753.
ID ’0000000000000000’ (sixteen zeros)
Default values returned without SPACEOPTIMIZE
When you query a property containing a default value using DQL, the default values are returned in
the following manner:
• The default character string value (a single blank) is returned as an empty string.
• DQL also returns a single blank found in a registered table column as an empty string.
• The default numeric value (zero) is returned as zero.
• NULLDATE values are returned as the word NULLDATE.
If you are using SQL (for example, in a user-written report writer), the default values are returned in
the following manner:
• The default character string value (a single blank) is returned as a single blank.
• SQL also returns a single blank found in a registered table column as a single blank.
• The default numeric value (zero) is returned as zero.
• NULLDATE values are returned as the actual date (for example, 1/1/1 in Oracle).
Default values returned with SPACEOPTIMIZE
For single valued properties, default values stored as NULLs return NULLs. For repeating properties,
the situation is a little more complicated.
In releases earlier than 6.6, a NULL could not be stored as a value of a repeating property, since the
NULL indicted the end of the repeating values. For example, Table 98, page 337 shows an object with

Using DQL
ID obj1 and five repeating integer attributes (properties). Property attr1 has three values, the first
value being the default value, attr2 has one value, attr3 has two values, attr4 has three values with the
last one being the default value, and attr5 has four values, the second and fourth have default values.
Table 98. Repeating values without SPACEOPTIMIZE
r_object_id i_position attr1 attr2 attr3 attr4 attr5
obj1 -1 0 1 2 3 4
obj1 –2 1 NULL 1 4 0
obj1 –3 2 NULL NULL 0 2
obj1 –4 NULL NULL NULL NULL 0
In the next table, Table 99, page 337, the same object is shown as it would if SPACEOPTIMIZE had
been specified on these five repeating integer properties. Default values are shown in bold. Note that
the first value for attr1 and the second value for attr5 are NULL, and the third value for attr4 and the
fourth value for attr5 are zero. In this case, a default value might be returned as a NULL or as a zero.
Since either value could be returned for the default, there is new behavior for NULL comparisons in
6.6. See section Testing for default and NULL values, page 338, for a description of this new behavior.
Table 99. Repeating values with SPACEOPTIMIZE
obj1 -1 NULL 1 2 3 4
obj1 –2 1 NULL 1 4 NULL
obj1 –3 2 NULL NULL 0 2
obj1 –4 NULL NULL NULL NULL 0
If a new value is added to attr4 and a new default value is added to attr5, the following changes
shown in Table 100, page 337 will occur. As before, the default values are shown in bold. Notice that
the third value of attr4 and the fourth value of attr5 have been changed to NULL.
Table 100. Added repeating values with SPACEOPTIMIZE
obj1 -1 NULL 1 2 3 4
obj1 -2 1 NULL 1 4 NULL
obj1 -3 2 NULL NULL NULL 2
obj1 -4 NULL NULL NULL 6 NULL
obj1 -5 NULL NULL NULL NULL 0
So, in summary, for repeating properties that use SPACEOPTIMIZE, all default values are stored as
NULLs, unless the last value is a default value, and that property uses NULLs to fill out its column of
repeating values. In that case, the default value is stored as was done in releases before 6.6, and is still
done for properties that do not use SPACEOPTIMIZE.

Using DQL
Testing for default and NULL values
Documentum provides predicates for use in WHERE clauses to test for the presence of default values
or NULLs. Table 101, page 338, briefly describes these predicates. For a complete list of predicates,
refer to Predicates, page 32, of the EMC Documentum Content Server DQL Reference guide.
Table 101. Predicates that test for NULL and default values
IS [NOT]NULL Used only for columns in registered tables. Tests
for the presence of NULL.
ANY...IS [NOT]NULL Tests for the presence of the NULL terminator in
a repeating property. This is primarily useful if
you are testing values in two or more repeating
properties at corresponding index levels.
[ANY]...IS [NOT]NULLDATE Tests for the presence of the NULLDATE or a
NULL. Use the ANY option if the property is
a repeating property.
[ANY]...IS [NOT]NULLSTRING Tests for the presence of the default string value
(a single blank) or a NULL in a property. Use
the ANY option if the property is a repeating
property.
[ANY]...IS [NOT]NULLINT Tests for the presence of the default numeric
value (a zero) or a NULL in a property. Use
the ANY option if the property is a repeating
property.
[ANY]...IS [NOT]NULLID Tests for the presence of the default ID value
(’0000000000000000’) or a NULL in a property.
Use the ANY option if the property is a repeating
property. NULLID was introduced in release
6.6.
In release 6.6, true NULL storage was introduced. A server.ini flag was also added to control how
predicates that qualify a single-valued property behave. The flag, extend_default_predicate_to_null, is
true by default. When this flag is true, for any single-valued property that uses SPACEOPTIMIZE, a
query that uses the default value to qualify the property will be extended to also cover the NULL
value as well.
For example, for an integer property attr1, that uses SPACEOPTIMIZE, the following predicate:
attr1 = 0
will be converted to:
(attr1 = 0 OR attr1 IS NULL)
so that this property will be qualified as if it has the default value.
For more information about querying repeating properties, refer to Repeating properties in queries,
page 339.

Using DQL
Default values and aggregate functions
Aggregate functions are functions that operate on a group of values and return a single value. For
example, count is an aggregate function. It counts the values in a property and returns the number.
Aggregate functions do not ignore the default values assigned to a property. These values are
included in any calculations performed by the function. For example, suppose a repeating property
value can range from 1 to 10. If an object has three actual values (4, 7, and 6) and one default
value (one zero) for the property, the AVG function adds 4, 7, 6, and 0, then divides the total by 4
to obtain the average:
(4 + 7 + 6 + 0)/4 = 4.25
If the function ignores the default value, the average is:
(4 + 7 + 6)/3 = 5.66
As another example, look at the following SELECT statement:
SELECT COUNT(DISTINCT ratings) FROM "recipe"
WHERE "object_name" = 'lemon cake'
This statement (assuming the property values for ratings described for the previous example) returns
the number 4 because the default value is counted.
You can use the WHERE clause to exclude the default value when you are selecting aggregate
functions. For example, the following statement returns the correct number of ratings for the lemon
cake recipe:
SELECT COUNT(DISTINCT rating) FROM "recipe"
WHERE "object_name" = 'lemon cake'
AND "rating" IS NOT NULLINT
Sorting and nulls
Different databases handle NULLs differently when they are sorting values in a table. Here is how
each repository handles NULLs if the sort order is ascending:
• In Oracle and DB2, NULLs are sorted to the bottom of the list.
• In MS SQL Server, NULLs are sorted to the top of the list.
Repeating properties in queries
Repeating properties store multiple values. In Documentum, many object types have repeating
properties. For example, because a document can have multiple authors, the authors property of
the dm_document type is a repeating property.

Using DQL
You can reference repeating properties in:
• The selected values list in a SELECT statement
• WHERE clause qualifications for SELECT, UPDATE...OBJECT, CHANGE...OBJECT, and
DELETE...OBJECT statements
• The update clauses of the CREATE...OBJECT, UPDATE...OBJECT, and CHANGE...OBJECT
statements
For information about using a repeating property in SELECT statement, as a selected value or in the
WHERE clause qualification, refer to Select, page 116.
Modifying repeating attributes
You can use the UPDATE...OBJECT statement to add, delete, and modify individual values for
repeating properties. In most of these operations, you must specify the index position of the value
you want to change. The index indicates the value’s position in the list of values assigned to the
repeating property. Index numbers begin with zero for the first value and increment by 1 for each
additional value.
For example, suppose the a recipe object type has a repeating property called ingredients. One recipe
has the following ingredients: eggs, sugar, cream cheese, and amaretto. Assuming the ingredients
were added to the ingredients property for a recipe object in the order listed, they would have the
following index values:
ingredient[0] (eggs)
ingredient[1] (sugar)
ingredient[2] (cream cheese)
ingredient[3] (amaretto)
Index positions are always specified inside square brackets after the name of the property.
Adding new values
To add a value to a repeating property, you can:
• Insert the new value, which lets you choose where to put the new value in the property’s value list
• Append the value, which automatically places the value at the end of the property’s value list
Inserting values
When you want to control where the new value is placed in the repeating property, use the insert
option as the update clause in the UPDATE...OBJECT statement. The syntax is:
UPDATE object_type OBJECTS
INSERT property_name[x] = value
...
For example, suppose that the authors of the Espresso Cheesecake recipe forgot to include one of
the ingredients for the crust. The ingredients for the crust are listed ahead of the ingredients for the

Using DQL
filling in the document. Consequently, the authors want to insert the forgotten ingredient ahead of
the filling ingredients so that it appears with the other crust ingredients. The ingredients for the crust
occupy positions 0 through 3 in the ingredients property, and the filling ingredients begin at position
4. The following statement inserts the forgotten ingredient at position 4.
UPDATE "recipe" OBJECTS
INSERT ingredients[4] = '2 T ground bittersweet chocolate'
WHERE "title" = 'Espresso Cheesecake'
The server inserts the requested value and renumbers all other values from position 4 on. The filling
ingredients now begin at position 5.
Inserting values into a repeating property never overwrites a current value. Any value currently at
the specified insertion point and any following values are always renumbered.
If you do not specify an insertion point, the server automatically inserts the new value in the first
position (property_name[0]).
You can identify the value you want to insert either as literal value or by using a subquery. If you
use a subquery, it can return only one value. If it returns multiple values, the INSERT statement
fails with an error.
Appending values
When you append values, the server automatically adds the new value to the end of the list of values
in the property. You do not have to specify an index value when you append. To illustrate, suppose
the authors want to add another ingredient to the Espresso Cheesecake recipe: strawberries to be
used as a garnish. To append this ingredient, they use the following statement:
UPDATE "recipe" OBJECT
APPEND "ingredients" = '1 pint strawberries'
WHERE "title" = 'Espresso Cheesecake'
You can identify the appended value as a literal value or by using a subquery. For example, perhaps
your company is reorganizing and employees currently working for Mr. Rico are being moved to a
group called frontline. The following statement uses a subquery to find those employees and append
them to the list of users in the frontline group:
UPDATE "dm_group" OBJECTS
APPEND "users_names"=(SELECT "user_name"
FROM "dm_user","employees" e
WHERE e."manager"='rico' AND "r_is_group"=F)
WHERE "group_name" = 'frontline'
When using a subquery, you can specify the maximum number of values that you want to append.
The syntax in the update clause is:
APPEND n property_name = subquery
where n represents the maximum number of appended values.
Updating values
To update a value, use the set option as the update clause.

Using DQL
For example, the authors of Heavenly Cheesecakes have decided to update one of the recipes in the
book. They want to replace the cream cheese requirement with a lower-fat substitute—ricotta cheese.
Knowing that cream cheese is the third ingredient in the ingredients list, they use the following
statement to make the substitution:
SET ingredients[2] = '1 lb ricotta cheese'
WHERE "title" = 'mocha cheesecake'
Note that the index value for cream cheese is [2]. This is because index values are counted from
zero, so the first two ingredients (eggs and sugar) have index values of zero and one ([0] and [1]),
respectively.
Deleting values
To delete a value in a repeating property, use the remove option as the update clause and specify
the index value associated with the value. To illustrate, the following example removes the fifth
ingredient in a list of ingredients for brownies:
REMOVE ingredients[4]
WHERE "object_name" = 'Mandarin Brownies'
If you do not specify which value to remove, the system automatically removes the first ([0]) value.
Forcing index correspondence in query results
Note: The query syntax described in this section may not be used in an FTDQL SELECT statement.
(For information about FTDQL SELECT statements, refer to Select, page 116.)
In some queries, you may only want objects for which the repeating property values are in the same
relative positions. For example, perhaps you want only those recipes for which a particular author
and keyword occupy the same index position.
To require index correspondence for expressions referencing repeating properties, use the following
syntax:
ANY ([NOT] predicate AND [NOT] predicate
{AND [NOT] predicate})
To force index correspondence, the predicates must be ANDed inside the parentheses. Using the OR
operator to link the predicates inside the parentheses increases the query’s performance but does not
force index correspondence. The query returns any object that contains one of the ORed values.
To illustrate, assume that the keywords untried, tested, approved, and rejected are assigned to recipes
in various stages of acceptance and that these keywords are always placed in the first position in the
keywords list, to correspond to the implicit version label (the numeric label) associated with the recipe.
The following statement finds only original recipes (version 1.0) that were rejected by the testers:
SELECT "r_object_id", "object_name" FROM "recipe"
WHERE ANY ("r_version_label" = '1.0'
AND "keywords" = 'rejected')
In the above example, for the recipes returned, r_version_label[0] = 1.0 and keywords[0] = rejected.

Using DQL
You can’t specify which position is searched when you force index correspondence. You can only
specify that the values be found in the same position. In the above example, it was easy to know that
the values are in the first position for all pairs because the implicit version label is always stored in
the first position. However, for other repeating properties this will not be true. For example, look
at the following statement:
UPDATE "recipe" OBJECTS
SET "subject" = 'lowfat meals'
WHERE ANY ("ingredients" IN ('skim milk','margarine')
AND "keywords" = 'fat free')
The statement updates the subject property for any recipe for which skim milk and fat free or
margarine and fat free occupy the same respective positions within their individual properties. For
one recipe, the positions might be the fourth: ingredients[3] = margarine and keywords[3] = fat
free. For another, the values might be found in the seventh position: ingredients[6] = skim milk
and keywords[6] = fat free. In all instances, the values are at the same index position within their
properties.
It is possible to combine both forms of the syntax. For example:
SELECT "r_object_id", "title" FROM "recipe"
WHERE ANY "ingredients" IN ('cream cheese','chocolate')
AND ANY("authors" = 'daphne' AND "keywords" IN ('cheesecake','brownies','mousse'))
This statement returns all recipes that have cream cheese or chocolate as an ingredient and also have
the author daphne paired with the keyword cheesecake, brownies, or mousse.
Performance note for MS SQL Server users
DQL turns repeating property predicates (and the FOLDER predicate) into ORed subselects
internally. Queries that contain ORed subselects run slowly against MS SQL Server, and performance
degrades in direct proportion to the number of SysObjects in your repository.
To improve performance, here are some suggested alternative ways to formulate queries:
Instead of:
SELECT "r_object_id" FROM "dm_sysobject"
WHERE ANY "authors" = 'a' OR ANY authors = 'b'
Use:
WHERE ANY ("authors" = 'a' OR "authors" = 'b')
Instead of:
WHERE FOLDER ('/Temp) OR FOLDER ('/System')
Use:
WHERE FOLDER ('/Temp', '/System')

Using DQL
Querying virtual documents
Virtual documents are documents that contain other documents. The documents they contain can
be simple documents or other virtual documents. Nesting virtual documents inside other virtual
documents creates a hierarchy of components within the top-level virtual document. Documentum
allows you to nest virtual documents to any depth you choose.
To identify the virtual document you want to query, you can use either the IN DOCUMENT clause or
the IN ASSEMBLY clause in a SELECT statement. The IN DOCUMENT clause allows you to choose
which set of the document’s components to query at runtime. The IN ASSEMBLY clause directs
the query to the virtual document’s snapshot, which is a previously selected set of the document’s
components.
If you are querying a virtual document in an UPDATE...OBJECT or DELETE...OBJECT statement, you
must use the IN ASSEMBLY clause.
Both the IN DOCUMENT and IN ASSEMBLY clauses include the optional keyword DESCEND. This
keyword directs the server to search any components contained by components that are themselves
virtual documents. The only exception is if the component is a reference link. In such cases, the server
can’t search for the component’s contained components. The search returns only the reference link.
Documentum also allows you to specify one particular component of a virtual document as the
target of a query. This feature can make it easy for users who are working on a virtual document to
pull out a specific part of the document for work. The NODE option of the IN ASSEMBLY clause
implements this feature.
Full-text searching and virtual documents
Like simple documents, virtual documents can have associated content files. Marking a virtual
document for indexing directs Content Server to index the content files associated with the virtual
document—not the components of the virtual document.
In SELECT statements, if you include a SEARCH clause and an IN DOCUMENT clause, the server
first searches to see whether the content files of the specified virtual document meet the full-text search
criteria. Then, as the server assembles the virtual document, the criteria in the SEARCH clause are
applied to the content files associated with each component.
If you want to conduct fulltext searches on the assembled document components, you must create
a file from the assembled document and associate it with the virtual document as a content file.
Use the following procedure:
To index an assembled document as a whole:
1. Assemble the document.
2. Print the document to a file.
3. Use Setfile or Setcontent to associate the resulting file with the virtual document as a content file.
Now you have a content file for the virtual document that represents the assembled document
and that can be indexed.

