In message <[email protected]>, John Nagle wrote:
> ("It should be written in C" is not an acceptable answer.)
I don't see why not. State machines that have to process input byte by byte
are well known to be impossible to implement efficiently in high-level
languages. That's why lex/flex isn't worth using. Even GCC has been moving
to hand-coded parsers, first the lexical analyzer, and more recently even
for the syntax parser (getting rid of yacc/bison), certainly for C.
--
http://mail.python.org/mailman/listinfo/python-list
