Skip to content

tsucres/ImpCompiler

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ImpCompiler

Javadoc

This is a compiler written in Java for a simple imperative language (referred as "Imp"). It supports all the basic features of a programming language, including:

  • integer variables
  • string constants
  • arithmetic expressions
  • boolean expressions
  • basic flow control instructions: if/else, while, for
  • functions
  • print/read statements

Its full grammar is presented in the grammar.md file. You can check the test folder for some example source codes.

The project is composed of 4 main parts:

  • The lexical analyzer, built with jFlex, outputting a sequence of lexemes and a list of detected identifiers
  • The parser, which builds a parser tree from the sequence of lexemes
  • The semantic analyzer, which basically just builds an AST from the parser tree and runs some basic verifications concerning the semantic of the input program
  • The code generator: generate and outputs an LLVM-IR code equivalent to the input program.

The produced LLVM code can then be ran using LLVM's lli command.

Build & Run

Build

To build the project, run

make 

You can also import it in your favorite IDE.

The previous command will compile the project in the bin directory, produce a jar in the dist folder and output the javadoc in the docs folder.

Produce jar

make jar

The jar will be created in the dist folder

Produce javadoc

The following will produce the javadoc in the docs folder

make doc

Run

$ java -jar dist/ImpCompiler.jar 
	
	Usage:  java -jar ImpCompiler.jar file.imp [out.ll] [--debug] [--exec]
	or	java Main file.imp  [out.ll] [--debug] [--exec]

	--debug		 prints the ParserTree and the AST
	--exec		 executes the resulting LLVM code (lli must be an accessible command)

Project structure

  • dist: contains the .jar
  • docs: contains the javadoc (available here)
  • test: contains a bunch of imp source codes (valid and syntactically/semantically invalid) used to test the compiler. They're sorted according to the functionalities of the Imp language they implement and to whether they are valid or not.
  • src:
    • The Lexer package implements the lexical analyzer. It's mainly composed of the model classes Symbol and LexicalUnit and of the jflex-generated class ImpLexer
    • The Parser package implements... well, the parser. It contains everything needed to build the parser tree and a tree of identifier tables (representing the variable scoping)
    • The AST package is made of a single class used to simplify the parser tree into an Abstract Syntax Tree
    • The LLVM package is basically a java binding for LLVM that handles the needed functionalities. It's independent to the Imp language, expect for the LLVMGenerator class which produce the LLVM-IR code from the (Imp) AST.

About

Compiler for a simple imperative language (called"Imp"). Uses jFlex for its lexical analyser. Produces LLVM-IR code.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published