https://bugs.exim.org/show_bug.cgi?id=1916
--- Comment #15 from Philip Hazel <p...@hermes.cam.ac.uk> ---
(In reply to Markus Elfring from comment #14)
> (In reply to Philip Hazel from comment #13)
>
> Thanks for another constructive feedback.
>
> * How are the chances to change the mentioned software inefficiency in
> significant ways?
>
> * Can your current regular expression engine perform cost analysis so that
> it could determine at which number of âkey wordsâ (or âbranchesâ) in
> the
> alternation it would make sense to work also with a corresponding trie
> variant which would be generated on demand as a special service?
>
> * Will advanced indexing approaches become feasible here?
Look, all this is very interesting, but PCRE is designed as a general regex
engine. I have no intention of trying to add anything special for matching
lists of fixed strings because the right thing to do is to write specialist
code for such applications.
--
You are receiving this mail because:
You are on the CC list for the bug.
--
## List details at https://lists.exim.org/mailman/listinfo/pcre-dev
