Horton

Re: Another regexp performance improvement: skip useless paren-captures

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andrew Dunstan <andrew@dunslane.net>
Cc: pgsql-hackers@lists.postgresql.org, Joel Jacobson <joel@compiler.org>
Date: 2021-08-05T14:36:21Z
Lists: pgsql-hackers 
Andrew Dunstan <andrew@dunslane.net> writes:
> I'm a bit worried about how you'll keep track of back-ref numbering
> since back-refs only count capturing groups, and you're silently turning
> a capturing group into a non-capturing group.

They're already numbered at this point, and we aren't changing the numbers
of the capturing groups that remain live.  There will be unused entries in
the regmatch_t array at runtime (corresponding to the zapped groups), but
that doesn't cost anything worth mentioning.

Now that you mention it, I am not sure whether there are any regression
test cases that specifically cover still being able to match \2 when
the first capture group went away.  Probably should add more cases...

			regards, tom lane

Commits

  1. Fix regexp misbehavior with capturing parens inside "{0}".

  2. Let regexp_replace() make use of REG_NOSUB when feasible.

  3. Fix bogus assertion in BootstrapModeMain().

  4. Avoid determining regexp subexpression matches, when possible.

  5. Check the size in COPY_POINTER_FIELD

  6. Make regexp engine's backref-related compilation state more bulletproof.