flex scanner too huge, suggestions?

Hi all,


I have a working flex/bison implementation to parse a text
protocol. The problem is that the scanner is huge.


Made with LFLAGS = -Cfe, size gives
42068(.text) + 24(.data) + 60(.bss) + 1687886(.rodata) = 1730038


Made with LFLAGS = -Cem, size gives
42496(.text) + 24(.data) + 60(.bss) + 508326(.rodata) = 550906


This is still pretty large for some of our target environments.


I haven't tried to disable features to determine what makes
it so huge, but the protocol is case-insensitive and context
(position) sensitive. I addressed the context sensitivity
with flex start states.


My flex file has the following %option line.
%option noyywrap 8bit batch case-insensitive perf-report


I'm looking for suggestions for either tweaking my flex
implementation or switching to another scanner generator.


The corresponding bison parser is solid, so I'm not really
looking for suggestions to replace the pair. (We initially
tried several other tools for the total parsing problem but
they all choked on this large, context sensitive protocol.)


Thanks,
-troy

Return to the comp.compilers page.
Search the comp.compilers archives again.