Having been through the experience of using bison to parse a few COBOL verbs for the preprocessor, I think it would be good to have a more powerful parser generator. In particular I had trouble with the identifier construct, having to write lots of ugly buggy C code to hack the token values so bison could cope. I know Laura etc have been doing some research on the parser generators out there. I had a play with btyacc and apart from the sparse documentation it looks quite good. Would you mind sharing the research you have done, with us? What PGs should we consider/avoid and why? Having completed the preprocessor I am keen to start on the compiler proper. What is the situation at present? Are we still waiting on the parser generator research? Can anything else be done in the meantime? Tim Josling
- To: gnu-cobol@lusars.net
- Subject: Parser generators for COBOL
- From: Tim Josling <tej@melbpc.org.au>
- Date: Mon, 13 Sep 1999 19:01:32 +1000
- Organization: Melbourne PC User Group
Having been through the experience of using bison to parse a few COBOL verbs for the preprocessor, I think it would be good to have a more powerful parser generator. In particular I had trouble with the identifier construct, having to write lots of ugly buggy C code to hack the token values so bison could cope. I know Laura etc have been doing some research on the parser generators out there. I had a play with btyacc and apart from the sparse documentation it looks quite good. Would you mind sharing the research you have done, with us? What PGs should we consider/avoid and why? Having completed the preprocessor I am keen to start on the compiler proper. What is the situation at present? Are we still waiting on the parser generator research? Can anything else be done in the meantime? Tim Josling