Using DQL
Querying registered tables
Registered tables are tables in the underlying RDBMS that have been registered with the repository.
You must register any RDBMS table that you wish to query using DQL. Registering a table creates
an object of type dm_registered for that table. (Refer to Register, page 110, in the EMC Documentum
Content Server DQL Reference guide for information about registering tables.)
To obtain information about the repository object that represents the table, specify the dm_registered
type in the FROM clause of the SELECT statement. The server searches for the object representing
the table. To query the table itself, specify the registered table’s name in the FROM clause. When
you query the table, the server searches the actual underlying RDBMS table—not the dm_registered
object type.
Referencing registered tables in queries
To avoid ambiguity in a query, it is often useful to reference a database table by its fully qualified
name; that is, owner_name.table_name.
The owner name is the name of the person who created the table. Tables created by the system when
a repository is created (for example, dm_sysobject_r or dm_sysobject_s) are owned by the repository
owner. For these tables, Oracle and DB2 uses the value in the repository owner’s user_db_name
property as the owner name. MS SQL Server prefixes the names of all tables created by the repository
owner with the alias dbo. This alias makes it possible to write applications that are portable across
MS SQL Server databases.
For application portability across all repositories, Documentum provides an alias, dm_dbo, that you
can substitute for the repository owner’s name in any fully qualified reference to registered tables.
(For MS SQL Server, when referencing registered tables, the aliases dm_dbo and dbo are equivalent.)
Security controls
Access to a registered table is controlled by the object-level permissions and table permits defined for
the table’s dm_registered object. The object-level permissions must give you at least Browse access
to the dm_registered object and the table permits must give you permission for the operation that
you want to perform. For example, if you want to update a table, the table permits must grant you
DM_TABLE_UPDATE permission. Table permits are defined for three user levels: owner, group, and
world. (For complete information about table permits and Documentum security, refer to the EMC
Additionally, the user account under which Content Server is running must have the appropriate
RDBMS permission to perform the requested operation on the specified table. (The actual name of
this permission depends on your RDBMS.) For example, if you want to update the table, the server
account must have permission in the RDBMS to update the table.
All three of these conditions must be met to gain access to a registered table. Even if a user has
permission to access the underlying table through the RDBMS, the user will not have access through
DQL unless the object-level permissions and table permits on the dm_registered object permit access.

Using DQL
Similarly, a user might have the correct object-level permissions and table permits, but if the server’s
user account in the RDBMS does not have permission, then access is denied.
Default object-level permissions and table permits
The REGISTER statement automatically sets the object-level permissions for a table’s dm_registered
object to default values. The statement also sets the default table permit to SELECT for the owner
(group and world have no default table permit).
You can change the object-level permissions and table permits by setting the appropriate properties
directly. TheEMC Documentum Content Server Administration and Configuration Guide provides more
information about these properties.
Caching queries
Some queries return the same results every time you run the query. For example, a payroll
application may ask for the names and social security numbers of the employees in the company.
Although the query results may change over a long period of time, they may not change from week
to week. Rather than incur the performance cost of rerunning the query that returns the users and
social security numbers each time the payroll application executes, you can cache the query results
on the client host. For information about persistent caching, refer to EMC Documentum Content
Server Fundamentals Guide.
Privileges, permissions, and queries
DQL query statements affect only those objects for which users have the appropriate permissions or
privileges.
When a SELECT statement references a SysObject type (dm_sysobject or any of its subtypes), by
default, only those objects of the referenced type for which the user has at least Browse permission
are retrieved. If the SELECT statement includes a SEARCH clause that accesses the object’s content,
then the user must have at least Read permission. However, the SELECT statement supports an
option to allow you to specify another base permission level. For example, you might use a query that
returns only objects for which a user has at least Relate permission.
If the object type specified in the SELECT statement is modified by the keyword PUBLIC, then the
statement retrieves only objects that have a world permission of at least Browse (or Read for full-text
searches) or objects that are owned by the user.
When a SELECT statement references a registered table, the user must have at least Browse
permission on that registered table for the query to succeed. In addition, Content Server must have
the SELECT privilege in the RDBMS.
The Superuser user privilege bypasses all Documentum security checks. However, Content Server
must still have the SELECT privilege in the RDBMS for a superuser to query registered tables.

Using DQL
You must have Write permission to update an object with the UPDATE...OBJECT statement. You
must have Delete permission for an object to delete it using the DELETE...OBJECT statement.

Using DQL

Appendix A
Using DQL Hints
This appendix describes how DQL hints are implemented and how to use them in your queries.
• General guidelines for all, page 350
• CONVERT_FOLDER_LIST_TO_OR, page 350
• FETCH_ALL_RESULTS N, page 350
• FORCE_ORDER, page 351
• FTDQL and NOFTDQL, page 351
• FT_CONTAIN_FRAGMENT, page 352
• GENERATE_SQL_ONLY, page 352
• GROUP_LIST_LIMIT N, page 352
• HIDE_SHARED_PARENT DQL Hint, page 353
• IN and EXISTS, page 353
• OPTIMIZE_ON_BASE_TABLE, page 354
• OPTIMIZE_TOP N, page 355
• RETURN_RANGE, page 356
• RETURN_TOP N, page 357
• ROW_BASED, page 359
• SQL_DEF_RESULT_SET N, page 361
• TRY_FTDQL_FIRST, page 362
• UNCOMMITTED_READ , page 362
• DM_LEFT_OUTER_JOIN_FOR_ACL, page 362
• Including multiple hints limiting rows returned, page 362
• Passthrough hints, page 363

Using DQL Hints
General guidelines for all
Database hints affect how a query is executed. When deciding whether to use a hint, it is
recommended that you execute the query with and without the hint. Compare the generated SQL
queries and compare the response time and the resource use to determine whether the hint is helpful.
You can use any of the hints except ROW_BASED and RETURN_RANGE in an FTDQL query. The
ROW_BASED hint may not be included in an FTDQL query and RETURN_RANGE is not FTDQL
compliant.
Additionally, you may not use the ROW_BASED hint in query that includes a lightweight object type
in the FROM clause.
CONVERT_FOLDER_LIST_TO_OR
A new DQL hint was added as a platform compatibility update to Content Server in release 6.7
SP1. It provides backward compatibility to the behavior of release 6.0 and earlier. Beginning with
release 6.5, a DQL FOLDER clause like:
... FOLDER ('/Temp', '/Temp/test_folder_1', '/System')...
is translated to an SQL statement with a clause like:
... dm_sysobject_r2.i_folder_id IN
('0c01068980000106', '0b01068980004509',
'0c001068980000107') ...
In some cases, this causes performance degradation compared with the translation used in the earlier
release. The earlier translation of the FOLDER clause is:
... dm_sysobject_r2.i_folder_id = '0c01068980000106'
or dm_sysobject_r2.i_folder_id = '0b01068980004509'
or dm_sysobject_r2.i_folder_id = '0c001068980000107' ...
To get the earlier translation, use the CONVERT_FOLDER_LIST_TO_OR DQL hint. For example:
... FOLDER ('/Temp', '/Temp/test_folder_1', '/System') ENABLE(CONVERT_FOLDER_LIST_TO_OR) ...
FETCH_ALL_RESULTS N
The FETCH_ALL_RESULTS N hint fetches all the results from the database immediately and
closes the cursor. The hint does not affect the execution plan, but may free up database resources
more quickly.
To fetch all the results, set N to 0.
On SQL Server, it is recommended that you use SQL_DEF_RESULT_SETS instead of the
FETCH_ALL_RESULTS hint. SQL_DEF_RESULTS_SETS provides the same benefits and is the
recommended way to access SQL Server databases.
Use FETCH_ALL_RESULTS if you want to reduce the resources used by the database server by
quickly closing cursors. On SQL Server, try FETCH_ALL_RESULTS if using SQL_DEF_RESULT_SETS
did not improve query performance.

Using DQL Hints
SQL Server, the hint, and subqueries
If you use this hint on query against a SQL Server database and the query includes a subquery, the
hint is not applied to the subquery.
FORCE_ORDER
The FORCE_ORDER hint controls the order in which the tables referenced in the query’s FROM
clause are joined. The tables may be RDBMS tables or object type tables.
Oracle and SQL Server implement this hint at the statement level. For example, suppose you issue the
following DQL:
SELECT object_name
FROM DM_SYSOBJECT ENABLE(FORCE_ORDER)
The generated SQL for Oracle is:
select /*+ ORDERED */ object_name from dm_sysobject_s;
The generated SQL for SQL Server is:
select object_name from dm_sysobject_s
OPTIONS (FORCE_ORDER)
Using FORCE_ORDER may not ensure that you obtain a particular join order. When you examine the
SQL query generated by a DQL query, you may find there are more tables in the FROM clause than
are present in the DQL query. This is because Content Server often uses views to generate the SQL for
many DQL queries and may also add ACL tables to the queries for security checks.
If you are considering using this hint, run the query without the hint and obtain the generated
SQL statement and the execution plan. If the join order in the generated query appears incorrect
and you believe that joining the tables in the order they appear in the DQL query will result in
better performance, run the query with FORCE_ORDER. Compare the results to the query execution
results without the hint.
If you use this hint, it is recommended that you retest the query with and without the hint occasionally
to ensure that the hint is still useful. Database changes can make the plan chosen by the hint incorrect.
FTDQL and NOFTDQL
The FTDQL hint supports the FTDQL functionality. If you include this hint in query, Content Server
attempts to execute the query as an FTDQL query. If the remaining syntax in the query conforms to
the required syntax for an FTDQL query, the query is executed as an FTDQL query. If the syntax does
not conform to FTDQL query rules, an error is returned.
If you include the NOFTDQL hint, the query is run as a standard SELECT query whether the syntax
satisfies the FTDQL query rules or not.
A standard query queries both the fulltext index and the database. Such a query typically contains a
SEARCH clause and a WHERE clause. If the WHERE clause is not compliant with the rules of
FTDQL or the query contains an explicit ENABLE(NOFTDQL), Content Server executes the SEARCH

Using DQL Hints
clause against the fulltext index and the WHERE clause against the database and then returns the
intersection of the results.
FT_CONTAIN_FRAGMENT
The FT_CONTAIN_FRAGMENT hint expands the range of matches for full-text searches that include
a search clause of the form LIKE ’%string%’ clause, such as:
subject LIKE '%work%'
When the query includes the DQL hint FT_CONTAIN_FRAGMENT, this clause returns all objects
whose subject contains "work” in any position, including where the string appears as part of a
larger string. For example, it returns any object whose subject contains "networking” or "workers”.
However, if the query does not include this hint (the default behavior), a full-text search that
uses this clause returns only those objects whose subject contains "work” (or a variant form of
the word if grammatical normalization is enabled) as a separate word. Limiting the results to
separate word matches improves the performance of the query. Beginning with release 6.0, the
default behavior was changed. In previous releases, the default was to return objects as though
FT_CONTAIN_FRAGMENT was specified.
The FT_CONTAIN_FRAGMENT hint affects only full-text searches. Database searches return partial
string matches regardless of this hint. The hint only applies to search clauses that include a percent
sign both before and after the literal string; The behavior for LIKE ’%work’ and LIKE ’work%’
are unaffected.
GENERATE_SQL_ONLY
The GENERATE_SQL_ONLY hint returns the SQL statement for a particular DQL statement (Select
statement only) without executing the statement on the database. This hint enables you to verify the
SQL statement for a DQL statement without querying the data on the database.
GROUP_LIST_LIMIT N
The GROUP_LIST_LIMIT hint may improve query performance when a user is a member of a
large number of groups. For example, suppose a user who is a member of 500 groups executes
the following query:
SELECT r_object_id FROM dm_document WHERE object_name LIKE '%dm%'
When this query is translated to SQL, a subquery is needed to perform the ACL checking because the
default group list limit for queries is 250. The performance is slower due to the need for the subquery.
You can use the DQL hint to override the default value for group list limit. For example:
SELECT r_object_id FROM dm_document
WHERE object_name LIKE '%dm%'
ENABLE(GROUP_LIST_LIMIT 600)
Using this hint overrides the default value and the DM_GROUP_LIST_LIMIT environment variable,
if that is set, as well as flushes and rebuilds the group list cache in the new size. For users whose

Using DQL Hints
group membership number is below the specified limit, generated SQL statements display group
names inline without subqueries.
HIDE_SHARED_PARENT DQL Hint
The HIDE_SHARED_PARENT DQL hint directs Content Server to return only the rows in the query
results that are not shared parents. In order to accomplish this, the server will add two additional
attributes, r_object_id and r_object_type to the SQL statement select list, if not already there (only
applies to dm_sysobject or any of its subtypes). The server will run the SQL query against the
database and then for each qualified row it will get the value of r_object_type, fetch the type object
and check to see if it is a shareable type. For non-shareable types it will just return the row, but for
shareable types it will do the following:
• Look in the sysobject cache for the r_object_id object, or issue an SQL statement if the object is not
in the cache, and examine the i_sharing_type.
• Return the row if i_sharing_type is empty, or skip the row if it is not empty (indicating that the
row is from a shared parent).
The results of a query that uses the HIDE_SHARED_PARENT DQL hint will not contain any shared
parents. A side effect of this hint is that queries will show shared parents without child objects, but
the shared parents will disappear from the results when a child LWSO is attached. Webtop uses this
hint by default.
This hint does affect performance, since there are the additional checks required to determine
whether the result contains shared parents.
IN and EXISTS
IN and EXISTS are mutually exclusive hints that you can use in a standard WHERE clause, in a
repeating property predicate that contains a subquery. These hints may not be used in a WHERE
clause in an FTDQL query.
The syntax is:
WHERE ANY [IN|EXISTS] property_name (subquery)
For example:
WHERE ANY IN "authors" IN
(SELECT "user_name" FROM "dm_user")
If you do not include either IN or EXISTS explicitly, when Content Server translates the query, it
includes one or the other automatically. Which hint Content Server chooses to include is determined
by the indexes present in the repository, the properties referenced in the query, and other factors.
The queries generated by each option are different and perform differently when executed. For
example, here is the generated SQL statement for the query in the previous example:
select all dm_sysobject.r_object_id
from dm_sysobject_sp dm_sysobject
where (dm_sysobject.r_object_id in
(select r_object_id from dm_sysobject_r

Using DQL Hints
where authors in
(select all dm_user.user_name
from dm_user_sp dm_user)))
and (dm_sysobject.i_has_folder=1 and
dm_sysobject.i_is_deleted=0)
If the DQL query used the EXISTS hint, the generated SQL statement would look like this:
select all dm_sysobject.r_object_id
from dm_sysobject_sp dm_sysobject
where (exists (select r_object_id
from dm_sysobject_r
where dm_sysobject.r_object_id = r_object_id
and authors in
(select all dm_user.user_name
from dm_user_sp dm_user )))
and (dm_sysobject.i_has_folder=1 and
dm_sysobject.i_is_deleted=0)
If you feel that a query that references a repeating property predicate that uses a subquery is
performing badly, run the query and examine the generated SQL statement to determine which hint
Content Server is adding to the generated SQL and note the performance time. Then, rewrite the
DQL query to include the hint that the server isn’t choosing. For example, if the server is choosing to
add the EXISTS hint to the generated SQL statement, rewrite the DQL query to include the IN hint.
Then, rerun the query to determine if the performance improves.
OPTIMIZE_ON_BASE_TABLE
The OPTIMIZE_ON_BASE_TABLE hint can optimize queries that only reference properties contained
within one type in the type hierarchy.
DQL statements that do not use this hint query against the view of the table for the type (the _sp or
_rp view). When the type hierarchy is deep, or the number of instances are large, a query run against
the view can be slow. In contrast, if all the attributes in the query are contained in the table for the
type, a query run only against that table can be much faster. Since the properties are contained in that
table, there is no need for the joins required to query attributes in the tables of the supertype. You
must evaluate this hint for your own environment before you decide to use it in production.
The OPTIMIZE_ON_BASE_TABLE hint has the following effects for these cases:
• Case 1: The type in the FROM clause does not have any supertype. For this case, the hint has no
effect, and the _sp and/or _rp views are used.
• Case 2: The referenced properties in the select list, WHERE clause, ORDER BY clause, or GROUP
BY clause, are from more than one type. For this case, the hint has no effect, and the _sp and/or
_rp views are used.
• Case 3: The type is not a subtype of dm_sysobject. A) If all properties are from the type specified
in the FROM clause, then the _s and _r tables of the type are joined (or only the _s table if all
attributes are single-valued). B) If all properties are from a supertype of the type specified in the
FROM clause, then the relevant _s and/or _r base table is used, plus the _s table of the referenced
type is joined with the _s table of the type in the FROM clause.
• Case 4: The type is a subtype of dm_sysobject. A) If all properties are from the type specified in
the FROM clause, then the hint has the same effect as for Case 3A, but if an ACL predicate is
needed (for non-superusers), or a version clause is needed (when DELETED is not specified), then

Using DQL Hints
the dm_sysobject_s table is also joined for the query. B) If all properties are from the supertype of
the type specified in the FROM clause the effect is the same as for Case 3B, but if an ACL predicate
or version clause is needed, then the dm_sysobject_s table is also joined for the query.
Here is an example of Case 3A, all properties from the specified type:
SELECT group_class, groups_names
FROM dm_audittrail_group
enable(OPTIMIZE_ON_BASE_TABLE)
And this listing is the resulting SQL:
SELECT all dm_audittrail_group.group_class, dm_repeating.groups_names
FROM dm_audittrail_group_s dm_audittrail_group,
dm_audittrail_group_r dm_repeating
WHERE dm_repeating.r_object_id=dm_audittrail_group.r_object_id
Here is an example of Case 3B, all properties from the supertype of the specified type:
SELECT event_name, object_name
FROM dm_audittrail_group
enable(OPTIMIZE_ON_BASE_TABLE)
And this listing is the resulting SQL:
SELECT all dm_audittrail_group.event_name, dm_audittrail_group.object_name
FROM (select dm_opt_base.* from dm_audittrail_s
dm_opt_base, dm_audittrail_group_s dm_opt_from
where dm_opt_base.r_object_id = dm_opt_from.r_object_id)
dm_audittrail_group
OPTIMIZATION_LEVEL level_1 level_2
Use the OPTIMIZATION_LEVEL hint against a DB2 database when you want to change the
optimization level for a particular query. The level you set in the level_1 argument is used for the
current query and after the query is finished, Content Server sets the optimization level to the level
specified in the level_2 argument.
Using this hint generates a SQL trace that looks like the following:
set current optimization level level_1
run query
set current optimization level level_2
This hint is ignored for all databases except DB2.
For more information or assistance on the level_1 and level_2 parameters, please contact your DB2
database vendor.
OPTIMIZE_TOP N
The OPTIMIZE_TOP N hint directs the database server to return the first N rows returned by a query
quickly. The remaining rows are returned at the normal speed.

