MODULAR AND DIDACTIC COMPILER DESIGN WITH XML INTER-PHASES COMMUNICATION

International Journal of Computer Science, Engineering and Information Technology (IJCSEIT)
Vol.12, No.1, February 2022
DOI : 10.5121/ijcseit.2022.12101 1
MODULAR AND DIDACTIC COMPILER DESIGN
WITH XML INTER-PHASES COMMUNICATION
Eduardo Adam Navas-López
Computer Science Department, Mathematics Faculty,
University of El Salvador, San Salvador, El Salvador
ABSTRACT
In Compiler Design courses, students learn how a program written in high level programming language
and designed for humans understanding is systematically converted into low level assembly language
understood by machines, through different representations. This article presents the design, educative
characteristics and possibilities of a modular and didactic compiler for a Pascal-like programming mini-
language that is super-set of Niklaus Wirth's PL/0. The main feature is that it implements the compiling
phases in such a way that the information delivered to each next one may be reflected as an XML
document, which can be studied separately. It is also shown that its design is suitable for being included as
learning tool into compiler design courses. It is possible to implement a compiler in a high-level language
like Python.
KEYWORDS
Compiler Design, XML intercommunication, Compiling phases, PL/0, Compiler Education.
1. INTRODUCTION
In Compiler Design courses, students learn how a program written in high level programming
language designed for humans understanding, is systematically converted into low level assembly
language understood by machines (or by virtual machines). Some researchers think that it is no
exaggeration to say that compilers and high-level languages are central to the information age [1].
But the discussion about including (excluding, or why not excluding) compiler design courses in
computer science undergraduate programs has long history, as Parnas [2] and Henry [3] show.
Recently, Gruner [4] and others [5] continues the discussion by arguing the importance of
learning about compiler design and construction, despite the reluctance of some students [3] and
some employers [2].
When studying compiler design, we always talk about the traditional phases: Lexical Analysis,
Syntactic Analysis, Semantic Analysis, Intermediate Code Generation, and Object Code
Generation [6]. But the default behavior of traditional professional compilers is to hide them to
give programmers a fast and effective response. This is reasonable when one just want an
executable file from a set of source files. Nevertheless, when studying compiler design and
construction, students would like to see the process performed by the compiler phases, or
products of that phases; however, traditional compilers do not allow displaying that kind of
information. Many educational efforts have been developed in this direction. See section 2.
Hall et al. [1] call to develop methodologies and repositories that enable the comparison of
methods and reproducibility of results, and to develop curriculum recommendations on compiler
technology. So, experiments, good experiences and educative proposals in compiler education are

International Journal of Computer Science, Engineering and Information Technology (IJCSEIT),
2
very important. Moreover, literature reviews reveal the importance of studying multiple semiotic
representations and different abstraction levels in computer science [7], and in algorithmic
thinking [8, sec. 2.3]. Thus, this article presents a modular and didactic compiler for a super-set
of PL/0, called here Language pl0+ (pee-ehl-zero-plus, or “Pe-eLe-Cero-Más” in Spanish),
programmed in Python 3.8, that can reveal the data that is delivered to each next compile phase in
the form of an XML document as different semiotic representations of a same computer program.
See source code and examples at [9]. The target assembly language is a modified version of PL/0
Machine instruction set, based on that used by [10, p. 210] and proposed by Niklaus Wirth [11, p.
331], called here Language p+ (pee-plus, or “Pe-Más” in Spanish). Also an interpreter for p+ is
included. See section 3.
It is also shown that it is possible and reliable to implement a compiler in a high-level language
like Python, such as argued by [12] and showed by [13]. After all, the most remarkable
accomplishment by far of the compiler field is the widespread use of high-level languages [1]. In
section 4, didactic features and possibilities of that compiler, their design, source language, target
language, and the interpreter are discussed. Finally, author's conclusions are presented in section
5.
2. RELATED WORK
Academic literature on compiler education dates back to the mid-1960s and due to the wide-
spread standardization of compiler construction, publications on this topic appear in only
irregular frequency and not in large numbers [4]. Several educative efforts have been performed
around compiler design, from games for computer languages implementation [14], and teaching
mathematics through a compiler [15], to assembly language simulators [16], and computer
architecture simulators [17]. Some efforts focuses on low level, like Jordan et al.'s compiler [13]
that was written in Python for translate C-code to GAMA32 processor instruction set, or Nakano
and Ito's compiler [18] that takes Perl source code to run on Spartan-3 kit hardware. However, the
review of this research focuses on high level.
2.1. Courses (re)design motivations
Most of compiler education research is motivated by enhancing compiler education itself: Aiken
[19], Resler and Deaver [20], Baldwin [21], Mernik and Žumer [22], Almeida-Martínez et al.
[23], and Urquiza-Fuentes et al. [24], have developed a diversity of tools for enhancing and
complementing their courses, and to share their proposals. A good compiler course combines data
structures, algorithms, and tools for students as they build a large piece of software that performs
an interesting and practical function [1].
Demaille et al. [25] present their selection of tools for dealing with a massive course (up to 250
students) while manage and facilitate evaluation of student projects, to highlight some rarely used
features that enhance (compiler construction) learning. Kundra and Sureka [26] presents their
Case-Based and Project-Based Learning Approaches on compiler design concepts. Some other
research is motivated by the idea that “Students will (most likely) never construct compilers in
their future works”, like Henry's [3]. Gruner [4] is worried because some universities nowadays
(2019) might feel tempted to dilute (if not entirely abolish) a number of classical courses (like
Compiler Construction) that are now being regarded as “too theoretical”, “not practical enough”,
or “not industrially relevant”, as revealed by [27, 28]. And he [4] exposes their experience and
their ideas about compiler construction and their curricular relevance, despite the fact that some
industry employers think that abstract academic projects are not very important [29]. Another
researchers, like Na and ShiMing [30] and Wang and Li [5], have institutional needs for justify

