Re: tin (current 1.9.3 snapshot) hangs with pcre problem

On Mon, Oct 15, 2007 at 02:41:48PM +0200, Olaf Dietrich wrote:
> > no it just gives the error message (= waitz for two seconds) for _each_
> > (depending on filter type, all inc. read, or just unread) subject in that
> > group -> that may take a while.
> 
> Nevertheless I wonder, whether this situation could be handled somewhat
> more user-friendly - there are currently 6000 unread articles in the
> group, so there is little chance to recover from the error messages.

the current code is something like:

filter_articles() loops over all articles (in the given scope) and tries to
apply the rule via test_regex() which looks if the rule has already been
compiled (in that case it would use the compiled result for speed reasons)
and if not tries to do so via compile_regex(). compile_regex() (which calls
pcre_compile() which does not have an error returncode except storing an
error message somewhere) fails, gives an error messag and returns with FALSE
which just means no match. outside compile_regex() we do not know anything
about the regexp except that it didn't match and we do not have a build
cache (which is also true for the first article in the loop).

feel free to improve that (but avoid pcre_compile2()).

urs
-- 
"Only whimps use tape backup: _real_ men just upload their important stuff
 on ftp, and let the rest of the world mirror it ;)" - Linus