Using DQL Hints
On SQL Server and DB2, you can include an integer value (as N) to define how many rows you want
returned quickly. On Oracle, the number is ignored.
For example, suppose you issue the following DQL query:
SELECT r_object_id FROM dm_sysobject
ENABLE (OPTIMIZE_TOP 4)
On SQL Server, the generated SQL statement is:
select r_object_id from dm_sysobject_s
OPTION (FAST 4)
On DB2, the generated SQL statement is:
select r_object_id from dm_sysobject_s
OPTIMIZE FOR 4 ROWS
On Oracle, the generated SQL statement is:
select /*+ OPTIMIZE_TOP */ r_object_id
from dm_sysobject_s
OPTIMIZE_TOP is recommended for use:
• When you want to return all the results but want the first few rows more quickly
• With the RETURN_TOP hint, to optimize the return of specified rows
• When the execution plan chosen for query is a bad plan and there is no obvious solution
Using OPTIMIZE_TOP if the query sorts the results or includes the DISTINCT keyword reduces the
efficiency of the hint.
RETURN_RANGE
The RETURN_RANGE hint specifies which rows are returned by a query sorted by the returned
values of specified properties. This hint is provided as a general way to paginate the results of
a query. The syntax of this hint is:
RETURN_RANGE starting_row ending_row [optimize_top_row] 'sorting_clause'
Table 102. RETURN_RANGE argument descriptions
starting_row Specifies the starting row number to return from the qualified
rows.
ending_row Specifies the ending row number to return from the qualified
rows.
optimize_top_row Specifies the number of top rows to be optimized by the
database optimizer. This parameter is optional. If not
specified, then there is no corresponding hint generated for
the converted SQL statement.
sorting_clause Specifies the attribute and its sorting sequence used to
determine the range. It defines the sequence of the qualified
results.

Using DQL Hints
The syntax of the sorting clause is:
'attribute_name[ASC|DESC][,attribute_name[ASC|DESC]...]'
Table 103. sorting_clause argument descriptions
attribute_name Specifies which attribute is used to sort the qualified rows.
ASC Sorts in ascending order. This is the default if no order is
specified.
DESC Sorts in descending order.
The RETURN_RANGE hint can appear in the outermost query and in subqueries. When in the
outermost query, RETURN_RANGE honors the object mode, like the RETURN_TOP hint. In other
words, if the ROW_BASED hint is used, then RETURN_RANGE returns the top rows as specified,
otherwise, it returns the specified top rows or objects based on the server.ini settings. If the
RETURN_RANGE hint appears in a subquery, it will only perform on the row level.
RETURN_TOP N
The RETURN_TOP N hint, by default, limits the number of rows returned by a query. Since an object
with repeating properties may consist of multiple rows, this hint may return fewer than N objects.
In release 6.5 SP2, a server.ini flag was added, return_top_results_row_based, true by default, that
controls whether RETURN_TOP returns the number of rows or the number of objects specified.
Setting this flag to false (return_top_results_row_based = F), causes the returned results to be limited
by the number of objects.
More details about RETURN_TOP N
DQL is an object based query language - and we need to understand whether RETURN_TOP 10
means to return 10 rows, or 10 objects. Databases are not aware of Documentum objects. 10 rows
does not always equal 10 objects. If a DQL query contains repeating attributes, each database row
returned does not translate into a separate Documentum object. If we make RETURN_TOP 10 mean
top 10 objects, we must limit results at the Content Server level. If we mean RETURN_TOP 10 to
mean top 10 rows, we can limit at the database level (for more efficiency). A query containing only
single valued attributes will behave the same for object or row-based results.
Database-specific implementations
The following sections describe how RETURN_TOP is handled for individual databases.

Using DQL Hints
SQL Server
On SQL Server, using this hint results in fewer database rows being touched by a query. The internal
behavior of query on SQL Server is:
1. The query executes on the database (touching whatever tables are required) and generates a
list of keys or lookup IDs inside the tempdb.
2. Each time a row is fetched from the cursor, the lookup ID accesses the actual table to return
the full result set with all columns.
When you include the RETURN_TOP hint, the second step is executed only as many times as the
hint directs. For example, if the hint is RETURN_TOP 20, then only 20 rows are returned and the
table is accessed only 20 times.
Including RETURN_TOP adds the SQL Server TOP hint to the generated SQL statement. For
example, suppose you issue the following DQL statement:
SELECT user_name FROM dm_user ENABLE (RETURN_TOP 3)
The generated SQL for SQL Server is:
select TOP 3 user_name from dm_user_s
Subqueries and the hint
If the query includes a subquery, the hint is not applied to the subquery.
DB2
On a DB2 database, including RETURN_TOP adds the FETCH FIRST N ROWS ONLY hint to the
generated SQL statement. For example, the following DQL query
SELECT user_name FROM dm_user ENABLE (RETURN_TOP 3)
is translated to the following SQL statement
select user_name from dm_user_s
FETCH FIRST 3 ROWS ONLY
It is recommended that you use RETURN_TOP in conjunction with OPTIMIZE_TOP. The previous
example becomes:
SELECT user_name FROM dm_user
ENABLE (RETURN_TOP 3, OPTIMIZE_TOP 3)
The statement generates the following SQL:
select user_name from dm_user_s
FETCH FIRST 3 ROWS ONLY OPTIMIZE FOR 3

Using DQL Hints
Oracle
If you include RETURN_TOP in a query running against Oracle, the returned results are not limited
at the database level, but by Content Server itself. Consequently, using RETURN_TOP on Oracle
results in no database performance gains, but may reduce network traffic.
Additionally, in Oracle implementations before release 6.5, RETURN_TOP N returned the number of
objects specified, not the number of rows. Use the server.ini flag, return_top_results_row_based, in
releases 6.5 SP2 and later to cause this same behavior.
Effects of a SEARCH clause
If the DQL query includes a SEARCH clause, to limit results to documents that are indexed, the
implementation of RETURN_TOP depends on whether the searched documents are public or not.
If all the searched documents are public, the returned results are limited at the fulltext index level. If
the searched documents are not all public, then the limits are imposed when Content Server performs
the security checking on the returned results.
Recommended use
Use RETURN_TOP when:
• Only a particular number of rows is needed from the database.
• The application accepts ad hoc queries from users and you want to limit the potential damage
an unbounded query might cause.
Using RETURN_TOP if the query sorts the results or includes the DISTINCT keyword reduces the
efficiency of the hint.
ROW_BASED
The ROW_BASED hint changes both the way query results are returned and the syntax rules for
the query itself.
This hint may not be used in FTDQL queries or queries that reference a lightweight object type
in the FROM clause.
Effects on returned results
The ROW_BASED hint forces Content Server to returns query results in a row format, as opposed to
an object-based format. The difference is most readily apparent if the query selects values from a
repeating property. In an object-based format, the server returns all selected repeating values for a

Using DQL Hints
particular object in one query result object. In a row-based format, the server returns each selected
repeating property value in a separate query result object.
For example, by default the following query returns results in an object-based format:
SELECT "r_object_id","title","authors" FROM "dm_document"
WHERE "subject"='new_book_proposal'
That query returns the authors values as a list of authors in one query result object for each returned
object. Table 104, page 360, shows how the results are returned. Each row in the table represents one
query result object and each column is one property in the query result objects.
Table 104. Example of object-based query results
090000015973a2fc Our Life and Times Jennie Doe Carol Jones
Hortense Smith
0900000123ac12f6 Life of an Amoeba James Does Jules Doe
There is one query result object (one row) for each object returned by the query. Now, add the
ROW_BASED hint to the query:
SELECT "r_object_id","title","authors" FROM "dm_document"
WHERE "subject"='new_book_proposal'
ENABLE(ROW_BASED)
Table 105, page 360, illustrates how Content Server returns the results for that query.
Table 105. Example of row-based query results
090000015973a2fc Our Life and Times Jennie Doe
090000015973a2fc Our Life and Times Carol Jones
090000015973a2fc Our Life and Times Hortense Smith
0900000123ac12f6 Life of an Amoeba James Doe
0900000123ac12f6 Life of an Amoeba Jules Doe
There is one query result object (one row) for each repeating property value returned. The returned
repeating property values for each object are not aggregated into one result object.
Effects on query syntax rules
In addition to the changes the hint causes in the returned results, including the hint also changes
some of the syntax rules for a query. The hint affects:
• Query syntax rules when a repeating property is a selected property
Repeating properties, page 123, describes how the hint affects query syntax rules when a repeating
property is selected.
• The use of repeating properties in expressions in WHERE clauses

Using DQL Hints
Using repeating properties in qualifications, page 143, describes how the hint affects the use of
repeating properties in WHERE clause qualifications.
• The behavior of the asterisk as a selected value
The asterisk (*) as a selected value, page 131, describes how the hint affects the returned values
for an asterisk.
SQL_DEF_RESULT_SET N
The SQL_DEF_RESULT_SET N is most useful on a SQL Server database. On a SQL Server database,
including SQL_DEF_RESULT_SET in a query causes the RDBMS to use default result sets to return
the query results rather than a cursor. Using default result sets is the way that Microsoft recommends
accessing a SQL Server database.
On other databases, it only controls how many rows are returned, and is identical to the
RETURN_TOP hint.
On SQL Server, using default result sets allows the RDBMS server to perform fewer logical read
operations. Table 106, page 361 shows some test examples:
Table 106. Comparison of logical reads with and without default result sets
Query Number of returned rows Number of logical
reads without DRS
Number of
logical reads
with DRS
SELECT object_name
FROM dm_document
WHERE r_object_
id=’0900014c8000210b’
1 36 5
SELECT object_name FROM
dm_document WHERE
r_creator_name=’user2’
100 1600 700
SELECT object_name FROM
dm_document WHERE
r_creator_name=’user1’
2000 22500 500
However, the reduction in I/O may be offset by higher CPU costs. Additionally, because using
default result sets requires Content Server to buffer the results of a query in memory, higher memory
use is a result. (When using default result sets, two queries cannot be open simultaneously. If a
query is issued, before another can be issued from the same session, all results from the first must be
processed. To work around this limitation, Content Server internally buffers the results of queries
that are executed using default result sets.)
To determine whether using SQL_DEF_RESULT_SETS is useful for a query, run the query with and
without the hint and trace the results using the SQL Server Profiler.
Setting N to 0 causes all results to be returned. Setting N to an integer value limits the number of
rows returned to that value.

Using DQL Hints
Using SQL_DEF_RESULT_SETS is recommended if:
• You are running on a SQL Server database.
• The query returns limited result sets.
• The query is a commonly executed query.
• Tests show that using the hint greatly improves query performance.
TRY_FTDQL_FIRST
The TRY_FTDQL_FIRST hint is useful if a query is timing out or exceeds a resource limit in the
full-text engine. When it is included in a query, the query is first executed as an FTDQL query, and if a
timeout or resource exceeded error occurs, the query is then retried as a standard query.
UNCOMMITTED_READ
Use the UNCOMMITTED_READ hint in read only queries, to ensure that the query returns quickly
even if another session is holding locks on the tables queried by the read only query.
This hint is useful only on SQL Server.
DM_LEFT_OUTER_JOIN_FOR_ACL
When this DQL hint is added and if the user is not a super user, Content Server generates an SQL
query that contains a LEFT OUTER JOIN on the dm_acl_s table. This operation improves the
performance of DQL queries issued by a non-super user by optimizing the user permit calculation
of resolved objects. This hint is effective when the user is a part of many groups. The following
DQL statement is an example of using this hint:
SELECT r_object_id FROM dm_document WHERE object_name = 'test' ENABLE
(DM_LEFT_OUTER_JOIN_FOR_ACL)
The query returns r_object_id of all documents with object name ’test’.
This hint overrides the computer variable DM_LEFT_OUTER_JOIN_FOR_ACL.
Including multiple hints limiting rows returned
If you include more than one hint that limits the rows returned by a query, the number of rows
returned is the least number of rows defined in an included hint. For example, suppose you issue the
following DQL statement:
SELECT object_name FROM dm_document ENABLE
(FETCH_ALL_RESULTS 10, RETURN_TOP 5)

Using DQL Hints
Content Server returns 5 rows for the query because the RETURN_TOP hint is the most constraining
hint in the list.
On SQL Server, if the list includes SQL_DEF_RESULT_SET, the query is always executed using
default result sets regardless of where the hint falls in the list of hints. For example, suppose you
issue the following statement:
SELECT object_name FROM dm_document ENABLE
(SQL_DEF_RESULT_SET 10, RETURN_TOP 5)
The query executes using default result sets but returns only 5 rows.
Passthrough hints
Passthrough hints are hints that are passed to the RDBMS server. They are not handled by Content
Server.
SQL Server has two kinds of hints: those that apply to individual tables and those that apply globally,
to the entire statement. To accommodate this, you can include passthrough hints in either a SELECT
statement’s source list or at the end of the statement. The hints you include in the source list must be
table-specific hints. The hints you include at the end of the statement must be global hints.
For DB2 and Oracle, include passthrough hints only at the end of the SELECT statement.
Syntax
To include a passthrough hint, you must identify the database for which the hint is valid. To
identify the target database, keywords precede the hints. The valid keywords are: ORACLE, and
SQL_SERVER. For example, the following statement includes passthrough hints for SQL Server
and DB2:
WHERE "object_name" ='test'
ENABLE (SQL_SERVER('ROBUST PLAN','FAST 4',
'ROBUST PLAN'))
For portability, you can include passthrough hints for multiple databases in one statement. The entire
list of hints must be enclosed in parentheses. The syntax is:
(database_hint_list {,database_hint_list})
where database_hint_list is:
db_keyword('hint'{,'hint})
db_keyword is one of the valid keywords identifying a database. hint is any hint valid on the specified
database.
For example:
SELECT object_name FROM dm_document doc dm_user u
WHERE doc.r_creator_name = u.user_name
ENABLE (ORACLE('RULE','PARALLEL'),
SQL_SERVER('LOOP JOIN','FAST 1')

Using DQL Hints
Error handling and debugging
It is possible to execute a DQL statement that the server considers correct DQL but is incorrect at the
RDBMS level if you incorrectly specify a passthrough hint. If this occurs, any errors returned are
returned by the RDBMS server, not Content Server. Not all databases return errors on database hints.
For example, Oracle does not return an error if a database hint is incorrect—it simply ignores the hint.
For the other databases, the error messages may be difficult to decipher.
To debug problems with queries containing database hints, you can turn on SQL tracing. By default,
Content Server runs with the last SQL trace option turned on. You can turn on full SQL tracing using
an IDfSession.setServerTraceLevel race method or the SET_OPTIONS administrative function. The
SET_OPTIONS administration method is described in SET_OPTIONS, page 317.