3
teaching compiler design courses due to non-theoretical orientation of some universities. Both
papers, [30] and [5], also discuss online education and web-based teaching on compiler design
courses. More relaxed, Sumii [31] developed a compiler in Objective Caml for supporting
functional paradigm, for demonstrate that functional languages are simple and efficient. Some
papers on compiler education present curricular changes or discussions on this topic around the
world, in India [26], China [5, 30], Japan [31, 18, 32], South Africa [4], and United States [1].
Compiler courses must clearly demonstrate to students the extraordinary importance, range of
applicability, and internal elegance of what is one of the most fundamental enabling technologies
of computer science [1].
2.2. Diversity of compiler design courses and projects
Most compiler design courses are performed such a way the students must to implement
compilers for a programming mini-language. Some examples are object oriented, like Cool [19],
or Game Programming Language (java-like) [3]. Some others are functional, like MinimL [21],
minCaml [31] and Sarkar et al.'s [33, 34] (subset of Scheme). But the most are
imperative/structured, much of them subsets of Pascal (or Pascal-like) like Resler and Deaver's
example [20] (Pascal-like), Language X [35] (subset of Pascal), Niklaus Wirth's PL/0 [22]
(subset of Pascal), Niklaus Wirth's Oberon [36] (Pascal-like), and SimplePascal [24]. Also there
is another, C-like [18], and also compilers for another imperative languages [4].
The implementation language also varies. Some use ANSI C [20], C++ [19, 3, 25], Java [21, 22],
Objective Caml [31], Scheme [34], Perl [18], and Python [13] (this last is not educative, but it is
relevant). The target architecture/target assembly language also differs. Some educative
compilers are very specialized for specific hardware, like GAMA32 processors [13], or
TINYCPU/Spartan-3 [18]. Another are defined for produce executable object code for more
popular architectures like SPARC [31], RISC [36], MIPS [19, 16, 25], or Intel-PC assembler [20,
21, 3]. Even some of them design their own pseudo-assembly language, like Evangelidis et al.'s
[35] (similar to PL/0 Machine instruction set). Some compiler education tools and compiler
design projects focuses so much in high-level compilation phases that lack of code generation
analysis, like [22], [34], [23] and [24]. Gruner [4] even proposes a compiler design project whose
target language is not low-level but classic BASIC with GOTO statement.
Given that compiler projects are expensive to create, it is surprising that there are no standard,
widely used compiler projects [19]. There are many ideas about compiler design course projects:
Henry [3] design a game-oriented programming mini-language and developed a back-end java
source code for a compiler. The students must to construct compiler using that back-end, and to
program some simple computer games in it. Demaille et al. [25] propose to use many mini-
languages using several auxiliary computer-aided tools both to construct compilers and to
evaluate compilers partial source code from students. Kundra and Sureka [26] have designed
their course around “real-life situation” case-based projects, to achieve meaningful learning.
Sarkar et al. [33, 34] propose the “nanopass” methodology that supports decomposing a compiler
into many small pieces. This decomposition simplifies the task of understanding each piece and,
therefore, the compiler as a whole. They argue the nanopass tools enable a compiler student to
focus on concepts rather than implementation details.
2.3. Pedagogical tools for compiler education
Computer programs, like many other dynamic and abstract processes, are often best understood
by observing graphical simulations of their behavior [22]. Compilers are not exceptions. Mernik

4
and Žumer [22] presents LISA, an integrated development environment in which users can
specify, generate, compile-on-the-fly, and execute programs in a newly specified language. LISA
produces an interpreter or a compiler for a defined language written in the Java language from a
formal language specification; that is, LISA is an IDE for a compiler-compiler. It shows
deterministic finite state automata animations, syntax tree animations, and semantic tree
animations. See LISA website [37]. Resler and Deaver [20] presents VCOCO (Visible COmpiler
COmpiler), a program that generates LL (1) visible compilers. It also is a compiler-compiler.
This program allows to view the produced compiler source code, during the source code of
programs is being compiled by it. Almeida-Martínez et al. [23] developed VAST, which allows
to display different syntax error recovery strategies. VAST has been divided in two parts:
VASTapi and VASTview. VASTapi is the part encharged of the language processing, its target is
to interpret the actions made by the input parsers. Finally, it has to create an intermiddle
representation, in a XML file, with the content of the Syntax Tree and the necessary information
which allows it visualization. VASTview is the part encharged of the visualization. Its function is
to interpret and represent visually the content of the XML created by VASTapi. Urquiza-Fuentes
et al. [24] developed SOTA (SymbOl Table Animation), an educational tool aimed at visualizing
the working of a symbol table during the source code analysis.
Compiler design courses often use compiler-writing tools to simplify projects, ranging from LEX
and YACC to more modern tools, like report [20], [35], [3], [25], [18], [23], and [26]. But
because these tools deliberately hide many of the details of how a compiler works to the
programmers, they do not aid a course whose goal is to develop an understanding of those
workings. Moreover, according to Mernik and Žumer [22], these tools usually have little or no
didactic value, they were not designed for educational purposes, but rather for experienced
compiler writers (see [13]) where efficiency, space optimizations, modularity, and portability of
generated evaluators were primary concerns. Because of that, several compiler education
researchers do not use that kind of tools, like [19], [21], [22], [34], [31] and [4].
2.4. Error messages and language barriers
Error messages from compilers also has impact on programming learning due to the feedback
[38] and are often cryptic and pose a barrier to success for novice programmers who have been
shown to have trouble interpreting them [39]. Hattori and Kameda [32] and Becker et al. [40, 41]
have developed tools for enhancing Java compiler error messages providing more information
and changing description text for some common errors (through special front-end for the
compiler), achieving good results. On the other hand, Zhou, et al. [42] have not found the
enhanced messages to be useful for help to reduce student errors or improve students’
performance in debugging. Becker et al. [43] present reliable evidence that this heuristic
technique “fix the first error and ignore the rest” is trustworthy.
Reestman and Dorn [44] expose and discuss how English-only reserved words, documentation,
IDEs, and compiler error messages undoubtedly are barriers to learning to code for non-English
native learners, even for low English proficiency programmers. Guo [45] surveyed users of a
Python-language programming website, and discovered that low English proficiency
programmers have problems learning coding because of the built-in English language nature of
programming. Also, Qian and Lehman [46] reported that students’ English ability was
significantly correlated with their success in learning to program.

5
2.5. Similar proposals of compiler learning compilers and tools
In Aiken's COOL [19] design, although it is modular, since each phase can be compiled
separately, communication between that phases is done through internal data structures,
implemented in classes [47]. So you can't see the output of each phase independently of the
compiler's source code. VCOCO [20] is a compiler-compiler front-end that shows and highlights
produced compiler (scanner and parser) source code lines during source program compiling. It
also highlights the grammar production related to each token of source program. The X-Compiler
from Evangelidis et al. [35] is an IDE with an advanced mode that displays the assembly code
statically, and dynamic information during program execution (registers, variable values, and
program output). Moreover, compiler code it-self is not part of the didactic tool. This three tools
do not show the different representations of the source program, just object code.
Baldwin's [21] MinimL is a didactic monolithic compiler (written in Java) that does allow to see
the different representations of the program, during the different compilation phases: As a
sequence of unformatted characters, as a sequence of lexemes, as a syntax tree, and as assembly
code (see [48]). That representations are user-friendly, but are not structured enough to be input
for another software or system. The phases cannot be stopped, and the information produced by
the phases cannot be modified to enter it in a subsequent phase. LISA [22, 37] is similar to
VCOCO [20] but it has graphical animation of syntax tree instead of compiler code highlighting.
However, because it is a compiler-compiler, its use requires not only the specification of the
source language syntax, but also the specification of an attribute grammar. This makes it a very
powerful tool, but difficult to use, especially makes it difficult to interpret diagrams of syntax
trees and evaluation trees with semantics. VAST [23] focuses on error recovery visualization of
syntactic analysis, and lacks other phases. It use XML intermediate representation for syntax tree,
but it is not showed to user, so it does not exploit their educative and technical value. It is not a
modular tool. Finally, for SOTA [24], same as for VAST [23], compiler code it-self is not part of
the didactic tool. Both are animation tools for visualize specific parts of the typical compiler
process.
3. COMPILER, LANGUAGES AND INTERPRETER DESIGN
Next sub-sections briefly describe pl0+ (source language), p+ (target language), the compiler,
their phases, and the interpreter.
3.1. Compiler design principles
The compiler has been designed thinking not of its speed of execution, but of the possibility of
using it for academic-pedagogical purposes. Furthermore, according to Sumii, an efficient
compiler means a compiler that generates fast code, not a compiler which itself is fast [31].
Augier et al.'s [12] work shows that the performance of scientific programs depends less on
languages than on the time spent on optimization and the developer skills to correctly use the
right tools. Their benchmarks demonstrate that dynamic languages like Python can actually be
good solutions to easily obtain good performance while retaining simplicity and readability. That
is why the Python language was selected for the compiler and interpreter.
Similar as Baldwin's [21, 48] MinimL compiler, this compiler can output different static
representations of the source program after the different compilation phases: As a linear sequence
of lexemes, as an arborescent syntax tree, as an arborescent semantic tree, as linearized
intermediate code, and as assembly-like code (p+ code). However, with this compiler, users can
alter the produced output of a phase, and then to input it to the next phase(s). User even can alter

