|Lex and Yacc (or Flex and Bison) for Win95??? email@example.com (Akintunde Omitowoju) (1997-02-16)|
|Re: Lex and Yacc (or Flex and Bison) for Win95??? WStreett@shell.monmouth.com (1997-02-20)|
|Re: Lex and Yacc (or Flex and Bison) for Win95??? firstname.lastname@example.org (Schauer Werner) (1997-02-20)|
|Re: Lex and Yacc (or Flex and Bison) for Win95??? a-bnc@MICROSOFT.com (Nicholas Carey (BEST)) (1997-02-22)|
|Re: Lex and Yacc (or Flex and Bison) for Win95??? email@example.com (The Flanagans) (1997-03-01)|
|Re: Lex and Yacc (or Flex and Bison) for Win95??? firstname.lastname@example.org (1997-03-16)|
|From:||The Flanagans <email@example.com>|
|Date:||1 Mar 1997 21:45:07 -0500|
|Keywords:||flex, yacc, Windows|
Akintunde Omitowoju wrote:
> Hello everyone:
> I am creating a subset of a C++ grammar for my thesis work, but I am not
> sure whether the code generated by Flex and Bison for the DOS version of
> the GNU compiler will work.
If Bison is truly yacc compatible, then any recent version will
have source and generated code that is ANSI C compliant. One problem I
ran into ( other than Unix vs. DOS file paths and names ) is the size
of a grammar can exceed certain memory limits of a 16-bit DOS
port. Specifically, I was working with an SQL grammar that had over
1,000 lookahead sets which exceeded the maximum size of the array that
handles that ( 64k segment barrier ) in yacc.
With almost no effort at all ( as another reply points out ), I
made 32-bit OS/2 and Windows NT versions of yacc with the greater
limits, processed the grammar... ( cut me off if I'm going on ) and
then compiled the generated automation under 16-bit DOS without any
problems ( I did have to use one compiler flag to put the largest
tables in their own data segments to get under 64k data ).
Bottom line: "it should be painless".
Return to the
Search the comp.compilers archives again.