Appendix B
Database footprint reduction of
dmr_content objects
Beginning in release 6.6, support for true NULL values was added to Content Server, for databases
that support it. Users can create and alter their own types to take advantage of the space savings
possible with this feature, but most internal Content Server types do not use it. The one exception
is dmr_content.
In order to use true NULL storage with dmr_content, you must set the server.ini parameter,
spaceoptimize to dmr_content in a new repository:
spaceoptimize=dmr_content
Currently, dmr_content is the only internal type that you can specify. For Oracle installations, all
the properties of dmr_content will be set to SPACEOPTIMIZE. For SQL Server and DB2, only the
character and string properties and the ID properties will be set to SPACEOPTIMIZE.
For single-valued properties, the total number of bytes saved per content object is 87 bytes for Oracle,
74 bytes for SQL server, and 36 bytes for DB2.
For repeating-valued properties, the space saving is not so easy to predict. If all the repeating
properties have one value, the default value, there is no space saving since a last value of default
is always stored, not as a NULL, but as the default value. However, if there are N values
for each repeating property, each with default value, then the space saving would be about
(N-1)*(propertysize).
The details of the space saving are listed in the table, Table 107, page 365.
Table 107. Dmr_content object space savings
Property Type Bytes Saved Single/Repeating
rendition integer 4 single
other_ticket integer 4 single
resolution integer 4 single
x_range integer 4 single
y_range integer 4 single
z_range integer 4 single

Database footprint reduction of dmr_content objects
encoding char(10) 2 on SQL Server, 1 on
others
single
loss integer 4 single
transform_path char(32) 2 on SQL Server, 1 on
others
single
set_client char(64) 2 on SQL Server, 1 on
others
single
set_file chare(255) 2 on SQL Server, 1 on
others
single
set_time time 8 on SQL Server, 9 on
Oracle
single
is_offline Boolean 2 single
i_contents char(4000) 2 on SQL Server, 1 on
others
single
is_archived Boolean 2 single
index_format ID 31 on SQL Server, 16
on Oracle, 15 on DB2
single
index_parent ID 31 on SQL Server, 16
on Oracle, 16 on DB2
single
i_rendition integer 4 repeating
i_px integer 4 repeating
i_py integer 4 repeating
i_encoding char(10) 2 on SQL Server, 1 on
others, for each entry
repeating
page_modifier char(16) 2 on SQL Server, 1 on
others, for each entry
repeating
r_content_hash chare(256) 2 on SQL Server, 1 on
others
single
i_parked_state integer 4 single
other_file_size double 4 single

Appendix C
IDQL and IAPI
This appendix describes the Interactive DQL (IDQL) and IAPI utilities. The utility is useful if you
want to test short scripts or perform short, small tasks in the repository. The IAPI utility allows you
to execute Docbasic scripts.
The appendix includes the following topics:
• Using IDQL, page 367
• Using IAPI, page 371
Using IDQL
The IDQL utility is an interactive tool that lets you enter ad hoc DQL queries against a repository.
IDQL is also a useful as a tool for testing and other tasks that support an application or installation
because it allows you to run scripts and batch files.
IDQL is included and installed with Content Server. It is found in %DM_HOME%bin
($DM_HOME/bin). Copy the utility’s executable from the bin directory to a directory you can access
or copy it to your local machine. Depending on your operating system, there may be multiple IDQL
utilities. For example, on a 64–bit Windows installation, you will find both a 32–bit and 64–bit version
of the utility, idql32 and idql64.
If you copy it to your local machine, you must also copy the client library (DMCL) to your local
machine if it is not already there.
Starting IDQL
You can start the utility from the operating system prompt or an icon. To start the utility from the
operating system prompt, use the following syntax:
On Windows (use idql32 for the 32–bit version, idql64 for the 64–bit version):
idql32 [-n] [-wcolumnwidth]
[-Uusername|-ENV_CONNECT_USER_NAME]
[-Pos_password|-ENV_CONNECT_PASSWORD][-Ddomain]
repository|-ENV_CONNECT_DOCBASE_NAME [-Rinput_file]
[-X][-E][-Ltrace_level][-zZ][-Winteger][-Ksecure_mode]
On UNIX:

IDQL and IAPI
idql [-n] [-wcolumnwidth]
[-Uusername|-ENV_CONNECT_USER_NAME]
[-Pos_password|-ENV_CONNECT_PASSWORD][-Ddomain]
repository|-ENV_CONNECT_DOCBASE_NAME [-Rinput_file]
[-X][-E][-Ltrace_level][-zZ][-Winteger][-Ksecure_mode]
The order of the arguments on the command line is not important.
If you created an icon for the utility, double-click the icon and enter your name and password. To
invoke the utility with additional command line arguments, use the icon’s Properties to add the
arguments to the command line before you double-click the icon.
Table 108, page 368, describes the command line arguments.
Table 108. IDQL command line arguments
-n Removes numbering and the prompt symbol from
input files. This argument is used primarily for
scripts.
-wcolumnwidth Sets the screen width for output. The default width
is 80 characters.
-Uusername
or
-ENV_CONNECT_USER_NAME
Identifies the user who is starting the session.
You can include the user name on the command
line with the -U flag, or you can use the
-ENV_CONNECT_USER_NAME argument, which
directs the system to look for the user name in the
environment variable DM_USER_NAME.
-Pos_password
or
-ENV_CONNECT_PASSWORD
Identifies the user’s operating system password.
You can include the password on the command
line with the -P flag, or you can use the
-ENV_CONNECT_PASSWORD argument, which
directs the system to look for the password in the
environment variable DM_PASSWORD.
If the user does not specify a password, the utility
prompts for the password. If the user enters the -P
flag without the password, IDQL uses the default
password, NULL. Passwords are case sensitive.
-Ddomain Identifies the user’s domain.
repository
or
-ENV_CONNECT_DOCBASE_ NAME
Identifies the repository. You can include the
repository name on the command line, or you
can use the -ENV_CONNECT_DOCBASE_NAME
argument, which directs the system to look for
the repository name in the environment variable
DM_DOCBASE_NAME.
-Rinputfile Specifies the name of an input file containing DQL
queries. This is an optional argument.
-X Allows prompt for domain.

IDQL and IAPI
-E Turns on echo.
-Ltrace_level Turns tracing on for the session. trace_level can be
any of the valid trace levels described in the Trace
method description.
-zZ Specifies Windows NT Unified Logon. This option
overrides any user name and password specified on
the command line.
-Winteger
or
-winteger
Sets the maximum column width when displaying
results. integer is interpreted as number of bytes. The
default is 2000 bytes.
Any value which exceeds the maximum row width
is truncated when displayed.
-Ksecure_mode Defines the type of connection you want to establish.
Valid values are:
• secure
• native
• try_secure_first
• try_native_first
The default is native.
After you start IDQL, the utility returns with its prompt. The IDQL utility has a numerical prompt
that begins with 1 and increments each time you press the carriage return or Enter key to move to
a new line on your display.
The IDQL commands
IDQL recognizes the commands listed in Table 109, page 369.
Table 109. IDQL commands
Command Description
go Executes a DQL statement and clears the query buffer.
clear Clears the query buffer.
quit Exits the utility.

IDQL and IAPI
Command Description
describe Provides a description of a specified object type or registered
table. The syntax formats for this command are:
describe [type] type_name
describe table table_name
@scriptfile Executes the specified script.
shutdown Issues a Shutdown method that shuts down the Content
Server gracefully. You must be a superuser or system
administrator to use this command. When shutdown is
complete, the utility exits.
trace Turns on tracing for the current session. By default, tracing is
set to level 5.
-Winteger
or
-winteger
Sets the maximum column width when displaying results.
integer is interpreted in bytes. The default is 2000 bytes.
Setting this affects the results display for all statements issued
after this command.
Each command must be entered on a separate line. A command cannot appear on the same line as a
query, nor can two or more commands appear on one line.
Entering queries
Enter a query by typing it at the IDQL prompt. The following IDQL session shows two queries
typed at the prompt:
1> BEGIN TRAN
2> go
3> SELECT *
4> FROM Engineering
5> WHERE Engineer LIKE 'Smith, %'
6> go
The query processor interprets semicolons (;) as statement terminators and ignores everything that
follows the semicolon on the line.
When you enter a query, IDQL places the query in a buffer. You can then execute the query or
clear the buffer.
The go command executes a query. Each go command can execute only one query. You cannot
execute multiple queries with one go command.
There are two ways to execute queries in a batch using IDQL:
• Use the input file option in the IDQL command line.
• Use the @scriptfile command.
The input file is a file that contains queries. You must place a go command after each query in the file,
and you must terminate the file with a quit command.

IDQL and IAPI
The results of each query are returned to standard output by default.
The @scriptfile command executes a specified script, which can include DQL queries. Like an input
file, queries in the script must be separated by go commands. You are not required to use a quit
command to terminate a script executed with @scriptfile.
Clearing the buffer
The following excerpt enters a query and then clears it from the buffer:
1> SELECT *
2> FROM Engineering
3> WHERE Engineer LIKE 'Smith, %'
4> clear
Entering comments
To enter a comment in an IDQL session, start the comment line with or -- .
Stopping IDQL
To close an IDQL session, enter the quit command at the IDQL prompt.
Using IAPI
The IAPI utility is a tool that you can use to execute Docbasic scripts. This utility is included and
installed with Content Server. It is found in %DM_HOME%bin ($DM_HOME/bin). Depending
on your operating system, there may be multiple IAPI utilities. For example, on a 64–bit Windows
installation, you will find both a 32–bit and 64–bit version of the utility, iapi32 and iapi64.
Starting IAPI
You can start the utility from the operating system prompt or an icon. To start the utility from the
operating system prompt, use the following syntax:
On Windows (use iapi32 for the 32–bit version, iapi64 for the 64–bit version):
iapi32 [-Uusername|-ENV_CONNECT_USER_NAME]
[-Pos_password|-ENV_CONNECT_PASSWORD] [-Llogfile] [-X][-E]
[-V-][-Q] repository|-ENV_CONNECT_DOCBASE_NAME [-Rinputfile]
[-Ftrace_level][-Sinit_level][-Iescape_character][-zZ]
[-Winteger][-Ksecure_mode]
On UNIX:
iapi [-Uusername|-ENV_CONNECT_USER_NAME]

IDQL and IAPI
[-Pos_password|-ENV_CONNECT_PASSWORD] [-Llogfile] [-X][-E]
[-V-][-Q] repository|-ENV_CONNECT_DOCBASE_NAME [-Rinputfile]
[-Ftrace_level][-Sinit_level][-Iescape_character][-zZ]
[-Winteger][-Ksecure_mode]
The order of the arguments on the command line is not important.
If you created an icon for the utility, double-click the icon and enter your name and password. To
invoke the utility with additional command line arguments, use the icon’s Properties to add the
arguments to the command line before you double-click the icon.
Table 110, page 372, describes the command line arguments.
Table 110. IAPI command line arguments
-Uusername
or
-ENV_CONNECT_USER_NAME
Identifies the user who is starting the session.
You can include the user name on the command
line with the -U flag or you can use the
-ENV_CONNECT_USER_NAME argument, which
directs the system to look for the user name in the
environment variable DM_USER_NAME.
-Pos_password
or
-ENV_CONNECT_PASSWORD
Identifies the user’s operating system password.
You can include the password on the command
line with the -P flag, or you can use the
-ENV_CONNECT_PASSWORD argument, which
directs the system to look for the password in the
environment variable DM_PASSWORD.
If the user does not specify a password, the utility
prompts for the password.
-Llogfile Directs the server to start a log file for the session.
You must include the file name with the argument.
You can specify a full path.
-X Allows prompt for domain.
-E Turns on echo.
-V- Turns verbose mode off. The utility runs in verbose
mode by default. If you turn verbose mode off,
the system does not display the IAPI prompt, error
messages, or any responses to your commands
(return values or status messages).
You can also change modes from the IAPI prompt
(refer toIAPI commands, page 374).
-Q Directs the utility to provide a quick connection test.
The utility attempts to connect and exits with a status
of zero if successful or non-zero if not. It provides
error messages if the attempt is not successful.

IDQL and IAPI
repository
or
-ENV_CONNECT_DOCBASE_ NAME
Identifies the repository. You can include the
repository name on the command line, or you
can use the -ENV_CONNECT_DOCBASE_NAME
argument, which directs the system to look for
the repository name in the environment variable
DM_DOCBASE_NAME.
-Rinputfile Identifies an input file containing API methods.
-Ftrace_level Turns tracing on for the session. trace_level can be any
of the valid trace levels described in the Trace method
description.
-Sinit_level Defines the connection level established when you
start the API session. Valid init levels and their
meanings are:
api, which opens an API session. User must issue an
explicit Connect to start a session.
connect, meaning undetermined
standard, which opens a session
-Iescape_character Defines an escape character for certain symbols such
as a forward slash (/).
-zZ Specifies Windows NT Unified Logon. This option
overrides any user name and password specified on
the command line.
-Winteger
or
-winteger
results returned from a query entered with a ?
command. integer is interpreted as number of bytes.
The default is 2000 bytes.
Any value which exceeds the maximum row width is
truncated when displayed.
-Ksecure_mode Defines the type of connection you want to establish.
Valid values are:
• secure
• native
• try_secure_first
• try_native_first
The default is try_native_first.
When you issue the iapi command, the system returns a status message that tells you it is
connecting you to the specified database. This is followed by a message giving you the session ID for
your session and then the utility’s prompt (API>).

IDQL and IAPI
IAPI commands
The IAPI commands affect how the IAPI utility functions. These commands do not affect the
repository. The utility accepts the commands listed in Table 111, page 374.
Table 111. IAPI commands
Command Description
$logfile Outputs to the specified file
@scriptfile Reads input from the specified file
%scriptfile Outputs a script
-V[+|-] Turns verbose mode on (+) or off (-).
By default, verbose mode is on. If you turn verbose
mode off, the system does not display the IAPI
prompt, error messages, or any responses to your
commands (return values or status messages).
-Winteger
or
-winteger
results returned from a query entered with a ?
command. integer is interpreted in bytes. The
default is 2000 bytes.
Setting this affects the results display for all methods
issued after this command.
? Allows you to enter a query or collection ID and
formats the output as a table.
The ? command allows you to enter a DQL SELECT statement in IAPI. When used with a SELECT
statement, the command is the equivalent of a Readquery method. The syntax is:
API>?,c,select_query_string
For example:
API>?,c,select r_object_id, object_name from dm_document
The results are formatted in a table-like block. For example:
r_object_idobject_name
-----------------------
090015c38000013Bfoo.txt
090015c3800002e4spam.html
090015c3800001c2fah.txt
You can also use the ? command to retrieve a collection’s content in one block. For example, suppose
you issue a readquery method, which returns a collection:
API>readquery,c,select r_object_id, object_name from dm_document
...
API>q0
Instead of using a Next method to retrieve each row of the collection, you can use ? to retrieve
them all at once:
API>?,c,q0
...

IDQL and IAPI
r_object_idobject_name
-----------------------
090015c38000013Bfoo.txt
090015c3800002e4spam.html
090015c3800001c2fah.txt
Entering method calls
Note: This use of IAPI is deprecated, as the DMCL that was the primary API for Content Server in
versions prior to Documentum 6 is deprecated in Documentum 6.
Call methods directly from the IAPI prompt by typing the method name and its arguments as a
continuous string, with commas separating the parts. For example, the following command creates a
new folder:
API> create,s0,dm_folder
If any parameter value contains a comma, you must enclose that parameter in single quotes. For
example, the following command creates a new document and assigns it a content file:
API> create,s0,dm_document
API> setfile,s0,last,'/home/janed/temp.doc,1',text
The name of the content file is in single quotes because it contains a comma.
When the method completes, the utility displays the return value. For example, the Create method
returns the object ID of the newly created object. So if you issue the Create command shown in the
previous example, the utility displays the object ID of the new folder:
API> create,s0,dm_folder
. . .
0b6385F20000561B
When you execute a method that assigns a value to an attribute, you do not specify the value as
part of the command syntax. Instead, IAPI prompts you for that value. For example, the following
exchange sets the name of the folder object created in the previous example:
API> set,s0,last,object_name
SET> cake_recipes
. . .
OK
The keyword last refers to the last created, fetched, checked in/out, or retrieved object ID (in this
case, the folder ID).
Entering comments
Enter a comment in an IAPI session by preceding it with the pound sign (#). This is useful when you
want to use scripts with IAPI.
Shutting down and restarting Content Server
The following command shuts down the Content Server:

IDQL and IAPI
API> apply,c,NULL,SHUTDOWN_WORKFLOW_AGENT,<TIMEOUT>,<|>,<timeout_value>
The TIME_OUT parameter is optional. If the parameter is not provided, then the Content Server
will take the value specified in server_config > system_shutdown_timeout as the timeout value for
shutdown.
If the TIME_OUT parameter is specified, the Content Server is shutdown depending on the values.
• > 0: The workflow agent will wait for the specified value to do a graceful shutdown and when
the timeout value is reached, it will shutdown ungracefully.
• = 0: The workflow agent will shutdown ungracefully.
• < 0: The workflow agent will do a graceful shutdown
The following command restarts the Content Server:
API> apply,c,NULL,START_WORKFLOW_AGENT
Quitting an IAPI session
To exit from IAPI, enter quit at the IAPI prompt:
IAPI> quit

Appendix D
Implementing Java Evaluation of
Docbasic Expressions
This appendix describes how to implement Java evaluation of Docbasic expressions defined for check
constraints or value assistance in the data dictionary˙. The appendix includes the following topics:
• Docbasic expression handling by Content Server, page 377
• How DFC Version 6 and later handles the expressions, page 378
• Migrating the expressions for pre-6 clients, page 378
• Disabling or re-enabling Java evaluation, page 380
• Docbasic expression components support, page 381
Docbasic expression handling by Content
Server
This section describes how Content Server describes how Docbasic expressions defined for check
constraints or value assistance in the data dictionary are stored in the repository
Figure 2, page 377, illustrates the architecture.
Figure 2. Repository storage of Docbasic expressions for object types
For each object type that has Docbasic expressions defined as check constraints or in conditional value
assistance for one or more properties, Content Server creates one expr code object and a number of
func expr objects. Each func expr objects records one Docbasic expression. The expr code object
contains the compiled code and pcode that implements the expressions recorded in the func expr
objects. The routine_id property in the func expr objects points to the expr code object.

Implementing Java Evaluation of Docbasic Expressions
How DFC Version 6 and later handles the
expressions
DFC Version 6 and later generates the Java code equivalent to the Docbasic expressions defined
for check constraints or value assistance at runtime. The result is cached in memory for the life of
the DFC process.
Migrating the expressions for pre-6 clients
Prior to Version 6, DFC required that the expressions be manually migrated to Java if you wished to
have the expressions evaluated as Java code. If you are running pre-6 clients, use the information
in this section to migrate the Docbasic expressions to Java code. This section describes how that
implementation was stored in the repository and the methods used to migrate the expressions.
Manually migrating a Docbasic expression for DFC Version 6 and later clients is not useful. DFC
Version 6 (and later) automatically generates Java code at runtime for Docbasic expressions and will
ignore the results of manual migration.
Repository storage of migrated expressions
Migrating an expression manually creates additional repository objects that are associated with the
objects created by Content Server for the expression.
If you migrate an expression to Java, the migration method compiles the expression into Java code. If
the expression is the first expression compiled into Java for the object type, the migration method
also creates a dmc_jar object and a dmc_validation_module object. The compiled code is stored with
the dmc_jar object. The jar object is related to the validation module object through a relationship
named dmc_module_to_jar.
The validation module object points back to the associated expr code object through a property and is
related to the func expr objects using dmc_validation_relation objects. Validation relation objects are
a subtype of dm_relation. The name of the relationship represented by validation relation objects is
dmc_expr_to_module. Both the validation module and validation relation objects have properties
used to manage the Java evaluation. Figure 3, page 378, illustrates this architecture.
Figure 3. Repository storage of manually migrated Docbasic expressions
If additional Docbasic expressions are added later for an object type, the migration method updates
the validation module and jar objects.
Migrating expressions to Java is currently a manual operation. EMC Documentum provides two
migration methods, described in Migrating Docbasic expressions to Java, page 379. You can migrate
expressions already defined in the data dictionary. You can migrate new expressions after the
expressions are defined using an ALTER TYPE or CREATE TYPE statement.