6
assembly code, and then execute it. Alternatively, users can write their assembly programs from
scratch, like in Language X [35]. Output from each phase has form of XML document. If user
indicates it, compiler will create a file containing corresponding pretty-printed XML document.
The default behavior of the compiler is to execute all the compilation phases without writing in
secondary memory the intermediate products of the intermediate phases. When errors and
warnings occur during compiling, compiler errors messages can be showed as traditional terminal
text through standard output (stdout), or as XML document through standard error output (stderr)
with structured information about errors.
As Reestman and Dorn [44] call, educative research and tools development should making
programming more accessible to others (Spanish-speaker low English proficiency students in this
case) as we all continue moving towards a more technological world. Because of that, compiler
source code, interpreter source code, compiler error messages, source language grammar, target
language mnemonics, and XML tags are in Spanish, not in English.
3.2. Source language
The source language is a super-set of the well-known PL/0 language created by Niklaus Wirth
[11]. Several variations are used in compiler education research [20, 35, 22, 24] and other
educational tools [49, 50]. Here it will be known as the pl0+ language and its grammar is
presented below:
<programa> ::= <bloque> '.'
<bloque> ::=
[<declaración_constantes>] [<declaración_variables>] <declaración_procedimiento>*
<instrucción> [';']
<declaración_constantes> ::= 'const' <identificador> '=' [ '+' | '-' ] <número>
( ',' <identificador> '=' [ '+' | '-' ] <número> )* ';'
<declaración_variables> ::= 'var' <identificador> ( ',' <identificador> )* ';'
<declaración_procedimiento> ::= 'procedure' <identificador> ';' <bloque> ';'
<instrucción> ::=
<identificador> := <expresión> |
'call' <identificador> |
'begin' <instrucción> ( ';' <instrucción> )* [';'] 'end' |
'if' <condición> 'then' <instrucción> [ 'else' <instrucción> ] |
'while' <condición> 'do' <instrucción> |
'read' <identificador> |
'write' <identificador> |
<nada>
<condición> ::= 'odd' <expresión> |
<expresión> ( '=' | '<>' | '<' | '>' | '<=' | '>=' ) <expresión>
<expresión> ::= [ '+' | '-' ] <término> ( [ '+' | '-' ] <término> )*
<término> ::= <factor> ( [ '*' | '/' ] <factor> )*
<factor>::= '-'* ( <identificador> | <número> | '(' <expresión> ')' )
<identificador> ::= <letra> ( <letra> | <dígito> | '_' )*
<número> ::= <dígito> <dígito>*
<letra> ::= 'a' | ... | 'z'
<dígito> ::= '0' | ... | '9'
<comentario> ::= '(*' <cualquier-caracter>* '*)'
It is a simple high-level programming language that allows procedure nesting, direct and indirect
recursion, it only has 32-bit integer variables and constants, and has the basic arithmetic and
relational operators for conditions. The procedures do not return any value, that is, there are no
functions. And it only has basic integer input and output instructions for standard input and
standard output respectively.
3.2.1. Example program to calculate and show Fibonacci numbers
１. (* Cálculo de los números de la serie de fibonacci:
２. f_0 = 1

7
３. f_1 = 1
４. f_n = f_{n-1} + f_{n-2}, n>1
５. *)
６. var n, f;
７. procedure fibonacci;
８. var i, (* Variable para hacer el recorrido *)
９. f_1, (* Número Fibonacci anterior *)
１０. f_2; (* Número Fibonacci anterior al anterior *)
１１. begin
１２. if n=0 then f:=1;
１３. if n=1 then begin
１４. f:=1;
１５. write f; (* Los primeros dos elementos son iguales *)
１６. end;
１７. if n>1 then begin
１８. f_1:=1;
１９. write f_1;
２０. f_2:=1;
２１. write f_2;
２２. i:=2;
２３. while i<n do begin
２４. f:=f_1+f_2;
２５. write f;
２６. f_2:=f_1;
２７. f_1:=f;
２８. i:=i+1;
２９. end;
３０. f:=f_1+f_2;
３１. end;
３２. end; (* fin del procedimiento *)
３３. begin
３４. read n;
３５. call fibonacci;
３６. write f;
３７. end.
3.3. Target Language
The target language is a variant of the p-code defined for PL/0 [10, p. 210]. It is a simple
assembly language and here it will be known as p+ language. Table 1 presents the definition of
its instructions and mnemonics.
Table 1. p+ language mnemonics.
Instruction Parameters Description
LIT <val> Loads literal <val> onto the stack
CAR <dif>
<pos>
Loads (Carga/Cargar in Spanish) the value of the variable that is at
position <pos> in the block defined at <dif> static levels from the
current block at the top of the stack
ALM <dif>
<pos>
Stores (Almacena/Almacenar in Spanish) datum on top of the stack
in the variable that is at position <pos> in the block defined at
<dif> static levels from the current block
LLA <dif>
<dir>
Calls (Llama/Llamar in Spanish) a procedure defined at <dif>
static levels from the current block, starting at address <dir>
INS <num> Instantiates a procedure, reserving space for the <num> variables of
the block that implements it (this number includes the cells necessary
for the execution of the code, which in the case of language p+, as in
the case of p-code, are 3 additional integers
SAL <dir> Unconditional jump (Salto in Spanish) to address <dir>
SAC <dir> Conditional jump (Salto Condicional in Spanish) to the address
<dir> if the value at the top of the stack is zero
OPR <opr> Arithmetic or relational operation, depending on the number <opr>.

