Details
-
Bug
-
Resolution: Done
-
P3: Somewhat important
-
4.7.0
-
None
Description
Commit c5bfbe27 (from Fri Oct 16 2009) contains a regular expression of yours that appears to make qdoc take about 8 times as long to run than it did before. Maybe this is due to the complexity of the expression, or perhaps it's due to a regression in the regexp engine. In either case, I'd appreciate it if you can take a look at it and maybe simplify it, otherwise I'll have to hack at it.
Warwick wrote:
> The problem, I strongly suspect is that the "[^&]" is too greedy. It should be "[^&\n]". The "[&\\\\]" should also be "[\n&\\\\]". This should prevent the regexp from ever looking beyond linebreaks.
>
> The principle is "find the line that contains a single-line-comment NOT in a string, then find the single-line-comment in that" (previously, it found single-line comments in strings, which screwed up URLs; I don't recall the messes it made, but it was a problem - revert the patch as diff the resulting HTML to see the problem).