Migrating Docbasic expressions to Java
There are two methods that create compiled Java code for Docbasic expressions stored with expr code
objects. These methods are:
• dmc_MigrateDbExprsToJava
Use this method if you want to migrate expressions defined for multiple object types.
• dmc_MigrateDbExprsToJavaForType
Use this method if you want to migrate expressions added to a single object type (or the type’s
properties) using an ALTER TYPE or if you want to migrate expressions defined while creating
a new type using CREATE TYPE.
Arguments for both methods let you select which Docbasic expressions to migrate to Java code. You
can choose to compile only new Docbasic expressions, or those that failed compilation previously, or
all expressions within the scope of the method.
The methods are executed using either a DFC script or using a DO_METHOD administration
method. The DO_METHOD syntax is:
dmAPIGet("apply,session,NULL,DO_METHOD,METHOD,S,migration_method_name,SAVE_
RESULTS,B,T,ARGUMENTS,S,argument_list
migration_method_name is either dmc_MigrateDbExprsToJava or dmc_MigrateDbExprsToJavaForType.
Use a space to separate arguments in argument_list.Table 112, page 379, lists the valid arguments for
argument_list. .
Table 112. Arguments for the Docbasic expression migration methods
-docbase docbase_name Name of the repository that contains the expressions you want to
migrate. This must a 5.3 (or later) repository.
-user user_name Name of a user who has at least Sysadmin privileges in the
specified repository.
-ticket login_ticket String containing a login ticket.
Note: For information about login tickets and generating those
tickets, refer to the EMC Documentum Content Server Administration
and Configuration Guide.
-select selection_choice Identifies which expressions you wish to migrate. selection_choice is
one of:
• new_only, includes only new expressions, that is, those for
which migration has not been previously attempted
• new_and_failed, includes all new expressions and those for
which migration previously failed
• new_and_disabled, includes all new expressions and those that
do not have a corresponding enabled Java version
• all, includes all expressions

Any existing Java code is replaced if currently migrated
expressions are recompiled.
For dmc_MigrateDbExprsToJava, expressions for all object types are
considered for inclusion. For dmc_MigrateDbExprsToJavaForType,
only expressions defined for the object type identified in the type
argument are considered for inclusion.
-maxerrs value Maximum number of errors that can occur before code generation
stops. By default, there is no maximum.
Setting this argument is recommended only if there are large
numbers of expressions to compile.
-listExprsOnly value Whether to actually perform the migration or only create a file that
records the expressions selected for migration. Setting value to true
directs the method to only create the file. Setting the value to false
directs the method to migrate the expressions.
If you set this argument to true, you must include the
SAVE_RESULTS argument in the DO_M,ETHOD command line.
-type type_name Name of the object type. Include this argument only if you are
executing the dmc_MigrateDbExprsToJavaForType method. Use
the type’s internal name; for example: dm_document.
It is strongly recommended that you include the SAVE_RESULTS argument on the DO_METHOD
command line regardless of the -listExprsOnly setting. The content of the generated results document
varies depending on the circumstances of the method’s execution:
• If -listExprsOnly is set to true, the results document contains a list of all expressions selected for
migration.
• If -listExprsOnly is false and non-fatal errors occur during migration, the results document
contains the errors.
• If -listExprsOnly is false and fatal errors occur during migration, the results document contains
the exception stack trace embedded in the HTML error message.
• If -listExprsOnly is false and no errors occur during migration, the results document contains
only the text "There were no errors.”
Both methods return 0 unless a fatal error occurred during execution. If a fatal error occurs, the
methods return 1. Fatal errors are reported using the standard method server mechanism.
Disabling or re-enabling Java evaluation
It is possible to disable Java evaluation of a particular Docbasic expression. If you disable Java
evaluation of an expression, the Docbasic library is used to evaluate the expression, rather than the
compiled Java code.

To disable Java evaluation, execute the dmc_SetJavaExprEnabled method. This method has an
argument that turns evaluation of a specified expression on or off. The method is executed by a DFC
script or a DO_METHOD administration method. Here is the syntax for the DO_METHOD call:
apply,session,NULL,DO_METHOD,METHOD,S,dmc_SetJavaExprEnabled,SAVE_
RESULTS,B,T,ARGUMENTS,S,argument_list
Table 113, page 381, lists the arguments for dmc_SetJavaExprEnabled. All arguments for this method
are required. Use a space to separate arguments in argument_list.
Table 113. dmc_SetJavaExprEnabled arguments
-docbase repository_name Name of the repository that contains the Java expression
implementation you want to disable or re-enable.
-user user_name Name of the user performing the operation. Use the user’s login name.
The user must have at least Sysadmin privileges in the repository.
-ticket login_ticket A string containing a login ticket generated for the specified user.
-func_expr_id
func_expr_obj_id
Object ID of the dm_func_expr object representing the Docbasic
expression
-enable value Set this to true to enable the Java evaluation of the Docbasic expression
identified in the -func_expr_id argument. Or, set it to false, to disable
Java evaluation of that expression.
Docbasic expression components support
This section lists and briefly describes the Docbasic components, such as operators, functions, and
constants, that are supported for Java migration.
Operators
Table 114, page 381, lists the Docbasic operators for which Java evaluation is supported.
Table 114. Docbasic operators supported by Java evaluation
Operator Operation performed
& String concatenation
* Multiplication
+ Addition or concatenation
– Subtraction
/ Floating point division
< Comparison (less than)

Operator Operation performed
<= Comparison (less than or equal to)
<> Comparison (not equal to)
= Comparison (equal to)
> Comparison (greater than)
>= Comparison (greater than or equal to)
Integer division
^ Exponentiation
And Logical or binary conjunction
Eqv Logical or binary equivalence (not Xor)
Imp Logical or binary implication
Is Compares two operands and returns true if they refer to the same
object; otherwise, returns false.
Because objects are not supported, using this operator always results in
a type mismatch error at runtime.
Like Compares two strings and returns true if the expression matches the
given pattern; otherwise, returns false.
Mod Returns the remainder of the expression_1/expression_2 as whole number
Not Returns either a logical or binary negation of an expression
Or Logical or binary disjunction on two expressions
Xor Exclusive Or operation
Supported functions for Java evaluation
Table 115, page 382, lists the Docbasic functions for which Java evaluation is supported.
Table 115. Docbasic functions supported for Java evaluation
Function Description
abs(expr) Returns the absolute value of expr
Asc(text$) Returns the integer containing the numeric code for the first
character of the text$ (0–255)
AscB or AscW Returns the byte or wide-character equivalents of Asc
Atn(number) Returns the inverse tangent of number
CBool(expr) Converts the expr to a Boolean

CDate(expr) or CVDate(expr) Converts the expr to a Date
Note: If expr is a string, it is expected that the date is in canonical
form unless a specific date format string was passed to the
validation method.
CDbl(expr) Converts expr to a Double
Choose(index, expr1,
expr2,...exprN)
Returns the expression whose position in the list of expr
arguments matches the value specified index. If none is found,
returns null.
Chr(code) or Chr$(code) Returns the character whose value is code
The Java implementation returns a String variant in both functions.
CInt(expr) Converts expr to a Docbasic integer.
The Java implementation converts expr to a Java short.
CLng(expr) Converts expr to a Docbasic long
The Java implementation converts expr to a Java int.
Cos(angle) Returns the cosine of angle where angle is assumed to be
expressed in radians
CSng(expr) Converts expr to a Docbasic Single
The Java implementation converts expr to a Java float.
CStr(expr) Converts expr to a String
CVar(expr) Converts expr to a Variant
Date Returns the current date as a Date Variant
Date$ Returns the current date as a String
The Java implementation will always be in canonical form or in
the format passed to the validation method.
DateAdd(interval$,
increment&, date)
Increments the date field identified in interval$ of the given date
by the amount specified in interval&
DateDiff(interval$, date1,
date2)
Returns a Date variant representing the number of given time
intervals between date1 and date2
DatePart(interval$, date) Returns an Integer representing a specific part of a date/time
expression
DateSerial(year, nonth, day) Returns a Date variant representing the specified date
DateValue(dateString$) Returns a Date variant representing the date specified by
dateString$
Day(date) Returns the day of the month represented by date
Exp(val) Returns the value of e raised to the power of val
Fix(number) Returns the integer part of number

Format or Format$(expr,
userFormat)
Returns a string formatted to user specification (supports both
named and unnamed formats)
Hex[$](number) Returns a String containing the hexadecimal equivalent of number
Hour(time) Returns the hour of the day encoded in time
If(condition, TrueExpression,
FalseExpression)
Returns TrueExpression if condition is true; otherwise, returns
FalseExpression
InStr([start,] search, find
[,compare])
Returns the first character position of string "find” within string
"search”
InStrB([start,] search, find
[,compare])
Returns the first character position of string "find” within string
"search”
Int(number) Returns the integer part of number
IsDate(expr) Returns true if expr can be legally converted to a date; otherwise,
returns false
IsNull(expr) Returns true if expr is a Variant variable that contains no valid
data; otherwise, returns false
IsNumeric(expr) Returns true if expr can be converted to a number; otherwise,
returns false
Item$(text$, first, last
[,delimiters$])
Returns all items between first and last within the specified
formatted text list
ItemCount(text$
[,delimiters$])
Returns an integer containing the number of items in the specified
delimited list
LBound(arrayVar
[,dimension])
Returns an integer containing the lower bound of the specified
dimension of the specified array variable
LCase[$](text) Returns the lowercase equivalent of text
Left[$](text, nChars) or
LeftB[$](text, nChars)
Returns the leftmost nChars characters (Left and Left$) or bytes
(LeftB and LeftB$) from text
Note: Use LeftB with caution. Multibyte characters can be split in
the middle, resulting in the construction of illegal strings.
Len(expr or lenB(expr) Returns the number of characters in expr or the number of bytes
required to store expr
Line$(text$, first [,last]) Returns a string containing a single line or a group of lines
between first and last
LineCount(text$) Returns an integer representing the number of lines in text$
Log(number) Returns a double representing the natural logarithm of number
LTrim[$](text) Returns text with the leading spaces removed
Mid[$](text, start [,length]) or
MidB[$](text, start [,length])
Returns a substring of the specified text, beginning with start for
length number of characters or bytes
Note: Use MidB with caution. Multibyte characters can be split in
the middle, resulting in the construction of illegal strings.

Minute(time) Returns the minute of the day encoded in the time argument
Month(date) Returns the month of the date
Now Returns a date variant representing the current date and time
Oct(number) or Oct$(number) Returns a string containing the octal equivalent of the specified
number
Random(min,max) Returns a long value greater than or equal to min and less than or
equal to max
Right[$](text,nChars) or
RightB[$](text,nChars)
Returns the rightmost nChars characters or bytes from the text.
RightB and RightB$ are used to return byte data from strings
containing byte data.
Note: Use RightB with caution. Multibyte characters can be split
in the middle, resulting in the construction of illegal strings.
Rnd[(number)] Returns a random Single number between 0 and 1
RTrim[$](text) Returns text with the trailing spaces removed
Second(time) Returns the second of the day encoded in the time argument
Sgn(number) Returns an integer indicating whether a number is less than,
greater than, or equal to 0
Sin(angle) Returns a double value specifying the sine of angle
Space[$](NumSpaces) Returns a string containing the specified number of spaces
Sqr(number) Returns a double representing the square root of number
Str[$](number) Returns a string representation of number
StrComp(string1, string2
[,compare])
Returns an integer indicating the result of comparing string1 and
string2
String[$](number, [CharCode
| text$])
Returns a string of length number consisting of a repetition of the
specified filler character
Swith(condition1, expr1
[,condition2,expr2...
[condition7,expr7]])
Returns the expression corresponding to the first true condition.
Tan(angle) Returns a double representing the tangent of angle
Time/Time$ Returns the system time as a string or as a Date variant
In the Java implementation, the time is in canonical form.
Timer Returns a Single representing the number of seconds that have
elapsed since midnight
TimeSerial(hour, minute,
second)
Returns a Date variant representing the given time with a date
of zero

TimeValue(time_string$) Returns a date variant representing the time contained in
time_string$
The date must be in canonical form or in a format that conforms to
the format string passed to the validation method.
Trim[$](text) Returns a copy of text with leading and trailing spaces removed
TypeName(varname) Returns the type name of the specified variable
UBound(arrayVar
[,dimension])
Returns an integer containing the upper bound of the specified
dimension of the specified array variable
UCase[$](text) Returns the uppercase equivalent of the specified string
Val(numberString) Converts a given string expression to a number
Vartype(variable) Returns an integer representing the type of data in variable
Weekday(date) Returns an integer value representing the day of the week given
by date
Sunday is 1, Monday is 2, and so forth.
Word$(text$, first [,last]) Returns a string containing a single word or sequence of words
between first and last
WordCount(text$) Returns an integer representing the number of words in text$
Year(date) Returns the year of the date encoded in the date argument
The value returned is between 100 and 9999, inclusive.
Unsupported functions for Java evaluation
The functions listed in Table 116, page 386 are not supported for Java evaluation. For detailed
information about these functions, refer to the Docbasic documentation.
Table 116. Docbasic functions not supported for Java evaluation
Names of unsupported functions
CCur(expr) EOF GetSession Pmt(Rate, NPer, Pv, Fv,
Due)
Clipboard$ Erl GetSetting([appname],
section, key[, default])
PPmt(rate, Per, Nper,
Pv, Fv, Due)
Command or
Command$
Err Input, Input$, InputB,
and InputB$(...)
PrinterGetOrientation
CreateObject(ClassID) Error and
Error$(errNum)
InputBox and
InputBox$
PrintFile(filename$)
CurDir and
CurDir$(drive)
External(pathOrOb-
jID)
IPmt(Rate, Per, Nper,
Pv, Fv, Due)
Pv(reate, NPer, Pmt,
Fv, Due)

Names of unsupported functions
CVErr FileAttr(fileNum, attr) IRR(ValueArray(),
Guess)
Rate(NPer, Pmt, Pv, Fv,
Due, Guess)
DDB(Cosot, Salvage,
Life, Period)
FileDateTime(file-
name)
IsError(expr) ReadIni$(section$,
item$[, filename$])
DDEInitiate(applica-
tion$, topic$)
FileExists(filename) IsMissing(variable) SaveFilename$[([ti-
tle$[, extensions$]])]
DDERequest[$](chan-
nel, DataItem$)
FileLen(filename) Loc(filenumber) Seek(filenumber)
Dir[$][(filespec$[,prop-
erties])]
FileParse(filename, op) Lof(filenumber) SetAppInfo(section$,
entry$, string$,
filename$)
DiskFree&([drive$]) FileType(filename) MacID(text$) Shell(command$[,
WindowStyle])
dmAPIExec FreeFile MIRR(valArray, rate,
reinvestRate)
ShellSync(command)
dmAPIGet Fv(Rate, Nper, Pmt,
Pv, Due)
MsgBox(msg) Spc(numspaces)
dmAPISet GetAppInfo(section$,
entry$, filename$)
NPer(rate, Pmt, Pv, Fv,
Due)
SYD(cost, salvage, life,
period)
dmExit GetAttr(filename$
[,class$])
Npv(rate, valArray) Tab(column)
Environ or
Environ$(varName
| varNum)
GetOption(name$ | id) OpenFilename$([([ti-
tle$[, extensions$]])]
TypeOf<objectVar> Is
ObjectType
Supported constants
Table 117, page 387, lists the Docbasic constants that are supported for Java evaluation.
Table 117. Docbasic constants supported for Java evaluation
Constant Description
ebBoolean Integer representing Boolean datatype
ebCurrency Integer representing Currency type
ebDataObject Integer representing a data object
ebDate Integer representing a Date type
ebEmpty Integer representing an Empty variant
ebError Integer representing an Error variant
ebInteger Integer representing the Integer type

Constant Description
ebLong Integer representing the Long type
ebNull Integer representing Null variant type
ebObject Integer representing the OLE automation object type
ebSingle Integer representing the Single type
ebString Integer representing the String type
ebVariant Integer representing the Variant type
Empty Constant representing an initialized variant
False Boolean constant
Nothing Null Object reference
Null Explicit Docbasic value distinguishable from Empty or Nothing. (Refer
to the Docbasic documentation for complete information.)
Pi Value of Pi
True Boolean constant
Unsupported constants
Table 118, page 388, lists the constants that are not supported for Java evaluation.
Table 118. Constants not supported for Java evaluation
Constant names
ebCancel ebNone ebSystem
ebCritical ebNormal ebSystemModal
ebHidden ebReadOnly ebVolume
Implicit objects
The implicit objects provided by Docbasic, such as the implicit object named Basic, are not supported
for use in expressions that are compiled for Java evaluation.