8
Parameters are the values that are currently at the top of the stack and
the result is pushed right there. Possible values for <opr> are:
1 Negative (additive inverse)
2 Addition (+)
3 Subtract (-)
4 Multiplication (*)
5 Division (/)
6 odd operator
8 Equal to (=)
9 Not equal to (<>)
10 Less than (<)
11 Greater than or equal to (>=)
12 Greater than (>)
13 Less than or equal to (<=)
RET Return of procedure
LEE Reads (Lee/Leer in Spanish) a value from standard input (stdin) and
stores it on the top of the stack
ESC Writes the value from the top of the stack to standard output (stdout)
3.4. Compiler Interface
The compiler interface is by command line. The general syntax to invoke it is as follows:
$ python3 compilador.py [-a] [-m] [-x] [--lex] [--sin] [--sem] [--gen] program
The options and their meanings are presented in table 2.
Table 2. Parameter options for compiler.
Short Option Long Option Description
-a --ayuda Shows a help message and ends immediately
-m --mostrar If the compilation is successful, it prints the result of the
process on stdout
-x --errores-xml Prints errors and warnings to stderr in XML format
--lex Executes lexical analysis phase
--sin Executes syntax analysis phase
--sem Executes semantic analysis phase
--gen Executes object code generation phase
Running the command attempts to compile the file program. If no particular phase is indicated,
all are assumed. If the compilation is successful, a file with a different extension is generated,
depending on the last phase executed. The extension of the pl0+ programs is assumed to be
.pl0+, the lexical analysis output extension is assumed to be .pl0+lex, the syntax analysis

9
output extension is assumed to be .pl0+sin, the semantic analysis extension is assumed to be
.pl0+sem, and the object code generation extension is .p+. Several combinations can be
executed:
$ python3 compilador.py program.pl0+
$ python3 compilador.py --lex program.pl0+
$ python3 compilador.py --lex --sin program.pl0+
$ python3 compilador.py program.pl0+ --lex --sin --sem
$ python3 compilador.py program.pl0+lex --sin --sem
$ python3 compilador.py program.pl0+sin --sem --gen
From the previous lines, the first one executes all the compilation phases on the source file
program.pl0+ and generates a file called program.p+. The second one only executes the
lexical analysis phase and generates a file called program.pl0+lex. The third one executes
the lexical and syntactic analysis phases and generates a file called program.pl0+sin. The
fourth one executes the lexical, syntactic and semantic analysis phases and generates a file called
programa.pl0+sem. The fifth one takes a file program.pl0+lex with the XML list of
tokens from a source program, runs the syntactic and semantic analysis phases, and generates a
file called program.pl0+sem. The sixth one takes a program.pl0+sin file containing the
syntax tree of a source program, runs the semantic analysis and object code generation phases,
and generates a file called program.p+. Input and output file extensions are resumed in figure
1. Obviously not all combinations are possible. For example, it cannot be requested to run both
the lexical analysis (--lex) and semantic analysis (--sem) phases only. In such cases, the
compiler will respond with an error message to the user. It should be noted that only the file of
the last executed compilation phase is created, and not the intermediate ones. In addition, a
CDATA element is included at the end of each output file, containing the source code of the
source program if it was received by the previous phase.
Figure 1. Scheme of compiler phases and related extensions
3.5. Lexical Analysis Description
In any compiler, the purpose of this phase is to convert the characters of the file that contains the
source program, into a linear sequence of the minimum elements with a meaning in the language
(and their eventual values). These minimal elements with meaning are called Lexical Elements or
Tokens [36, 6]. For each lexical element of the pl0+ programs, the lexical analysis phase
generates an XML tag that represents it. Each tag has column, line, and length attributes, which
indicate respectively the column where the item begins, the line it is in, and the length of the
lexical item it represents.

10
The reserved words of the pl0+ language are: 'begin', 'call', 'const', 'do', 'end', 'if', 'odd',
'procedure', 'then', 'var', 'while', 'else', 'write' and 'read'. Each one of them is
represented by a label with the same name but uppercase. Each identifier are represented by an
IDENTIFICADOR-tag with a name attribute. Integer literals are represented by NUMERO-tags
with a value attribute. The others lexemes are represented by XML tags with names in table 3.
Table 3. XML Tag names for symbols.
Symbol Tag name Symbol Tag name
= igual > mayor_que
:= asignacion <= menor_igual
, coma >= mayor_igual
; punto_y_coma + mas
( parentesis_apertura - menos
) parentesis_cierre * por
<> diferente / entre
< menor_que . punto
The result of applying the lexical analysis (up to the first 16 lines of source program) on the
Fibonacci numbers program (see section 3.2.1) is presented in Appendix A.
3.6. Syntactic analysis description
The purpose of this phase is to build a syntax tree —and at the same time check whether it is
possible to build one— from the sequence of lexical elements provided by the lexical analysis
phase. This tree represents how the source program, as a linear sequence of terminal symbols, is
derived from the initial symbol of the source language grammar [36, 6]. In this case, from the
non-terminal symbol <programa>. In this compiler, the syntax tree is represented by means of
an arborescent XML document. The general structure of the syntax tree for any pl0+ program is
as follows:
<arbol_de_sintaxis>
<programa>
<bloque>
...
</bloque>
</programa>
<fuente> <![CDATA[...]]> </fuente>
</arbol_de_sintaxis>
According to the pl0+ syntax, every program is made up of a main codeblock (<bloque>).
Every block is represented as a sequence of constant, variable and procedure declarations and
optionally a statement:
<bloque>
<constante columna="15" linea="7" nombre="fib_1" valor="1"/>
...
<variable columna="4" linea="8" nombre="n"/>
...
<procedimiento columna="10" linea="10" nombre="fibonacci">
<bloque> ... </bloque>
</procedimiento>
...

</bloque>
Another statements are represented like in table 4.

11
Table 4. Statement representation in parsing tree.
Statement Example XML fragment
Assignment n := ...; <asignacion variable="n">

</asignacion>
procedure
calls
call fibonacci; <llamada procedimiento="fibonacci"/>
begin/end begin ... end <secuencia>

</secuencia>
if if ... then ...
[else ...]
<condicional>
<condicion operacion="...">

</condicion>


</condicional>
while while ... do ... <ciclo>
<condicion operacion="...">

</condicion>

</ciclo>
read read n <leer variable="n"/>
write write f <escribir simbolo="f"/>
The parsing of the Fibonacci numbers program (section 3.2.1) is shown as a complete example in
Appendix B.
3.7. Semantic analysis description
There are several tasks that must be performed during the semantic analysis phase. These are
described in [6]. In this compiler, these tasks are: (1) Put a unique id code to each identifier
(variable, constant or procedure) and codeblocks (<bloque> tags) to be referenced later. The id
code includes information about the symbol type and the scope in which it is declared. (2) Verify
duplication of symbols in the same scope. (3) Check for procedure identifiers referenced in an
expression or in an assignment (which is not valid in pl0+). (4) Check for referenced constant
identifiers on the left side of an assignment. (5) Verify that each referenced identifier/symbol is in
a valid scope.
In the case of a valid syntax tree as input, the output of this phase is a revised syntax tree,
essentially the same as the input tree, but including that id code for each identifier and each
codeblock. In Appendix C are presented the first 20 lines of the result of this phase for the
Fibonacci numbers program (corresponding to first 12 lines from the source program, see section
3.2.1).
3.8. Intermediate and object code generation description
The purpose of intermediate code generation is to create a representation of the program that can
be easily translated into a low-level language (assembler or binary) without the abstractions of
high-level languages [6]. In this compiler, the last phase is carried out with the objective of
transforming the revised syntax tree into “machine code” for a virtual machine that exclusively
interprets programs in the p+ language. But first, an intermediate representation is created in an
XML document that includes a tag for each instruction of the corresponding language p+. In fact,

