[antlr-interest] Re: Replacing a operator precedence parser

[dan.stanger at ieee.org]

I could generate c++ code, and link it into lisp, or I could write
a back end to produce lisp code.  I think that is not the issue.
The curent parser is not modular, and is very tightly coupled to
the lexer and input streams.  However the interface to the math
engine is very simple.
I am more interested in how to build the grammar and lexer on a higher
level, and how I would design the grammar.  I assumed that I would
use predicates to identify operators, I expected the whole thing to
be built using tables.
Dan Stanger

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/