Appendix E
DQL Quick Reference
This appendix contains only the formal syntax descriptions of the DQL statements and the DQL
reserved words. For a full description of each statement, refer to Chapter 2, DQL Statements.
The DQL statements
This section contains the formal syntax for each of the DQL statements.
Abort
ABORT [TRAN[SACTION]]
Alter Aspect
ALTER ASPECT aspect_name ADD
[[NO]OPTIMIZEFETCH]
ALTER ASPECT aspect_name ADD_FTINDEX ALL
ALTER ASPECT aspect_name DROP_FTINDEX ALL
ALTER ASPECT aspect_name ADD_FTINDEX property_list
ALTER ASPECT aspect_name DROP_FTINDEX property_list
Alter Group
ALTER GROUP group_name ADD members
ALTER GROUP group_name DROP members
ALTER GROUP group_name SET ADDRESS email_address

DQL Quick Reference
Alter Type
type_modifier_list [PUBLISH]
MODIFY (property_modifier_clause)[PUBLISH]
ADD property_def {,property_def}[PUBLISH]
DROP property_name {,property_name}[PUBLISH]
ALTER TYPE type_name ALLOW ASPECTS
ADD|SET|REMOVE DEFAULT ASPECTS aspect_list
ALTER TYPE type_name ENABLE PARTITION
ALTER TYPE type_name SHAREABLE [PUBLISH]
Begin Tran
BEGIN TRAN[SACTION]
Change...Object
CHANGE current_type OBJECT[S] TO new_type[update_list]
[IN ASSEMBLY document_id [VERSION version_label] [DESCEND]]
Commit
COMMIT [TRAN[SACTION]]
Create Group
CREATE GROUP group_name [WITH] [ADDRESS mail_address]
[MEMBERS member_list]
Create...Object
CREATE type_name OBJECT update_list

DQL Quick Reference
Create Type
CREATE TYPE type_name [(property_def {,property_def})]
[WITH] SUPERTYPE parent_type
To create a partitionable type:
CREATE PARTITIONABLE TYPE type_name
[WITH] SUPERTYPE NULL
To create a shareable type:
CREATE SHAREABLE TYPE type_name
[WITH] SUPERTYPE parent_type [PUBLISH]
To create a lightweight type:
CREATE LIGHTWEIGHT TYPE type_name
SHARES shareable_type
[AUTO MATERIALIZATION |
MATERIALIZATION ON REQUEST |
DISALLOW MATERIALIZATION]
[FULLTEXT SUPPORT [
NONE |
LITE ADD ALL
BASE ADD ALL |
]
[PUBLISH]
Delete
DELETE FROM table_name WHERE qualification
Delete...Object
DELETE [PUBLIC]type_name[(ALL)]
[correlation_variable]
[WITHIN PARTITION (partition_id {,partition_id})
OBJECT[S]

DQL Quick Reference
Drop Group
DROP GROUP group_name
Drop Type
DROP TYPE type_name
Execute
EXECUTE admin_method_name [[FOR] object_id]
[WITH argument = value {,argument = value}]
Grant
GRANT privilege {,privilege} TO users
Insert
INSERT INTO table_name [(column_name {,column_name})]
VALUES (value {,value}) | dql_subselect
Register
REGISTER TABLE [owner_name.]table_name (column_def {,column_def})
[[WITH] KEY {column_list)]
[SYNONYM [FOR] 'table_identification']
Revoke
REVOKE privilege {,privilege} FROM users
Select
The syntax for a standard query is:
[WITHIN PARTITION (partition_id{,partition_id})
| IN DOCUMENT clause

DQL Quick Reference
| IN ASSEMBLY clause]
[SEARCH [FIRST|LAST]fulltext_search_condition
[GROUP BY value_list]
[HAVING qualification]
[UNION dql_subselect]
[ORDER BY value_list]
where the IN DOCUMENT clause is:
IN DOCUMENT object_id [VERSION version_label]
[DESCEND][USING ASSEMBLIES]
[WITH binding condition]
[NODESORT BY property {,property} [ASC|DESC]]
and the IN ASSEMBLY clause is:
IN ASSEMBLY object_id [VERSION version_label]
[NODE component_id][DESCEND]
The syntax for an FTDQL SELECT statement is:
[SEARCH fulltext_search_condition
[ORDER BY SCORE]
Table 119, page 393, lists the constraints imposed on the clauses in an FTDQL query.
Table 119. Summary of FTDQL query rules
Applicable to Rule
FTDQL queries in general An FTDQL query must contain one of the following:
• SEARCH clause
• One of: SCORE, SUMMARY, or TEXT as a selected value
• ENABLE(FTDQL)
The query may not contain:
• An IN DOCUMENT or IN ASSEMBLY clause
• The FIRST or LAST keyword in a SEARCH clause
• A UNION, GROUP BY, or HAVING clause
selected values list The selected values list must satisfy the following rules:
• Only the following keywords are acceptable: CONTENTID,
ISCURRENT, OBJTYPE, SCORE, SUMMARY, SYSOBJ_ID,
TEXT, and THUMBNAIL_URL
• Literals or expressions of any type may not be included.
• An asterisk (*) is not allowed.

DQL Quick Reference
Applicable to Rule
AS clause The AS clause is acceptable with no restrictions.
FROM clause The following rules apply to the FROM clause:
• You may include only one object type reference, and that
reference must be to dm_sysobject or a SysObject subtype.
• You may not include references to a registered table.
SEARCH clause The SEARCH clause may not reference the FIRST or LAST
keyword. There are no restrictions on the full-text search
condition. However, note that the SEARCH DOCUMENT
CONTAINS syntax is the preferred way to specify a SEARCH
clause.
WHERE clause A WHERE clause in an FTDQL query is subject to the
following rules:
• Any repeating properties referenced in the clause must be
of type string or ID.
• Only the DQL UPPER and LOWER functions are allowed.
The functions SUBSTR, MFILE_URL, and all aggregate
functions are not acceptable.
• The following predicates are not allowed:
— BETWEEN
— NOT LIKE
— NOT FOLDER
— NOT NULL
— [NOT] CABINET
— ONLY
— TYPE
— <> ” (empty string)
• The IN and EXISTS keywords are not allowed.
• Any valid form of the FOLDER predicate (except NOT
FOLDER) may be used. The DESCEND option is also
allowed.

DQL Quick Reference
Applicable to Rule
• The following rules apply to the LIKE predicate:
— The LIKE predicate may be used with pattern matching
characters, but not against an ID attribute, only a string
attribute.
— The LIKE predicate can be used with an ESCAPE clause,
but it is ignored.
• The keywords TODAY, YESTERDAY, TOMORROW may
not be used in the DATE function.
• The following rules apply to all expressions in the WHERE
clause:
— Expressions may not contain the ISREPLICA or USER
keywords.
— Expressions may use any comparison operator.
— Expressions may use an arithmetic operator, but they
may not be used to form a compound expression.
— Expressions that compare one property to another are
not allowed. For example, subject=title is invalid.
— Expressions in the following format are not allowed:
property_name operator('literal' operator
'literal)
— Expressions that force index correspondence between
repeating properties are not allowed. Such expressions
AND together expressions that reference repeating
properties in the format:
predicate(repeating_attr_expr AND
repeating_attr_expr)
(For more information about forcing index
correspondence, refer to Forcing index correspondence
in query results, page 342.)
• The following additional rules apply to expressions in the
format first_expression operator second_expression
— The first_expression is limited to one of the following:
property name upper(property name)
lower(property name)
— The second_expression is limited to one of the following:

DQL Quick Reference
Applicable to Rule
literal value upper(literal value)
lower(literal value) the DATE() function
— The operator may be any valid operator.
ORDER BY clause If included, this clause must be:
ORDER BY SCORE
SCORE must be referenced by name, not position, in the
selected values list, also.
ENABLE clause There are no restrictions on this clause.
Unregister
UNREGISTER [TABLE] [owner_name.]table_name
Update
UPDATE table_name SET column_assignments
WHERE qualification
Update...Object
UPDATE [PUBLIC]type_name [(ALL)][correlation_var]
[WITHIN PARTITION partition_id {,partition_id}]
OBJECT[S] update_list
DQL reserved words
If you using DQL reserved words as object or property names, enclose name in double quotes when
using in a DQL query.

DQL Quick Reference
Table 120. DQL reserved words
Initial letter Reserved words
A
ABORT ALTER ASSEMBLIES
ACL AND ASSEMBLY
ADD ANY ASSISTANCE
ADD_FTINDEX APPEND ATTR
ADDRESS APPLICATION AUTO
ALL AS AVG
ALLOW ASC
B
BAG BOOL BUSINESS
BEGIN BOOLEAN BY
BETWEEN BROWSE
C
CABINET COMMENT CONTAINS
CACHING COMMIT CONTENT_FORMAT
CHANGE COMPLETE CONTENT_ID
CHARACTER COMPONENTS COUNT
CHARACTERS COMPOSITE CREATE
CHAR COMPUTED CURRENT
CHECK CONTAIN_ID
D
DATE DELETED DISTINCT
DATEADD DEPENDENCY DM_SESSION_DD_
LOCALE
DATEDIFF DEPTH DOCBASIC
DATEFLOOR DESC DOCUMENT
DATETOSTRING DESCEND DOUBLE
DAY DISABLE DROP
DEFAULT DISALLOW DROP_FTINDEX
DELETE DISPLAY

DQL Quick Reference
E
ELSE ENFORCE EXEC
ELSEIF ESCAPE EXECUTE
ENABLE ESTIMATE EXISTS
F
FALSE FOR FT_OPTIMIZER
FIRST FOREIGN FULLTEXT
FLOAT FROM FUNCTION
FOLDER FTINDEX
G
GRANT GROUP
H
HAVING HITS
I
ID INT IS
IF INTEGER ISCURRENT
IN INTERNAL ISPUBLIC
INSERT INTO ISREPLICA
J
JOIN
K
KEY
L
LANGUAGE LIGHTWEIGHT LITE
LAST LIKE LOWER
LATEST LINK
LEFT LIST
M
MAPPING MEMBERS MONTH
MATERIALIZE MFILE_URL MOVE
MATERIALIZATION MHITS MSCORE
MAX MIN
MCONTENTID MODIFY

DQL Quick Reference
N
NODE NOTE NULLID
NODESORT NOW NULLINT
NONE NULL NULLSTRING
NOT NULLDATE
O
OF ON ORDER
OBJECT ONLY OUTER
OBJECTS OR OWNER
P
PAGE_NO PERMIT PRIVATE
PARENT POLICY PRIVILEGES
PARTITION POSITION [1] PROPERTY
PATH PRIMARY PUBLIC
Note: [1] The POSITION keyword, previously supported in SELECT queries
against the full-text index, is no longer supported.
Q
QRY QUALIFIABLE
R
RDBMS RELATE REPORT
READ REMOVE REQUEST
REFERENCES REPEATING REVOKE
REGISTER REPLACEIF
S
SCORE SMALLINT SUPERTYPE
SEARCH SOME SUPERUSER
SELECT STATE SUPPORT
SEPARATOR STORAGE SYNONYM
SERVER STRING SYSADMIN
SET SUBSTR SYSOBJ_ID
SETFILE SUBSTRING SYSTEM
SHAREABLE SUM
SHARES SUMMARY

DQL Quick Reference
T
TABLE TO TRANSACTION
TAG TODAY TRUE
TEXT TOMORROW TRUNCATE
TIME TOPIC TYPE
TINYINT TRAN
U
UNION UNREGISTER UPDATE
UNIQUE USER UPPER
UNLINK USING
V
VALUE VERSION VIOLATION
VALUES VERITY
W
WHERE WITHOUT WEEK
WITH WORLD
WITHIN WRITE
Y
YEAR YESTERDAY

Appendix F
Document Query Language Examples
This appendix contains examples of Document Query Language (DQL) SELECT statements. The
examples begin with basic SELECT statements that retrieve information about objects using the
standard SELECT clauses, such as the WHERE clause and GROUP BY clause. Then the examples
show SELECT statements that make use of the DQL extensions to the statement, such as the ability
to use folder and virtual document containment information or registered tables to retrieve object
information.
This appendix does not attempt to describe the syntax of the DQL SELECT statement in detail. For a
complete description of SELECT, refer to Select, page 116.
Basic examples
This section presents basic SELECT statements. These examples demonstrate the use of standard
clauses such as the WHERE, GROUP BY, and HAVING clauses, as well as how to use aggregate
functions in a query. Some of these examples also demonstrate how to query repeating properties.
The simplest format
The basic SELECT statement has the format:
SELECT target_list FROM type_name
The target_list identifies the value or values that you want to retrieve and type_name identifies the
types or registered tables from which you want to retrieve the requested information.
The following example returns the user names of all the users in the repository:
This next example returns the names of the users and their user privileges within a repository:
SELECT "user_name","user_privileges" FROM "dm_user"

Using the WHERE clause
The WHERE clause restricts the information retrieved by placing a qualification on the query. Only
the information that meets the criteria specified in the qualification is returned. The qualification
can be simple or complex. It can contain any of the valid predicates or logical operators, such
as AND, OR, or NOT.
The following example returns the user names of all users that belong to a specified group:
WHERE "user_group_name"='engr'
The next example retrieves all the types that do not have SysObject as their direct supertype and
have more than 15 properties:
SELECT "name" FROM "dm_type"
WHERE "super_name" != 'dm_sysobject'
AND "attr_count" > 15
The following example returns the types that were created one day before the current date together
with the date and time that the types were last modified:
SELECT "name", "r_modify_date" FROM "dm_type"
where "r_creation_date" >= DATE(YESTERDAY)
Searching repeating properties in a WHERE Clause
To search a repeating property for a specific value, you use the ANY predicate in the WHERE clause
qualification unless you have included the ROW_BASED hint in the query. When you use this
predicate, or if the query includes ROW_BASED, if any of the values in the specified repeating
property meets the search criteria, the server returns the object.
The following example searches the authors property to return any document that has Jim as one of
its authors:
WHERE ANY "authors" = 'jim'
This next example searches the authors property and returns any document that has either Jim or
Kendall as an author:
WHERE ANY "authors" IN ('jim', 'kendall')
The following example searches the keywords property and returns any document that has a key
word that begins with hap:
WHERE ANY "keywords" LIKE 'hap%'
Using aggregate functions
DQL recognizes several aggregate functions. Aggregate functions are functions that operate on a
set and return one value. For example, the count function is an aggregate function. It counts some

number of objects and returns the total. (For a full description of all the aggregate functions that you
can use in DQL queries, refer to Aggregate functions, page 22.)
This example counts the number of documents owned by the user named john:
WHERE "owner_name"='john'
The following example returns the dates of the oldest and newest documents in the repository:
SELECT MIN(DISTINCT "r_creation_date"),MAX(DISTINCT "r_creation_date")
FROM "dm_document"
This final example returns the average number of properties in Documentum types:
SELECT AVG(DISTINCT "attr_count") FROM "dm_type"
Using the GROUP BY clause
The GROUP BY clause provides a way to sort the returned objects on some property and return an
aggregate value for each sorted subgroup. The basic format of the SELECT statement when you
want to use the GROUP BY clause is:
SELECT property_name,aggregate_function FROM type_name
GROUP BY property_name
The property named in the target list must be the same as the property named in the GROUP BY
clause.
The following example retrieves the names of all document owners and for each, provides a count of
the number of documents that person owns:
SELECT "owner_name",count(*) FROM "dm_document"
Using the HAVING clause
The HAVING clause is used in conjunction with the GROUP BY clause. It restricts which groups are
returned.
The following example returns the owner names and a count of their documents for those owners
who have more than 10 documents:
HAVING COUNT(*) > 10
The ORDER BY clause
The ORDER BY clause lets you sort the values returned by the query. The clause has the format:
ORDER BY num [ASC|DESC] {,num [ASC|DESC] }
or
ORDER BY property [ASC|DESC] {,property [ASC|DESC]}

The num identifies one of the selected values by its position in the selected values list. The property
identifies one of the selected values by its name. ASC (ascending) and DESC (descending) define the
order of the sort and must be specified for each element specified in the ORDER BY clause.
The following example returns the owner’s name, the object’s title, and its subject for all SysObjects.
The results are sorted by the owner’s name and, within each owner’s name group, by title:
SELECT "owner_name","title","subject" FROM "dm_sysobject"
ORDER BY "owner_name", "title"
This next example returns the owner names and a count of their documents for those owners who
have more than 10 documents. The results are ordered by the count, in descending order.
GROUP BY "owner_name" HAVING COUNT(*) < 10
ORDER BY 2 DESC
Using the asterisk (*) in queries
DQL lets you use the asterisk in the target list. If the query does not include the ROW_BASED hint
and the asterisk is the selected value, the FROM clause may specify only a type name or a registered
table—it cannot specify both. If the query includes the ROW_BASED hint, the FROM clause may
include either or both object types and registered tables. The values returned by the asterisk depend
on what is specified in the FROM clause and whether the ROW_BASED hint is included. For
examples and complete details, refer to The asterisk (*) as a selected value, page 131.
Searching cabinets and folders
In Documentum, objects of all SysObject subtypes except cabinets are stored in cabinets, and
sometimes, in folders within those cabinets. It may, at times, be advantageous to restrict the search
for an object or objects to a particular cabinet or folder. Or, you may want to determine what a
particular cabinet or folder contains. To make these operations possible, DQL provides the CABINET
and FOLDER predicates for use with the WHERE clause. The CABINET predicate lets you specify a
particular cabinet to search. The FOLDER predicate lets you specify a particular folder or cabinet
to search. (You can use the FOLDER predicate for cabinets as well as folders because cabinets are
subtypes of folders.)
Use the cabinet or folder’s folder path or its object ID to identify it in the query. A folder path has
the format:
/cabinet_name{/folder_name}
To use the object ID, you must use the ID function and the object ID. The following examples illustrate
the use of both folder paths and object IDs with the CABINET and FOLDER predicates.
This example returns all the folders contained in Research folder, which is identified by its object ID:
SELECT "object_name" FROM "dm_folder"
WHERE FOLDER (ID('0c00048400001599'))
The next example returns all the objects in the Marketing cabinet sorted by their type and, within
each type, ordered alphabetically. The Marketing cabinet is identified by its folder path.
SELECT "object_name","r_object_type" FROM "dm_sysobject"

