Replacing GNU grep revisited

[David Schultz]

On Sat, Jun 21, 2003, Sean Farley wrote:
> In January, Pedro Giffuni started a thread about replacing GNU's grep in
> the system.  Interestingly, I did not know about the grep thread on
> hackers until later.  James Howard had interjected in February with
> mention that he had gotten patches to speed up freegrep.  I had sent him
> those patches in late January, but I have not heard from him since.
> Hopefully, he is OK and just busy.
> 
> I have placed the patches up on Geocities¹ for others to try out.  They
> get freegrep fairly close to the performance of GNU's grep.  Also
> included is a small patch to regex to squeak a bit more performance out
> of it, but I am not certain if it actually helps or not.
> 
> BTW, Postgres is using a newer version of the regex library--swiped from
> TCL--that FreeBSD uses.  It supports multibyte characters.  Is it time
> for an update?

dds@ has expressed some interest in compiling the FSMs for regexps
into native code, which would make it blazingly fast.  See
cvs-all at .  As a practical matter, there are only a couple of
zealots who care what kind of license grep is under, so replacing
GNU grep with something that's ``almost as good as GNU grep'' is a
regression IMO.  If we were talking about a kernel module or
library, of course, that would be a different matter.