BITS: Introduction to relational databases and MySQL - Schema design

Introduction to MySQL

● Introduction
● Installation
● SQL
● Schema design
● Perl

BITS/VIB Bioinformatics Training – Jun 29, 2011 – Luc Ducazu <luc@daphnia.com>

Database schema
● Although you can execute DDL commands from
the MySQL monitor directly, this is not often
done
● There are tools that allow you to design a schema
graphically and generate the CREATE TABLE ...
statements
● Examples:
– MySQL workbench
– Dia

MySQL workbench
● Available as a standard package on some Linux
distros (eg Fedora)
● Available as Windows MSI, Linux DEB or RPM
package or source archive from
http://dev.mysql.com/downloads/workbench/
● To install a DEB package:
# dpkg i package.deb
● To install a RPM package:
# rpm Uvh package.rpm

Database schema
● Once the schema is designed, MySQL workbench
can generate a 'script' containing all SQL
statements to create the tables and other objects:
File >
Export >
Forward Engineer SQL CREATE Script
● This 'script' can be executed as usual from the
MySQL monitor

Inserting rows
● To populate tables, use the INSERT SQL
statement:
mysql> insert into tbl
       (col1, col2, ...)
       values
       (val1, val2, ...) [,
       (valx, valy, ...) , ...]
● With:
– tbl the name of the table
– col1, col2, ... a list (subset) of column names
– val1 value for col1
– val2 value for col2

Inserting rows examples
● Example (biodb version 1)
insert into modorg
(id, class, genus, species,
nchr, gsize, draft)
values
(1, “Bacteria”, “Escherichia”, “coli”,
1, 4.639, “19970905 00:00:00”)
● Note that strings and dates have to be properly
quoted

Inserting rows
● You can leave out the column list if
– a value is given for all columns
– the values are specified in the order dictated by the
schema
mysql> insert into tbl
       values
       (val1, val2, ...) [,
       (valx, valy, ...) , ...]

Changing rows
● To change one or more rows, use the UPDATE
SQL statement:
mysql> update tbl
set col1=expr1 [, col2=expr2, ...]
[where cond]
● With:
– col1 the column to change
– expr1 the new value
– cond the row filter if unspecified, all rows of the
table will be updated

Changing rows examples
● To change the C elegans number of
chromosomes to 7:
update modorg
set nchr = 7
where genus = “caenorhabditis”
and species = “elegans”;
● To change the draft date to the current date:
update modorg
set draft = now();

Deleting rows
● To remove rows, you use the DELETE SQL
statement:
delete from tbl
[where cond]
● With:
– cond the row filter if unspecified, all rows of the
table will be deleted
– note: since you can only remove entire rows, there is
no need to specify column names

Deleting rows examples
● To remove all in model organisms with a genome
publishing date before 2000:
delete from modorg
where year(draft) < 2000;

Perl
● Perl is a scripting language that has excellent text
manipulation capabilities
● Many biological 'databases' are available as flat
text files
● Perl is very handy in the automation of the
population of MySQL databases

Automated population
● There are basically two ways Perl can help to
insert data into tables:
– The Perl script generates USE, INSERT, ... SQL
statements. You can than execute these statements
using the MySQL monitor:
$ perl myscript.pl | mysql
– The Perl script connects to the database and executes
SQL statements directly: DBI

Perl DBI
● Perl DBI provides a programming interface that
abstracts most of the RDBMS specifics
● In principle it should be possible to port scripts,
written for other RDBMS's (like PostgreSQL), to
MySQL with only minimal effort: all you have to
do is change the connection string
● Packages to install (Ubuntu)
– libdbiperl: Perl DBI
– libdbdmysqlperl: MySQL driver for DBI

Perl DBI connecting
● Here is a minimal program:

#!/usr/bin/perl w
use strict;
use DBI;

my $dbh = DBI>connect(
     “DBI:mysql:host=localhost;database=biodb”,
     “user”,
     “password”) or die;
...
$dbh>disconnect();

Perl DBI connecting
● Some highlights:
– use DBI;
Load Perl DBI library
– DBI>connect(connection string);
Connect to a database
● it is a MySQL database server
● the DB server is running on the local machine
● you can provide the name of the database
● you can provide a user name and password
The connect() function returns
● a database handle ($dbh) on success
● false on failure
– $dbh>disconnect() to clean up resources

Perl DBI
● To execute a SQL statement that does not return a
result set, eg INSERT, DELETE, use do():
my $n = $dbh>do(stmt);
● This function
– requires a valid database handle
– returns the number of rows affected, if no rows are
affected, a special value is returned: 0E0 (evaluates
true)
– false in case of an error
● You can use execute() as well

Perl DBI
● To execute a SQL statement that returns a result
set, eg SELECT use the following recipe:
1. Prepare a statement:
my $sth = $dbh>prepare(stmt);
2. Send the query to the database server:
$sth>execute();
3. Read the result, row by row:
my @row = $sth>fetchrow_array();
my $ref = $sth>fetchrow_hashref();
4. Clean up resources:
$sth>finish();

Perl DBI examples
● To list all classes in modorg:
...
my $qry = “select distinct class ”
        . “from modorg ”
        . “order by class”;
my $sth = $dbh>prepare($qry);
$sth>execute();
while(my @row = $sth>fetchrow_array())
{
   print “$row[0]n”;
};
$sth>finish();
...

Perl DBI examples
● To list all organisms in modorg:
...
my $qry = “select genus, species ”
        . “from modorg ”
        . “order by genus”;
$sth>execute();
while(my $ref = $sth>fetchrow_hashref())
{
   print $ref>{“genus”} . “ “
       . $ref>{“species”} . “n”;
};
$sth>finish();
...

Perl DBI examples
● To insert a bunch of rows into modorg:
...
my $qry = “insert into modorg ”
        . “(id, class, genus, species) ”
        . “values (?, ?, ?, ?)”;

$sth>execute(11, “Mammels”,
      “Sus”, “scrofa”);
...

Quoting
● Never allow arbitrary user input in SQL
statements

BITS: Introduction to relational databases and MySQL - Schema design

More Related Content

What's hot (20)

Viewers also liked (6)

Similar to BITS: Introduction to relational databases and MySQL - Schema design (20)

More from BITS (20)

Recently uploaded (20)

BITS: Introduction to relational databases and MySQL - Schema design