WHERE CABINET ('/Marketing')
ORDER BY "r_object_type","object_name"
This final example returns all the objects in the Correspondence folder in the Marketing cabinet.
Again, the objects are sorted by their type and ordered alphabetically within each type group. A
folder path is used to identify the correct folder.
SELECT "object_name","r_object_type" FROM "dm_sysobject"
WHERE FOLDER ('/Marketing/Correspondence')
ORDER BY "r_object_type","object_name"
The CABINET and FOLDER predicates have an optional keyword, DESCEND, that directs the server
to return not only those objects directly contained within a cabinet or folder but to also return the
contents of any folders contained in that folder or cabinet, and so forth. (There is a limit of 25,000
contained folders within the specified folder when you include DESCEND in the predicate.)
The following example returns the name, object ID, and type of all objects in the Clippings folder,
including the contents of any folders that are contained by the Clippings folder. The Clippings
folder resides in the Marketing cabinet.
SELECT "object_name","r_object_id","r_object_type"
FROM "dm_sysobject"
WHERE FOLDER ('/Marketing/Clippings', DESCEND)
Querying registered tables
A registered table is a table from your underlying RDBMS that has been registered with the repository.
This registration allows you to specify the table in a DQL query, either by itself or in conjunction
with a type. The following examples illustrate each possibility. (For a full discussion of the rules
concerning the use of registered tables in SELECT statements, refer to Chapter 2, DQL Statements.)
The first example returns all the name of all products and their owners from the registered table
named ownership:
SELECT "product","manager" FROM "ownership"
This second example joins the registered table called ownership to the user-defined type called
product_info to return the names of all the objects in Collateral folder, along with the names of the
managers who own the products described by the objects:
SELECT p."object_name", o."manager"
FROM "product_info" p, "ownership" o
WHERE p."product" = o."product" AND
FOLDER ('/Marketing/Collateral', DESCEND)
Querying virtual documents
DQL provides a clause that facilitates querying virtual documents. The clause is the IN DOCUMENT
clause. The IN DOCUMENT clause lets you search a particular virtual document.
The examples in this section are based on the virtual document shown in Figure 4, page 406:

Figure 4. Virtual document model
Determining the components
The following example returns only the directly contained components of Book1:
IN DOCUMENT ID('Book1_object_id')
This query returns Book1, Chapter 1, and Chapter 2. Book1 is included because a virtual document is
always considered to be a component of itself.
To return all the components of a virtual document, including those that are contained in its
components, you use the keyword DESCEND. For example:
IN DOCUMENT ID('Book1_object_id') DESCEND
The above example returns Book1, Chapter 1, Doc 1, Doc2, Chapter 2, and Doc 3, in that order.
Note: The components of a virtual document are returned in a pre-determined order that you cannot
override with an ORDER BY clause.
You can limit the search to a particular object type by specifying the type in the FROM clause. For
example, the following statement returns only the documents that make up Book1 (Doc 1, Doc 2,
and Doc 3):
IN DOCUMENT ID('Book1_object_id')

Index
* (asterisk)
in queries, 404
wildcard in search queries, 142
% (percent sign)
pattern matching character, 36
_ (underbar)
[] (square brackets) for index values, 162
(*) asterisk
as selected value, 131
ROW_BASED hint, effect of, 132
A
ABORT (statement), 44
accent marks, in search strings, 142
access controls for registered tables, 345
ACLs
object types, assigning to, 81
setting default, 60
add
partition, 277
administration methods
BATCH_PROMOTE, 171
CAN_FETCH, 173
CHECK_CACHE_CONFIG, 174
CHECK_RETENTION_EXPIRED, 177
CHECK_SECURITY, 181
CLEAN_DELETED_OBJECTS, 183
CLEAN_LINKS, 184
DB_STATS, 186
DELETE_REPLICA, 188
DESTROY_CONTENT, 190
DO_METHOD, 192
DROP_INDEX, 199
ESTIMATE_SEARCH, 201
EXEC_SQL, 203
executing
Documentum Administrator, 165
EXECUTE statement, 165
EXPORT_TICKET_KEY, 205
FINISH_INDEX_MOVES, 207
GET_INBOX, 216
GET_LAST_SQL, 219
GET_PATH, 221
GET_SESSION_DD_LOCALE, 223
HTTP_POST, 224
IMPORT_REPLICA, 228
IMPORT_TICKET_KEY, 230
LIST_AUTH_PLUGINS, 232
LIST_RESOURCES, 233
LIST_SESSIONS, 237
LIST_TARGETS, 240
LOG_OFF, 242
LOG_ON, 243
MAKE_INDEX, 245
MARK_AS_ARCHIVED, 248
MARK_FOR_RETRY, 249
MIGRATE_CONTENT, 251
MODIFY_TRACE, 273
MOVE_INDEX, 275
PING, 286
PURGE_AUDIT, 287
PURGE_CONTENT, 293
PUSH_CONTENT_ATTRS, 294
RECOVER_AUTO_TASKS, 296
REGISTER_ASSET, 298
REINDEX_PARTITIONABLE_
TYPE, 300
REORGANIZE_TABLE, 301
REPLICATE, 303
RESET_TICKET_KEY, 305
RESTORE_CONTENT, 306
ROLES_FOR_USER, 308
scope, 166
SET_APIDEADLOCK, 310
SET_CONTENT_ATTRS, 313
SET_OPTIONS, 317
SET_STORAGE_STATE, 320
SHOW_SESSIONS, 322
TRANSCODE_CONTENT, 324
UPDATE_STATISTICS, 327

Index
WEBCACHE_PUBLISH, 329
administrative methods
FIX_LINK_CNT, 208
GENERATE_PARTITION_SCHEME_
SQL, 209
GET_FILE_ULR, 214
MIGRATE_TO_LITE, 265
PARTITION_OPERATION, 277
aggregate functions
as selected values, 125
AVG, 24
COUNT, 23
examples of use, 402
MAX, 23
MIN, 23
SUM, 24
aggregate functions, default values
and, 339
ALL (keyword), 95, 122
ALL (predicate), 34
ALTER ASPECT (statement), 47
ALTER GROUP (statement), 45
ALTER TYPE (statement), 51
AND (logical operator), 40
ANSI format for dates, 16
ANY keyword in WHERE clauses, 143
ANY...comparison operator (predicate), 35
ANY...IN subquery (predicate), 34
ANY...IS NULL (predicate), 35
ANY...IS NULLDATE (predicate), 35
ANY...IS NULLID (predicate), 35
ANY...IS NULLINT (predicate), 35
ANY...IS NULLSTRING (predicate), 35
ANY...LIKE (predicate), 34
API (Application Program Interface)
IAPI utility, 371
application server
sending request to, 224
Apply method
operations, 166
archived documents, restoring, 306
archiving
audit trail entries, 248
archiving content files, 293
arithmetic expressions as selected
values, 127
arithmetic operators, 32
ascending sort order, 148
ASCII (scalar function), 20
aspect properties
modifying, 47
aspects
adding defaults to object types, 57
adding/dropping properties, 47
removing defaults from object
types, 58
assemblies
IN ASSEMBLY clause, 139
SELECT statements and, 138
asterisk (*)
as selected value, 131
in queries, 404
ROW_BASED hint, effect of, 132
wildcard in search queries, 142
audit trail entries
i_is_archived, setting, 248
removing from repository, 287
auditing
audit trail entries, removing, 287
authentication
LIST_AUTH_PLUGINS administration
method, 232
LIST_SESSIONS description, 238
tracing, 318 to 319
authentication plugins
list of, obtaining, 232
automatic activities
RECOVER_AUTO_TASKS
administration method, 296
AVG (aggregate function), 24
B
BATCH_PROMOTE administration
method, 171
BEGIN TRAN (statement), 67
BITAND (scalar function), 20
BITCLR (scalar function), 20
BITSET (scalar function), 20
bitwise, 20
brackets for index values, 162
C
ca store
moving content to/from, 257
CABINET (predicate), 39
cabinets
CABINET (predicate), 39
querying (examples of), 404

Index
cache config objects
consistency checking, 174
cached queries, 346
CAN_FETCH
case sensitivity
scientific notation, 14
case sensitivity, in search strings, 141
CHANGE OBJECT (statement), 68
CHANGE...OBJECT (statement), 334
character strings
converting dates to, 26
dates within, 15
dfc.compatibility.truncate_long_values
preference, 14
lengths, 14
literal formats, 14
lowercase and LOWER function, 21
pattern arguments, 16
returning substrings, 22
short date formats, 16
uppercase and UPPER function, 21
character strings, ASCII function, 20
check constraints, 86, 88
CHECK_ACL function, 28
CHECK_CACHE_CONFIG
(administration method), 174
CHECK_SECURITY (administration
method), 181
CLEAN_LINKS
client applications
checking for server connection, 286
short date format defaults, 16
client code pages
tracing, 318
client machines
tracing information for, 318
client roles
determining, 308
code page requirements
SYNONYM FOR clause, 113
columns
defining, 111
inserting values into, 107
updating, 155
comments
IAPI format, 375
IDQL format, 371
comments, in DQL, 43
COMMIT (statement), 71
common area
removing files from, 318
component specifications
dropping, 62
removing, 62
specifying, 62
component_specification, 89
connection brokers
obtaining list of targets, 240
tracing information, 318
connections
checking with PING, 286
constraints
check (data validation), 86, 88
creating with ALTER TYPE, 64
creating with CREATE TYPE, 86
dropping with property_drop_
clause, 65
dropping with type drop clause, 62
removing, 62
specifying, 61
CONTAIN_ID (keyword), 128
content addressable storage
PUSH_CONTENT_ATTRS
content files, 303
archiving, 293
deleting replicas, 188
fetching, 173
importing replicas, 228
MIGRATE_CONTENT administration
method, 251
obtaining file path, 221
removing, 190
removing from content-addressed
storage, 190
restoring archived, 306
transformations, requesting, 324
content objects
destroying, 190
SET_CONTENT_ATTRS
Content Server
common area, removing files from, 318

Index
Docbasic expressions, handling of, 377
history_cutoff startup parameter, 322
history_sessions startup
parameter, 322
listing active sessions, 322
listing connection broker targets, 240
obtaining resource information, 233
obtaining session information, 237
objects with expired retention,
finding, 177
content-addressed storage systems
content, removing, 190
tracing operations in, 318
CONTENTID (keyword), 127
contents
adding with SETFILE, 161
CONVERT_FOLDER_LIST_TO_OR, 350
correlation names, 136
COUNT (aggregate function), 23
create
partition, 277
CREATE...GROUP (statement), 72
CREATE...OBJECT (statement), 74
CREATE...TYPE (statement), 77
cursors, RDBMS, 187
D
data dictionary
attribute information, defining, 83
component specification, 89
current locale, obtaining, 223
default values for properties, 335
defining
default property values, 85
mapping table information, 85
value assistance information, 84
defining type information, 87
Java evaluation of Docbasic
expressions, 377
PUBLISH (DQL keyword), 56, 82
update_modifier specification, 83, 87
data partitioning, 58, 82
data validation, check constraints, 86
database
footprint reduction, dmr_content, 365
database hint
DM_LEFT_OUTER_JOIN_FOR_
ACL, 362
database hints
CONVERT_FOLDER_LIST_TO_
OR, 350
FETCH_ALL_RESULTS, 350
FORCE_ORDER, 351
FT_CONTAIN_FRAGMENT, 352
FTDQL and NOFTDQL, 351
GERATE_SQL_ONLY hint, 352
GROUP_LIST_LIMIT DQL hint, 352
HIDE_SHARED_PARENT DQL
hint, 353
IN and EXISTS, 353
including multiple hints, 362
OPTIMIZE_ON_BASE_TABLE, 354
OPTIMIZE_TOP, 355
passthrough hints, 363
RETURN_RANGE, 356
RETURN_TOP, 357
ROW_BASED, 359
SQL_DEF_RESULT_SET, 361
TRY_FTDQL_FIRST, 362
UNCOMMITTED_READ, 362
using, 350
datatypes
lengthening character string types, 57
date functions
DATEADD, 26
DATEDIFF, 25
DATEFLOOR, 26
DATEFLOOR_LOCAL, 27
DATETOSTRING, 26
DATETOSTRING_LOCAL, 28
date literals, 15
DATEADD (date function), 26
DATEDIFF (date function), 25
DATEFLOOR (date function), 26
DATEFLOOR_LOCAL (date function), 27
dates
ANSI format, 16
date literals, 15
earliest acceptable year, 15
functions, 24
input formats, 15
localized formats, 16
NOW (keyword), 17
output formats, 18
storage in repository, 18

Index
TODAY (keyword), 17
TOMORROW (keyword), 17
truncating, 26
using in queries, 15
YESTERDAY (keyword), 17
DATETOSTRING (date function), 26, 28
DB_STATS administration method, 186
dbo alias, 345
deadlocks, simulating, 310
debug tracing, 318
default aspects
adding to object types, 57
removing from object types, 58
default values
aggregate functions and, 339
default property specification, 64
groups for a type, 61
private and public groups, 73
returned format, 336
storage area for types, 60
testing for, 338
delete
partition, 277
DELETE (statement), 91, 335
DELETE...OBJECT (statement), 93, 335
DELETE_REPLICA administration
method, 188
DEPTH (keyword), 128
DESCEND (keyword), 138
descending sort order, 148
DESTROY_CONTENT administration
method, 190
DFC (Documentum Foundation Classes)
Docbasic expression handling, 378
preference, 14
diacritical marks, in search strings, 142
digital shredding
tracing, 318
disabling
assume_user program, 197
DISTINCT (keyword), 122
distributed storage
deleting replicas, 188
fetching content files, 173
importing replicas, 228
replicating files, 303
dm_bp_batch method, 171
dm_CheckCacheConfig job, 175
dm_dbo alias, 345
DM_LEFT_OUTER_JOIN_FOR_ACL, 362
dm_registered object type, 345
DM_SESSION_DD_LOCALE
(keyword), 19
dmc_MigrateDbExprsToJava method, 379
dmc_MigrateDbExprsToJavaForType
method, 379
dmc_SetJavaExprEnabled method, 381
DMCL (client library)
version number returned by
LIST_SESSIONS, 238
dmi_audittrail_attrs
SELECT constraint, 135
dmr_content
database footprint reduction, 365
DO_METHOD administration
method, 192
DO_METHOD procedures
running as server, 197
DO_METHOD result object properties, 194
Docbasic, 377
See also Docbasic expression migration
constants
supported for Java evaluation, 387
unsupported for Java
evaluation, 388
executing scripts, 192
expression handling by Content
Server, 377
expression handling, by DFC Version
6, 378
expression handling, pre-6 DFC, 378
expressions, migrating to Java, 377
functions
supported for Java evaluation, 382
unsupported for Java
evaluation, 386
operators, supported for Java
evaluation, 381
value_assistance_modifier, 84
Docbasic expression migration
dmc_MigrateDbExprsToJava
method, 379
dmc_MigrateDbExprsToJavaForType
method, 379
implicit objects, 388
results document, 380
supported expression components, 381
Docbasic expressions

Index
Java evaluation, disabling, 380
Docbasic expressions migration
return values, 380
domain, 78
double quotes, using, 335
DQL (Document Query Language)
basic query statements, 333
CHECK_ACL function, 28
comments, in-line, 43
comparison operators, 32
date functions, 24
examples, 401
functions, 19
GET_LAST_SQL administration
method, 219
ID function, 30
keywords, 18
literals, 13
logical operators, 40
MFILE_URL function, 31
predicates, 32
queries, caching, 346
registered tables, referencing, 345
repeating properties, referencing, 339
reserved words, 396
tracing generated SQL, 318
user privileges and, 346
DQL hints
ROW_BASED, 124
using, 148
DROP GROUP (statement), 97
DROP TYPE (statement), 98
DROP_INDEX administration method, 199
dump files
date format in, 18
E
ENV_CONNECT_DOCBASE (iapi/idql
argument), 368, 373
ENV_CONNECT_PASSWORD (iapi/idql
argument), 368, 372
ENV_CONNECT_USER_NAME (iapi/idql
argument), 368
error messages
date formats in, 18
escape character, 37
ESTIMATE_SEARCH (administration
method), 201
EXEC_SQL administration method, 203
EXECUTE statement
described, 99
EXISTS (database hint), 353
EXISTS (predicate), 34
EXPORT_TICKET_KEY administration
method, 205
expressions
Java evaluation, disabling, 380
external applications
deadlocks, simulating, 310
executing, 192
external procedures
method objects and, 195
external store
moving content from, 257
F
FALSE (keyword), 19
FETCH_ALL_RESULTS, 350
files
removing from server common
area, 318
FINISH_INDEX_MOVES administration
method, 207
FIRST keyword, 140
FIX_LINK_CNT administrative
method, 208
floating point literals, 14
FOLDER (predicate), 39
nested folder limit, 39
folders
FOLDER (predicate), 39
FORCE_ORDER, 351
formats
date literals, 15
date output, 18
folder path, 39
integer literals, 13
setting with DATETOSTRING
function, 26
value assistance, 84
FROM clause
disambiguating types and tables, 136
inline view specification, 136
table name specification, 135
type name specification, 134