12
the interpreter interprete.py executes files in that intermediate XML representation of p+
language and not the assembly-fashion p+ language. The general translation rules to the p+
language are very similar to those described in [11, sec. 5.10 and 5.11]. Additional information
has been added to each instruction tag (as informacion-tag child node) for enhancing human
reading. But theese are not needed for interpreter execution.
The result of the code generation phase (selected lines) of the Fibonacci numbers program (see
section 3.2.1) is presented in Appendix D.
3.9. Interpreter interface
The interpreter interface is by command line. The general syntax to invoke it is as follows:
$ python3 interprete.py [-a] [-d] object-program.p+
The options and their meanings are presented in table 5.
Table 5. Parameter options for interpreter.
Short
Option
Long
Option
Description
-a --ayuda Shows a help message and ends immediately
-d --
depurar
Executes each instruction doing a wait, so that the user
can observe the value of the virtual machine registers
3.10. Error management and report
In this compiler, errors are recorded in two internal lists that are passed from phase to phase
while that is possible, differentiating between “errors” and “warnings”. Errors are those
fragments of code that make compilation impossible because it is too difficult or impossible to
determine the programmer's intent. And warnings are fragments of code that allow compilation,
because the programmer's intent can be assumed, but the source program is not fully valid, and
the result of the compilation may not quite match the programmer's true intent. This compiler, as
well as many others, displays errors first, then warnings, and in each group, the errors are ordered
with respect to their appearance in the source code. Here is a pl0+ program with multiple errors,
and then the standard output of the compiler when trying to compile it:
１. const n=50;
２. var f, i;
３. begin
４. i := 2 % 4;
５. f := 9 - i * 2
６. if n<>1 then begin
７. i:=2;
８. while i+5 <= n+2 do begin
９. f1:=f; i:=i+1;
１０. end
１１. end;
１２. end.
Compiler output:
ERROR****************
Fase de origen:sin
Línea 4: Falta un operador
i := 2 % 4;

13
----------^
ERROR****************
Fase de origen:lex
Línea 4: Caracter inválido.
i := 2 % 4;
-----------^
ERROR****************
Fase de origen:sem
Línea 9: Referencia a variable no declarada
f1:=f; i:=i+1;
------------^
ADVERTENCIA**********
Fase de origen:sin
Línea 5: Falta un ';'
f := 9 - i * 2
------------------^
Errors and warnings can also be displayed as a structured XML file via standard error output
(stderr). The following file represents the same error report for the previous error-full program
but with the -x (or --errors-xml) option included when trying to compile it:
<?xml version="1.0" ?>
<errores_y_advertencias>
<errores>
<error columna="10" linea="4"><mensaje>Falta un operador</mensaje>
<contexto><![CDATA[ i := 2 % 4;]]></contexto>
<fase nombre="sin">Fase de análisis sintáctico</fase>
</error>
<error columna="11" linea="4"><mensaje>Caracter inválido.</mensaje>
<contexto><![CDATA[ i := 2 % 4;]]></contexto>
<fase nombre="lex">Fase de análisis léxico</fase>
</error>
<error columna="12" linea="9"><mensaje>Referencia a variable no declarada</mensaje>
<contexto><![CDATA[ f1:=f; i:=i+1;]]></contexto>
<fase nombre="sem">Fase de análisis semántico</fase>
</error>
</errores>
<advertencias>
<error columna="18" linea="5"><mensaje>Falta un ';'</mensaje>
<contexto><![CDATA[ f := 9 - i * 2]]></contexto>
<fase nombre="sin">Fase de análisis sintáctico</fase>
</error>
</advertencias>
</errores_y_advertencias>
To reduce the frequency of false error detections, it was decided to apply —at least in the
syntactic analysis phase— the heuristic rule that indicates that normally there is only one error in
one same line of code, similar to the rule stated in [43].
4. DIDACTIC FEATURES, PROJECT IDEAS AND RESEARCH POSSIBILITIES
In this section a list of didactic features and didactic possibilities are discussed briefly,
considering the technical possibilities offered by this design. Also project ideas and research
possibilities are described. The main advantage of using the compiler it-self is the possibility to
visualize the input program in its different representations: as high-level source code, as a
sequence of lexemes, as a syntax tree, as a semantic tree, and as assembly-like code. An
important additional feature is the possible interoperability with future tools and IDEs. Thanks of
XML intercommunication, compiler is relatively easy to connect to eventual future tools like
front-end editors, front-end animators (like SOTA [24]) or low-level code generators for another
architectures. Among others, it can easily connected to a graphical visualizer for syntax tree
(before and after semantic analysis) like those presented by [13, fig. 7-13], [22, fig. 4-5], and [23,
fig. 2]. Translation of source language and translation of XML tags can be a relatively easy task

14
for students to become familiar with compiler source code. Moreover, an interesting task would
be add a language option for XML tags (Spanish, Portuguese, English, French, etc.).
Some compiler-centered projects include several diverse enlargements of source language, like
including for-loops, repeat-loops, do-while; including arrays, for-each-loops, booleans, floats,
strings; including structures, functions, parameters, pointers, exceptions, dinamic memory,
garbage collection, classes, etc. Also including additional phases for optimization like those
proposed by [6], maybe in the form of nanopasses [34]. Adding phases is easy. The compiler is
implemented as a set of Python files organized as follows:
compilador.py --> main compiler interface
fases/
__init__.py --> Global settings and functions
lex.py --> Lexical analysis
sin.py --> Syntactic analysis
sem.py --> Semantic analysis
gen.py --> Code generation
interprete.py --> p+ interpreter interface
The __init__.py file contains a list called fasesDisponibles (“available phases” in
English) that determines which phases can be invoked from the main compiler interface. A new
phase can be added, including command line options by simply adding the information of the
module name, the file it is in, the name of the translation function in that module, the output
extension of that phase, and a description to be used in error messages. There is no need of
modify compiler main program. Other challenging ideas include modify the compiler code for
dealing with cascade error messages like presented by [40, 41] and [32]; and, of course,
enhancing error messages like done by [38, 39, 40, 41, 42].
Beyond compiler courses projects, there is some research ideas that can be developed from this
compiler design. By example, it is useful to refine observation and study of the effect of different
message styles on how well and quickly students identify errors in programs [51]. This can be
done by attaching compiler's standard input and outputs (specially stderr) to a collect-information
front-end (like those of [32] or [41]). More generally it can be used for explore multi-core
compiling techniques (see [52]). Finally, it can be integrated to more complex environments like
research-oriented operating system (similar to Amoeba [53] or Oberon [49, 50]), or research-
oriented Database Management Systems.
5. CONCLUSIONS
Compiling a minuscule language exposes students to most of what a compiler does without
overwhelming them [21]. So, in this work, a modular and didactic compiler for a minimal Pascal-
like language has been presented. It has XML inter-phases communication and error reporting,
which makes it appropriate to examine and study the different representations of a program, from
high-level source code to assembly-like object code. Without a software tool like this, the
compiler design topics are much harder to understand and treat [22]. It is a starting point tool for
teaching compiler design through many possible projects and extension ideas. So, this design is
suitable for being included into compiler design courses. Source code, with in-line
documentation, can be downloaded from [9]. Further research should be done in compiler
education field around this tool, as Hall et al. [1] say, “the compiler community must convey the
importance and intellectual beauty of the discipline [of compiler design] to each generation of
students”.

