Author: Philip Hazel Date: To: Ralf Junker CC: pcre-dev Subject: Re: [pcre-dev] Internal errors and crashes with quantified
subroutines
On Wed, 23 Nov 2011, Ralf Junker wrote:
> I fully agree from a functionality point of view. But it could still be
> nice to keep the internals in sync so malloc() kicks in at the same
> number of forward references regardless of LINK_SIZE.
I have just committed a patch that does this. I built in a larger
maximum number of forward references of around 200,000 just as a safety
measure. An incidental fix as part of this is that filling in repeated
forward references is now much faster (it wasn't remembering the
previous one, and was searching for it each time. Doh.)
