OBC - Simple version of C NASM X86 linux Compiler

This project implements a compiler for a simplified version of C with error recovery using python and Lark parsing library

What is the difference between this language and C?

• only has 2 data types, int and void
• there is no preprocessor

it does not support the following C features

• unary operators
• bitwise operators
• ternary operators
• structs
• pointers

for more information about it's syntax read src/grammar.lark

Builtin functions and variables

This language consists of the following builtin functions

declaration	description
void putchar(int c)	writes in the stdout the first byte indicated by the variable c.
void putstr(int str[])	writes in the stdout all the first bytes indicated by the variable str, it must be null-ended.
void putint(int n)	writes in the stdout an integer n.
int getchar(void)	returns single byte read from stdin.
int getint(void)	returns integer read from stdin.
int SIZEOFINT	represents the size in bytes of int variables (4)

Prerequisites

• python >= 3.6 (some source files have literal string interpolation)
• pip
• nasm

Installing

Install all dependecies described in requirements.txt using pip

pip3 install -r requirements.txt

Usage

cd to the src directory

cd src

and call the compiler on some source file

./obc.py [SOURCE FILE PATH]

by default the following files representing the compiler's IO will be created:

• tokens.dot,tokens.pdf, representing the lexer output
• syntax_tree.dot,syntax_tree.pdf, representing the syntax analyzer output
• symtable.dot,symtable.pdf,semantic_tree.dot,semantic_tree.pdf, representing the semantic analyzer output
• tac.txt, representing the intermediate code generator output
• code.asm, representing the code generator output
• object.o, representing the asembler output (NASM)
• [program_name], representing the linker output (ld)

for more options use /obc.py --help