15
REFERENCES
[1] M. Hall, D. Padua, and K. Pingali, “Compiler research: the next 50 years”, Communications of the
ACM, vol. 52, no. 2, pp. 60–67, 2009, http://doi.org/10.1145/1461928.1461946.
[2] D. L. Parnas, “Education for computing professionals”, Computer, vol. 23, no. 1, pp. 17–22, 1990,
http://doi.org/10.1109/2.48796.
[3] T. R. Henry, “Teaching compiler construction using a domain specific language”, In Proceedings of
the 36th SIGCSE technical symposium on Computer science education, pp. 7–11, 2005,
[4] S. Gruner, “Back to BASIC in compiler construction”, In Proceedings 48th Annual Conference of the
Southern African Computer Lecturer's Association, pp. 158–166. 2019,
http://hdl.handle.net/2263/72031.
[5] N. Wang and L. Li, “A hybrid teaching reform scheme of compiler technology course based on
engineering education”, In Proceedings of the 5th International Conference on Information and
Education Innovations, pp. 16–19, 2020, http://doi.org/10.1145/3411681.3411695.
[6] A. V. Aho, M. S. Lam, R. Sethi, and J. D. Ullman, Compiladores – Principios, técnicas y
herramientas, México: Pearson Educación, 2008, ISBN 978-970-26-1133-2.
[7] C. Mirolo, C. Izu, V. Lonati, and E. Scapin, “Abstraction in Computer Science Education: An
Overview”, Informatics in Education, vol. 20, no. 4, pp. 615–639, 2021,
https://doi.org/10.15388/infedu.2021.30.
[8] E. A. Navas-López, “Una caracterización del desarrollo del densamiento algorítmico de los
estudiantes de las carreras de licenciatura en matemática y licenciatura en estadística de la sede
central de la Universidad de El Salvador en el período 2018-2020”, MSc. Thesis, University of El
Salvador, El Salvador, 2021, http://doi.org/10.13140/RG.2.2.17298.99524.
[9] E. A. Navas-López, “Compilador para pl0+ e intérprete de p+ escrito en Python 3.8”, 2022,
http://dx.doi.org/10.13140/RG.2.2.16448.66560/1.
[10] C. P. Salas McClyskey, “An attributed translation grammar for PL/0”, International journal of
computer mathematics, vol. 21, no. 3-4, pp. 205–227, 1987,
http://doi.org/10.1080/00207168708803567.
[11] N. Wirth, Algorithms + Data Structures = Programs, Prentice-Hall, 1976, ISBN 0-13-022418-9.
[12] P. Augier, C. F. Bolz-Tereick, S. Guelton, and A. V. Mohanan, “Reducing the ecological impact of
computing through education and python compilers”, Nature Astronomy, vol. 5, no. 4, pp. 334–335,
2021, http://doi.org/10.1038/s41550-021-01342-y.
[13] W. Jordan, A. Bejo, and A. G. Persada, “The development of lexer and parser as parts of compiler for
GAMA32 processor's instruction-set using python”, In 2019 International Seminar on Research of
Information Technology and Intelligent Systems (ISRITI), pp. 450–455, 2019,
http://doi.org/10.1109/ISRITI48646.2019.9034617.
[14] D. Rodríguez-Cerezo, A. Sarasa-Cabezuelo, M. Gómez-Albarrán, and J.-L. Sierra, “Serious games in
tertiary education: A case study concerning the comprehension of basic concepts in computer
language implementation courses”, Computers in Human Behavior, vol. 34, pp. 558–570, 2014,
http://doi.org/10.1016/j.chb.2013.06.009.
[15] D. Duffy and J. O'Donoghue, “Use of the compiler as a mathematics teaching resource”,
International Journal of Mathematical Education in Science and Technology, vol. 23, no. 5, pp. 637–
650, 1992, http://doi.org/10.1080/0020739920230501.
[16] K. R. Vollmar and P. Sanderson, “MARS: An education-oriented MIPS assembly language
simulator”, In Proceedings of the 37th SIGCSE technical symposium on Computer science education
(SIGCSE '06), 2006, http://doi.org/10.1145/1121341.1121415.
[17] H. Öztekin, F. Temurtaş and A. Gülbağ, “An assembler and compiler program design for an
educational purpose computer architecture simulator being used computer and electrical engineering
education”, Electronic Letters on Science and Engineering, vol. 5, no. 2, pp. 9–16, 2009,
https://dergipark.org.tr/en/pub/else/issue/29318/313752.
[18] K. Nakano and Y. Ito, “Processor, assembler, and compiler design education using an fpga”, In 2008
14th IEEE International Conference on Parallel and Distributed Systems, pp. 723–728. 2008,
http://doi.org/10.1109/ICPADS.2008.71.
[19] A. Aiken, “Cool: A portable project for teaching compiler construction”, ACM Sigplan Notices, vol.
34, no.7, pp. 19–24, 1996, http://doi.org/10.1145/381841.381847.

