call for fanatics

[Jim Blandy <jimb@red-bean.com>]

> The cheap answer is that my heavily optimized extract.scm is still >
> 5x slower than the perl version & is spending the bulk of its time in
> regexp crap.  I had to throw out regexp-substitute/global because it
> was too slow to get this close.  Not to mention that straight regexp
> stuff is 5x slower in guile than in perl, & >10x slower than in awk.
> 
> Or did you mean that you wanted to whether or not this was where the
> regexp time was being lost...

Um, yeah, I wanted to know what to fix.  :)

Dumb question: you're not recompiling the regexps each time, right?
You've hoisted the call to make-regexp out of the loop, haven't you?

If you can narrow down the speed difference to something specific ---
the I/O calls, or the regexp engine, or the garbage collection,
etc. --- then we can start thinking about how to work on this.
Thing is, I don't really have time for this myself.

Basically, I need someone to appoint themselves Guile's Text Czar.  Or
the Guile Textpatriot.  Or whatever.  Maybe a group.  Someone who can
take some real application (the extractor is a perfect test bed),
figure out what fundamental changes we need to bring it up to speed on
Guile, and make those changes.  Besides the obvious rules (the rest of
Guile has to continue to work, you gotta support the R5RS and SCSH I/O
primitives, it's gotta be somewhat generally helpful, not just hacks
for that one app), it's open season.