C parser (C parser written in nML)

Sunae Seo
ROPAS
Computer Science
Korea Advanced Institute of Science and Technology

Courtesy
This program is based on ckit

Introduction
This package contains a parser for C programming language implemented in nlex, nyacc and nmlc. The executable of this package gets C source code (after preprocessing: this parser ignores lines starting "#") and produces parse tree represented as a set of nML data types. It is based on ckit, C front end written in SML. In ckit, they implemented type checker, and using the type information, ckit produces abstract syntax tree of the C source code. But we didn't implement type checker and abstract syntax tree generation part. So, in our implementation, the result of program execution is a parse tree of C source code, which is simpler and has less information than abstract syntax tree.

Implementation
After unpacking the source, you can see source code files and README file. In the code files, there are parser related rule files(lexer.nl, parser.ny), parse tree definition file(parseTree.n), files for printing and so on. Detailed information of all files is in README file. There are two kinds of printing options.

-pp : print the parse tree of the given C program
-pl : print the lex tokens till the current parsing position.

Restrictions

• Macro lines are ignored as if they were comments.
• This parser recoginizes types defined by "typedef" keyword. (but types defined other files cannot be recognized as a proper type)
• The ranges of proper values of each literal(char, int, float) in C are not the same as those in nML. So there might be overflow or underflow problems. If you don't want to lose information from the difference of two language literals, please treat them as string(that is, modify parts of the code).

Download
  cparser.tar.gz