16
[20] R. D. Resler and D. M. Deaver, “VCOCO: a visualization tool for teaching compilers”, ACM SIGCSE
Bulletin, vol. 33, no. 3, pp. 199–202, 1998, http://doi.org/10.1145/290320.283123.
[21] D. Baldwin, “A compiler for teaching about compilers”, In Proceedings of the 34th SIGCSE technical
symposium on Computer science education, pp. 220–223, 2003,
[22] M. Mernik and V. Žumer, “An educational tool for teaching compiler construction”, IEEE
Transactions on Education, vol. 49, no. 1, pp. 61–68, 2003, http://doi.org/10.1109/TE.2002.808277.
[23] F. J. Almeida-Martínez, J. Urquiza-Fuentes, and J. Á. Velázquez-Iturbide, “Educational
visualizations of syntax error recovery”, IEEE EDUCON 2010 Conference, 2010,
http://doi.org/10.1109/educon.2010.5492467.
[24] J. Urquiza-Fuentes, F. Manso, J. Á. Velázquez-Iturbide, and M. Rubio-Sánchez, “Improving
compilers education through symbol tables animations”, In Proceedings of the 16th annual joint
conference on Innovation and technology in computer science education, pp. 203–207, 2011,
[25] A. Demaille, R. Levillain, and B. Perrot, “A set of tools to teach compiler construction”, ACM
SIGCSE Bulletin, vol. 43, no. 3, pp. 68–72, 2008, http://doi.org/10.1145/1597849.1384291.
[26] D. Kundra and A. Sureka, “An experience report on teaching compiler design concepts using case-
based and project-based learning approaches”, In 2016 IEEE Eighth International Conference on
Technology for Education (T4E), pp. 216–219, 2016, http://doi.org/10.1109/T4E.2016.052.
[27] C. Scaffidi, “Employers’ needs for computer science, information technology and software
engineering skills among new graduates”, International Journal of Computer Science, Engineering
and Information Technology, vol. 8, no. 1, pp. 1–12, 2018, http://doi.org/10.5121/ijcseit.2018.8101.
[28] N. Baidya, W. D. Mathuram, A. Renavikar, J. Kim, J. R. Jupally, and R. R. Guduru, “An overview of
computer science master degree programs in the state of california”, International Journal on
Integrating Technology in Education, vol. 7, no. 3, pp. 1–9 2018,
http://doi.org/10.5121/ijite.2018.7301.
[29] C. Scaffidi, “A survey of employers’ needs for technical and soft skills among new graduates”,
International Journal of Computer Science, Engineering and Information Technology, vol. 8, no. 5/6,
pp. 11–21, 2018, http://doi.org/10.5121/ijcseit.2018.8602.
[30] W. Na and Z. ShiMing, “Construction of compiler technology course in application-based
university”, In 2013 International Conference on Education Technology and Information System, pp.
633–636, 2013, https://doi.org/10.2991/icetis-13.2013.143.
[31] E. Sumii, “MinCaml: A simple and efficient compiler for a minimal functional language”, In
Proceedings of the 2005 workshop on Functional and Declarative Programming in Education, pp.
30–41, 2005, http://doi.org/10.1145/1085114.1085122.
[32] S. Hattori and H. Kameda, “Knowledge based compiler with e-TA for software engineering
education”, In Knowledge Based Software Engineering, A. Čaplinskas, H. Pranevičius and T.
Nakatani, Eds, pp. 265–278, 2010, http://www3.muroran-
it.ac.jp/wits/~hattori/pdf/hattori2010jckbse.pdf.
[33] D. Sarkar, O. Waddell, and R. K. Dybvig, “A nanopass infrastructure for compiler education”, ACM
SIGPLAN Notices, vol. 42, no. 9, pp. 201–212, 2004, http://doi.org/10.1145/1016848.1016878.
[34] D. Sarkar, O. Waddell, and R. K. Dybvig, “Educational pearl: A nanopass framework for compiler
education”, Journal of Functional Programming, vol. 15, no. 5, pp. 653–667, 2005,
http://doi.org/10.1017/S0956796805005605.
[35] G. Evangelidis, V. Dagdilelis, M. Satratzemi, and V. Efopoulos, “X-compiler: Yet another integrated
novice programming environment”, In Proceedings IEEE International Conference on Advanced
Learning Technologies, pp. 166–169. 2001, http://doi.org/10.1109/ICALT.2001.943890.
[36] N. Wirth, Compiler construction, Zürich, Swiss: Addison-Wesley, 2005. ISBN 0-201-40353-6.
[37] Institute of Computer Science, University of Maribor, “LISA website”, 2006,
https://labraj.feri.um.si/lisa/.
[38] G. Marceau, K. Fisler, and S. Krishnamurthi, “Measuring the effectiveness of error messages
designed for novice programmers”. In Proceedings of the 42nd ACM technical symposium on
Computer science education, pp. 499-504, 2011, https://doi.org/10.1145/1953163.1953308.
[39] B. Hartmann, D. MacDougall, J. Brandt, and S. R. Klemmer, “What would other programmers do:
suggesting solutions to error messages”. In Proceedings of the SIGCHI Conference on Human
Factors in Computing Systems, pp. 1019-1028, 2010, https://doi.org/10.1145/1753326.1753478.

17
[40] B. A. Becker, “An effective approach to enhancing compiler error messages”, In Proceedings of the
47th ACM Technical Symposium on Computing Science Education, pp. 126–131, 2016,
[41] B. A. Becker, G. Glanville, R. Iwashima, C. McDonnell, K. Goslin, and C. Mooney, “Effective
compiler error message enhancement for novice programming students”, Computer Science
Education, vol. 26, no.2-3, pp. 148–175, 2016, http://doi.org/10.1080/08993408.2016.1225464.
[42] Z. Zhou, S. Wang, Y. Qian, “Learning From Errors: Exploring the Effectiveness of Enhanced Error
Messages in Learning to Program”, Frontiers in Psychology, 2021,
https://doi.org/10.3389/fpsyg.2021.768962.
[43] B. A. Becker, C. Murray, T. Tao, C. Song, R. McCartney, and K. Sanders, “Fix the first, ignore the
rest: Dealing with multiple compiler error messages”, In Proceedings of the 49th ACM technical
symposium on computer science education, pp. 634–639, 2018,
[44] K. Reestman and B. Dorn, “Native language's effect on java compiler errors”, In Proceedings of the
2019 ACM conference on international computing education research, pp. 249–257, 2019,
[45] P. J. Guo, “Non-Native English Speakers Learning Computer Programming: Barriers, Desires, and
Design Opportunities”, In Proceedings of the 2018 CHI Conference on Human Factors in Computing
Systems, pp. 1–14, 2018, https://doi.org/10.1145/3173574.3173970.
[46] Y. Qian, and J. D. Lehman, “Correlates of Success in Introductory Programming: A Study with
Middle School Students”, Journal of Education and Learning, vol. 5, no 2, p. 73-83, 2016,
http://doi.org/10.5539/jel.v5n2p73.
[47] A. Aiken, “A Tour of the Cool Support Code”, Stanford University,
http://theory.stanford.edu/~aiken/software/cooldist/handouts/cool-tour.pdf.
[48] D. Baldwin, “Class MinimLCompiler Documentation”, SUNY Geneseo Department of Computer
Science, https://www.geneseo.edu/~baldwin/miniml/doc/geneseo/cs/miniml/MinimLCompiler.html.
[49] N. Wirth y J. Gutknecht, Project Oberon – The Design of an Operating System and Compiler, Edition
2005, https://people.inf.ethz.ch/wirth/ProjectOberon1992.pdf.
[50] N. Wirth, The Programming Language Oberon, Revision 1.10.2013 / 3.5.2016,
https://people.inf.ethz.ch/wirth/Oberon/Oberon07.Report.pdf.
[51] M.-H. Nienaltowski, M. Pedroni, and B. Meyer, “Compiler error messages: What can help novices?”,
In Proceedings of the 39th SIGCSE technical symposium on Computer science education, pp. 168–
172, 2008, http://doi.org/10.1145/1352135.1352192.
[52] A. Barve, and B. K. Joshi, “Issues in Implementation of Parallel Parsing on Multi-Core Machines”,
International Journal of Computer Science, Engineering and Information Technology, vol. 4, no. 5,
pp. 51–29, 2014, http://doi.org/10.5121/ijcseit.2014.4505.
[53] A. S. Tanenbaum y G. J. Sharp, The Amoeba Distributed Operating System, Vrije Universiteit, 1998,
https://www.cs.vu.nl/pub/amoeba/Intro.pdf.
APPENDIX A - LEXICAL ANALYSIS ON FIBONACCI NUMBERS PROGRAM
<lexemas>
<VAR linea="6" columna="0" longitud="3"/>
<IDENTIFICADOR nombre="n" linea="6" columna="4" longitud="1"/>
<coma linea="6" columna="5" longitud="1"/>
<IDENTIFICADOR nombre="f" linea="6" columna="7" longitud="1"/>
<punto_y_coma linea="6" columna="8" longitud="1"/>
<PROCEDURE linea="7" columna="0" longitud="9"/>
<IDENTIFICADOR nombre="fibonacci" linea="7" columna="10" longitud="9"/>
<VAR linea="8" columna="4" longitud="3"/>
<IDENTIFICADOR nombre="i" linea="8" columna="8" longitud="1"/>
<IDENTIFICADOR nombre="f_1" linea="9" columna="8" longitud="3"/>
<IDENTIFICADOR nombre="f_2" linea="10" columna="8" longitud="3"/>
<BEGIN linea="11" columna="4" longitud="5"/>
<IF linea="12" columna="8" longitud="2"/>