Index
FT_CONTAIN_FRAGMENT DQL
hint, 352
FTDQL DQL hint, 351
FTDQL queries. See SELECT statement
distinct_query_results server.ini key
and, 122
full-text indexes
MARK_FOR_RETRY administration
method, 249
querying, 201
SEARCH clause, 140
virtual documents and, 344
full-text indexing
tracing operations, 273
full-text keywords, 127
functions
aggregate, 339
Apply method, 166
ASCII, 20
bitwise, 20
CHECK_ACL, 28
date, 24
DATEADD, 26
DATEDIFF, 25
DATEFLOOR, 26
DATEFLOOR_LOCAL, 27
DATETOSTRING, 26
DATETOSTRING_LOCAL, 28
defined, 19
EXECUTE statement, 100
GROUP BY clause and, 146
ID, 30
LOWER, 21
MFILE_URL, 31, 126
SUBSTR, 22
UPPER, 21
G
GENERATE_PARTITION_SCHEME_SQL
administrative method, 209
GET_FILE_URL administrative
method, 214
GET_INBOX administration method, 216
GET_LAST_SQL administration
method, 219
GET_PATH administration method, 221
GRANT (statement), 105
GROUP BY clause
described, 146
groups
alter group (statement), 45
method), 181
creating, 72
default for type, 61
private, 46
public, 46
removing, 97
H
HAVING clause
defined, 146
hierarchy, searching, 138
history_cutoff (startup parameter), 322
history_sessions (startup parameter), 322
HTTP_POST administration method, 224
I
i_is_archived property, 248
IAPI (utility)
commands for, 374
comments, entering, 375
executing methods, 375
exiting, 376
ID function, 30
ID literal format, 15
IDQL (utility)
buffer, clearing, 371
commands, 369
idql command syntax, 367
queries, entering, 370
quiting, 371
starting, 367
IMPORT_REPLICA administration
method, 228
IMPORT_TICKET_KEY administration
method, 230
IN (database hint), 353
IN (predicate), 33
IN and EXISTS, 353
IN ASSEMBLY clause
DELETE...OBJECT and, 95
SELECT and, 139

Index
unioned selects and, 147
IN DOCUMENT clause
described, 137
NODESORT BY option, 138
unioned selects and, 147
WITH option, 138
inboxes
GET_INBOX administration
method, 216
index positions
defined, 340
forcing correspondence, 342
indexed properties
querying, 140
indexes. See object type indexes
input formats for dates, 15
INSERT (statement), 107, 335
integer literal format, 13
internationalization
client code pages, tracing, 318
object types
altering and, 56
creating and, 82
session locales, tracing, 318
IS NULL (predicate), 33
IS NULLDATE (predicate), 33
IS NULLID (predicate), 33
IS NULLINT (predicate), 33
IS NULLSTRING (predicate), 33
ISCURRENT (keyword), 127
ISPUBLIC (keyword), 127
ISREPLICA (keyword), 129
J
Java
evaluation of Docbasic expressions,
disabling, 380
java servlets, invoking, 224
jobs
dm_CheckCacheConfig, 175
JOIN
implied, 134
K
keywords
DQL, 18
full-text, 127
miscellaneous, 129
PUBLISH, 56, 82
THUMBNAIL_URL, 130
USER, 18, 130
virtual document, 128
L
LAST keyword, 140
LAUNCH_DIRECT, run_as_server
and, 197
LEFT JOIN, 134
LEFT OUTER JOIN, 134
lifecycles
batch promotion, 171
defining default for object types, 61
lightweight type
create, 82
LIKE (predicate)
described, 33
pattern matching with, 36
link record object type, 184
linked storage areas
removing unneeded objects from, 184
linking objects with UPDATE...
OBJECT, 163
links
removing with UPDATE...
OBJECT, 164
Linuxprocess launch, tracing, 319
LIST_AUTH_PLUGINS administration
method, 232
LIST_RESOURCES administration
method, 233
LIST_SESSIONS administration
method, 237
LIST_TARGETS administration
method, 240
literals
character string, 14
date, 15
date keywords, 17
escape character and pattern
matching, 37
floating point, 14
ID format, 15
integer, 13
pattern matching, 36
types of, 13
locales

Index
determining current, 223
DM_SESSION_DD_LOCALE
keyword, 19
locking
tracing Windows, 318
log files
LOG_OFF administration method, 242
LOG_ON administration method, 243
method, 263
LOG_OFF administration method, 242 to
243
LOG_ON administration method, 243
logical operators
described, 40
precedence order, 41
login ticket key
EXPORT_TICKET_KEY administration
method, 205
method, 230
RESET_TICKET_KEY administration
method, 305
login ticket keys
tracing import/export, 319
login tickets
tracing, 319
logs
tracing and, 273
LOWER (scalar function), 21
M
MAKE_INDEX administration
method, 245
mapping information, dropping, 65
mapping table specification, 63, 85
MARK_AS_ARCHIVED administration
method, 248
MARK_FOR_RETRY administration
method, 249
MAX (aggregate function), 23
max_nqa_string (server.ini key), 55, 81
maximums
characters in string literals, 14
floating point literal values, 14
number of properties in ALTER TYPE
statements, 56
values returned by MAX aggregate
function, 23
method objects
run_as_server property, 197
running procedures as server, 197
method server
tracing execution, 319
methods
dm_bp_batch, 171
dmc_SetJavaExprEnabled, 381
executing with DO_METHOD, 165
IAPI execution, 375
tracing method server, 319
MFILE_URL (DQL function), 31, 126
method, 251
log file, 263
MIGRATE_TO_LITE administrative
method, 265
MIN (aggregate function), 23
minimum values
returned by MIN aggregate
functions, 23
MODIFY_TRACE administration
method, 273
MOVE_INDEX administration
method, 275
N
NODESORT BY, 138
NOFTDQL DQL hint, 351
non-qualifiable properties
constraint on string datatypes, 55, 81
NOT (logical operator), 40
NOT EXISTS (predicate), 34
NOT IN (predicate), 33
NOT LIKE (predicate)
described, 33
pattern matching with, 36
NOW (keyword), 17
NULL values
in properties, 335
sorting, 339
testing for, 338
NULLDATE (keyword), 18
O
object IDs
IN DOCUMENT clause and, 137

Index
object type indexes
creating, 245
DROP_INDEX administration
method, 199
FINISH_INDEX_MOVES
moving, 275
object types
adding default aspects, 57
changing to another, 334
changing with ALTER TYPE, 51
component specifications,
removing, 62
component_specification, 89
constraints, removing, 62
creating, 77
data dictionary information,
defining, 87
defining default lifecycle for, 61
deleting, 98
mapping table specification, 61, 87 to
88
names, quoting in applications, 43
properties
maximum allowed number of, 80
properties, adding, 56
referencing, 335
referencing in queries, 134
removing default aspects, 58
removing indexes for, 199
object-level permissions
DQL statements and, 346
objects
batch promotion, 171
changing to another type, 334
changing type, 68
creating, 74
deleting, 93
linking with UPDATE...OBJECT, 163
moving to new folder/cabinet, 164
unlinking with UPDATE...
OBJECT, 163
Update...Object (statement), 158
OBJTYPE (keyword), 127
operators
arithmetic, 32
comparison, 32
logical, 40
OPTIMIZE_ON_BASE_TABLE, 354
OPTIMIZE_TOP, 355
OR (logical operator), 40
Oracle
RETURN_TOP, 359
SYNONYM FOR code page
requirement, 113
ORDER BY clause
SELECT statement, 148
OUTER JOIN, 134
P
paginate
using RETURN_RANGE, 356
PARENT (keyword), 129
parentheses to change precedence of logical
operators, 41
partition, 58, 82
add, 277
detach, 277
merge, 277
split, 277
switch, 277
partition scheme, 277
PARTITION_OPERATION administrative
method, 277
partitioning types, 58, 82
passthrough database hints, 363
pattern matching
characters, 36
escape character, 37
percent sign (%)
performance
SEARCH clauses, 202
searches and PUBLIC keyword, 133
WHERE clause, 144
permits, registered tables, 345
persistent client caching
CHECK_CACHE_CONFIG
dm_CheckCacheConfig job, 175
PING administration method, 286
precedence
date formats, 16
logical operators, 41
predicates
ANY...comparison operator, 35
ANY...IN, 34
ANY...IS NULL, 35

Index
ANY...IS NULLDATE, 35
ANY...IS NULLID, 35
ANY...IS NULLINT, 35
ANY...IS NULLSTRING, 35
ANY...LIKE, 34
CABINET, 39
comparison operators, 32
defined, 32
FOLDER, 39
pattern matching, 36
repeating property predicates, 34
single-valued property predicates, 33
SysObject, 38
TYPE, 38
where clause components, 142
WHERE clause components, 32
printing, assume user feature and, 197
private
groups, 46, 73
privileges
granting, 105
revoking, 114
programs
assume user, 197
properties, 339
See also repeating atttributes
adding to type definition, 56
aspect properties, selecting, 123
assume_user_location and disabling
assume user program, 197
data dictionary information,
defining, 83
default values, defining, 85
preference, 14
dropping, 57
lengthening character string types, 57
mapping table information,
defining, 85
modifiers, 63
modifying, 57
names, quoting in applications, 43
non-qualifiable, selecting, 123
NULL values and, 335
querying indexed, 140
referencing, 335
string length constraint, 55, 81
property drop clause, 65
property values, retrieving, 122
public
groups, 46, 73
objects, selecting, 127
PUBLIC keyword
creating public groups, 72
DELETE...OBJECT statement, 95
SELECT statements and, 133
PUBLISH keyword, 56, 82
publishing, to Web site, 329
PURGE_AUDIT administration
method, 287
PURGE_CONTENT administration
method, 293
PUSH_CONTENT_ATTRS administration
method, 294
Q
qualifying table names, 136
queries
database hints, 349, 364
executing SQL, 203
FROM clause limitations, 134
query performance
GROUP_LIST_LIMIT DQL hint, 352
HIDE_SHARED_PARENT DQL
hint, 353
query result objects
ROW_BASED DQL hint effects, 359
query results
returning as object-based, 124
ROW_BASED DQL hint, 124
queue items
marking for indexing retry, 249
R
RDBMS
deleting table rows, 91
inserting table rows, 107
operation statistics, 186
passthrough hints for SELECT, 148
permissions for Superusers, 346
registered tables
described, 345
permissions for, 345
referencing in queries, 345
security, 345
REORGANIZE_TABLE administration
method, 301

Index
unregistering tables, 153
UPDATE_STATISTICS administration
method, 327
updating rows, 155
RECOVER_AUTO_TASKS administration
method, 296
REGISTER (statement), 110
REGISTER_ASSET administration
method, 298
registered tables
creating, 110
deleting rows, 91
described, 345
DQL predicates for, 33
inserting rows, 107
permissions, default, 346
querying, 125
referencing in queries, 134, 345
security, 345
SYNONYM clause, 112
table names, qualifying, 136
unregistering, 153
updating rows, 155
removing
repeating property values, 163
REORGANIZE_TABLE administration
method, 301
repeating properties
adding values, 340 to 341
comparison expressions, use in, 144
compound ANY predicates, use in, 144
deleting values, 342
forcing index correspondence, 342
index positions, 340
NULL values and, 335
predicates for, 34
removing multiple values, 163
truncating, 163
updating values, 342
values
appending, 162
inserting, 163
removing, 163
WHERE clause use, 143
replicas
deleting, 188
determination of, 129
importing, 228
REPLICATE administration method, 303
replicating, 303
repositories
audit trail entries, removing, 287
method, 230
owner name aliases, 345
repository sessions
query caching, 346
reserved words
table of, 396
RESET_TICKET_KEY administration
method, 305
RESTORE_CONTENT administration
method, 306
retention policies
tracing use, 318
retention_trace tracing option, 318
RETURN_RANGE
described, 356
RETURN_TOP
described, 357
SEARCH clause and, 359
REVOKE (statement), 114
roles. See client roles
ROLES_FOR_USER administration
method, 308
ROW_BASED DQL hint, 359
affect on asterisk as selected value, 132
RPC calls, tracing, 318
RPC logging
turning off, 242
turning on, 243
RPC tracing, 318
run_as_server property, 197
S
scalar functions
ASCII, 20
BITAND, BITCLR, BITSET, 20
LOWER, 21
SUBSTR, 22
UPPER, 21
scheme
partition, 277
SCORE (keyword), 127

Index
SEARCH clause
accents and diacritical marks in, 142
asterisk as wildcard, 142
case sensitivity, 141
DOCUMENT CONTAINS format, 140
RETURN_TOP and, 359
SELECT statement, 140
words and phrases, specifying, 140
search string
words and phrases, specifying, 140
search strings
accent marks in, 142
asterisk as wildcard, 142
case sensitivity, 141
diacritical marks in, 142
security
method), 181
DQL statements and, 346
SELECT (statement)
ALL keyword, 122, 134
arithmetic expressions as selected
values, 127
AS clause, 132
assigning result property names, 132
asterisk as selected value, 131
DELETED keyword, 134
described, 334
disambiguating types and tables, 136
DISTINCT (keyword), 122
DQL functions as selected values, 125
DQL hints, 148
FROM clause, 133
FROM clause limitations, 134
GROUP BY clause, 146
HAVING clause, 146
IN ASSEMBLY clause, 139
IN DOCUMENT clause, 137
inline view specification, 136
keywords as selected values, 127
LITE keyword, 134
MFILE_URL as selected value, 126
ORDER BY clause, 148
permissions and, 346
retrieving column values, 125
retrieving property values, 122
SEARCH clause, 140, 201
table name specification, 135
type name specification, 134
UNION clause, 147
WHERE clause, 142
where clause qualification
components, 142
WITH option, 138
SELECT statement
aspect properties, referencing, 122
aspect properties, selecting, 123
described, 116
dmi_audittrail_attrs constraint, 135
FTDQL queries, 121
non-qualifiable properties,
selecting, 123
server.ini file, session information
parameters, 322
server.ini parameter
spaceoptimize, 365
servers. See Content Server, 346
session locales
tracing, 318
sessions
listing active, 322
listing current and historical, 237
SET_APIDEADLOCK administration
method, 310
SET_CONTENT_ATTRS administration
method, 313
SET_OPTIONS administration
method, 317
SET_STORAGE_STATE administration
method, 320
SETFILE clause, 161
shareable type
alter type, 58
create, 82
short date formats
defaults, 18
described, 16
SHOW_SESSIONS administration
method, 322
SOME predicate, 34
sorting (ORDER BY clause), 148
spaceoptimize
server.ini parameter, 365
split
partition, 277
SQL (Structured Query Language)
executing queries, 203
tracing last statement, 318
tracing statements, 318
SQL Server

Index
FETCH_ALL_RESULTS hint
constraint, 351
RETURN_TOP, 358
RETURN_TOP hint constraint, 358
SQL_DEF_RESULT_SET, 361
square brackets for index values, 162
statistics
RDBMS, determining, 186
storage areas
archiving files, 293
making read-only, 320
moving offline/online, 320
restoring archived content, 306
setting default, 60
string datatypes
non-qualifiable property
constraint, 55, 81
SUBSTR (function), 22
SUM (aggregate function), 24
SUMMARY (keyword), 128
Superuser user privilege
RDBMS permissions, 346
switch
partition, 277
SYNONYM clause, 112
SYNONYM FOR clause internationaliztion
requirements, 113
syntax
dates, 15
SYSOBJ_ID (keyword), 128
SysObjects
predicates, 38
system administration
execute (statement), 99
T
table permits, 345
tables
deleting rows, 91
inserting rows, 107
querying, 125
registering, 110
unregistering, 153
updating rows, 155
tablespaces
moving type indexes, 275
TEXT (keyword), 128
THUMBNAIL_URL (keyword), 130
thumbnails
generating, 298
TODAY (keyword), 17
TOMORROW (keyword), 17
tracing
operations, 318
full-text index operations, 273
login ticket keys, 319
login tickets, single-use, 319
retention policy use, 318
RPC calls, 243
trace file storage, 317
turning on/off, 317
workflow agent operations, 319
transactions
aborting explicit, 44
committing, 71
starting explicit, 67
TRANSCODE_CONTENT administration
method, 324
transformations, requesting, 324
TRUE (keyword), 19
truncating repeating properties, 163
TRY_FTDQL_FIRST DQL hint, 362
TYPE (predicate), 38
types
alter type statement, 51
create...object (statement), 74
deleting, 98
dropping properties, 57
modifying properties, 57
setting default ACL, 60
U
UNCOMMITTED_READ (database
hint), 362
underbar ( _ )
UNION clause, 147
unlinking
objects with UPDATE...OBJECT, 163
Unregister (statement), 153
UPDATE (statement), 155, 334
update modifier, 63
UPDATE...OBJECT (statement), 158, 334
update_modifier specification, 83, 87
UPDATE_STATISTICS administration
method, 327
UPPER (scalar function), 21

Index
URLs
MFILE_URL (DQL function), 31
THUMBNAIL_URL (keyword), 130
use_estimate_search (server.ini key), 202
USER (keyword), 18, 130
user privileges
granting, 105
revoking, 114
user privileges, DQL statements and, 346
users
method), 181
USING ASSEMBLIES, 138
V
va_clause (data dictionary), 84
value assistance
dropping, 65
format, 84
specification, 63
VERSION (keyword), 138
virtual documents
full-text indexes and, 344
IN DOCUMENT clause, 137
keywords for, 128
querying, 344
W
Web sites
publishing to, 329
WEBCACHE_PUBLISH administrative
method, 329
WHERE clause, 142
predicates, 32
valid qualifications, 142
wildcards for pattern matching, 36
workflow agent
tracing, 319
X
XML store
migrating content to/from, 259
Y
YESTERDAY (keyword), 17

Cs 7.2 dql

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to Cs 7.2 dql (20)

Recently uploaded (20)

Cs 7.2 dql