18
<igual linea="12" columna="12" longitud="1"/>
<NUMERO valor="0" linea="12" columna="13" longitud="1"/>
<THEN linea="12" columna="15" longitud="4"/>
<asignacion linea="12" columna="21" longitud="2"/>
<IF linea="13" columna="8" longitud="2"/>
<igual linea="13" columna="12" longitud="1"/>
<THEN linea="13" columna="15" longitud="4"/>
<BEGIN linea="13" columna="20" longitud="5"/>
<asignacion linea="14" columna="13" longitud="2"/>
<WRITE linea="15" columna="12" longitud="5"/>
<END linea="16" columna="8" longitud="3"/>
...
APPENDIX B - SYNTACTIC ANALYSIS ON FIBONACCI NUMBERS PROGRAM
<arbol_de_sintaxis>
<programa>
<bloque>
<variable linea="6" columna="4" nombre="n"/>
<variable linea="6" columna="7" nombre="f"/>
<procedimiento linea="7" columna="10" nombre="fibonacci">
<bloque>
<variable linea="8" columna="8" nombre="i"/>
<variable linea="9" columna="8" nombre="f_1"/>
<variable linea="10" columna="8" nombre="f_2"/>
<secuencia linea="11" columna="4">
<condicional linea="12" columna="8">
<condicion linea="12" columna="11" operacion="comparacion">
<identificador linea="12" columna="11" simbolo="n"/>
<numero linea="12" columna="13" valor="0"/>
</condicion>
<asignacion linea="12" columna="20" variable="f">
</asignacion>
</condicional>
</condicion>
</asignacion>
<escribir linea="15" columna="18" simbolo="f"/>
</secuencia>
</condicional>
<condicion linea="17" columna="11" operacion="mayor_que">
</condicion>
<asignacion linea="18" columna="12" variable="f_1">
</asignacion>
<escribir linea="19" columna="18" simbolo="f_1"/>

19
</asignacion>
<escribir linea="21" columna="18" simbolo="f_2"/>
<asignacion linea="22" columna="12" variable="i">
</asignacion>
<ciclo linea="23" columna="12">
<condicion linea="23" columna="18" operacion="menor_que">
<identificador linea="23" columna="18" simbolo="i"/>
</condicion>
<suma linea="24" columna="22">
<identificador linea="24" columna="19" simbolo="f_1"/>
</suma>
</asignacion>
</asignacion>
<identificador linea="30" columna="21" simbolo="f"/>
</asignacion>
<asignacion linea="28" columna="16" variable="i">
<identificador linea="28" columna="19" simbolo="i"/>
</suma>
</asignacion>
</secuencia>
</ciclo>
</suma>
</asignacion>
</secuencia>
</condicional>
</secuencia>
</bloque>
</procedimiento>
<leer linea="34" columna="9" variable="n"/>
<llamada linea="35" columna="9" procedimiento="fibonacci"/>
</secuencia>
</bloque>
</programa>
<fuente><![CDATA[...]]></fuente>
</arbol_de_sintaxis>
APPENDIX C – SEMANTIC ANALYSIS ON FIBONACCI NUMBERS PROGRAM
<arbol_de_sintaxis_revisado>
<programa>
<bloque codigo="b0">
<variable linea="6" columna="4" nombre="n" codigo="v0_0"/>
<variable linea="6" columna="7" nombre="f" codigo="v0_1"/>
<procedimiento linea="7" columna="10" nombre="fibonacci">
<bloque codigo="b0_0">
<variable linea="8" columna="8" nombre="i" codigo="v0/0_0"/>
<variable linea="9" columna="8" nombre="f_1" codigo="v0/0_1"/>
<variable linea="10" columna="8" nombre="f_2" codigo="v0/0_2"/>

20
<identificador linea="12" columna="11" simbolo="n" codigo="v0_0"/>
</condicion>
<asignacion linea="12" columna="20" variable="f" codigo="v0_1">
</asignacion>
...
APPENDIX D – INTERMEDIATE CODE OF FIBONACCI NUMBERS PROGRAM
<codigo_pmas>
<salto_incondicional direccion="0" parametro="27">
<informacion inicio_de_procedimiento="--PRINCIPAL--" codigo="b0"/>
</salto_incondicional>
<salto_incondicional direccion="1" parametro="2">
<informacion columna="10" linea="7" inicio_de_procedimiento="fibonacci"
codigo="b0_0"/>
</salto_incondicional>
<instanciar_procedimiento direccion="2" parametro="6">
<informacion columna="10" linea="7" inicio_de_procedimiento="fibonacci"
codigo="b0_0"/>
</instanciar_procedimiento>
<cargar_variable direccion="3" diffnivel="1" parametro="3">
<informacion linea="12" columna="8">Inicio de condicional (if-then)</informacion>
<informacion codigo="v0_0" linea="12" columna="11" variable="n"/>
</cargar_variable>
...
<cargar_variable direccion="59" diffnivel="0" parametro="4">
<informacion codigo="v0_1" linea="36" columna="10" variable="f"/>
</cargar_variable>
<escribir direccion="60"/>
<retornar direccion="61">
<informacion fin_de_procedimiento="--PRINCIPAL--"/>
</retornar>
<ensamblador><![CDATA[
0 SAL - 55
1 SAL - 2
2 INS - 6
3 CAR 1 3
...
59 CAR 0 4
60 ESC - -
61 RET - -
]]></ensamblador>
<fuente><![CDATA[...]]></fuente>
</codigo_pmas>
AUTHOR
Eduardo Adam Navas-López is full time Professor at University of El Salvador.
Received the B.Sc. Degree in Computer Science from Central American University
“José Simeón Cañas” and M.Sc. Degree in Mathematics Education from University of
El Salvador. His research interests are Computer Science Education, Mathematics
Education, Computer Graphics, Algorithmic Thinking, and Computational Thinking.

MODULAR AND DIDACTIC COMPILER DESIGN WITH XML INTER-PHASES COMMUNICATION

More Related Content

What's hot (12)

Similar to MODULAR AND DIDACTIC COMPILER DESIGN WITH XML INTER-PHASES COMMUNICATION (20)

Recently uploaded (20)

MODULAR AND DIDACTIC COMPILER DESIGN WITH XML INTER-PHASES COMMUNICATION