- From: Bjoern Hoehrmann <derhoermi@gmx.net>
- Date: Wed, 09 Apr 2008 04:10:16 +0200
- To: olivier Thereaux <ot@w3.org>
* olivier Thereaux wrote: > In a qa-dev thread about profiling, Ville found a few tricks to make > HTML::Encoding much faster on some pathological cases. The thread > includes patches. Would you look into them? > > http://lists.w3.org/Archives/Public/public-qa-dev/2008Apr/0007.html Please feel free to forward this response to the list or archive it on www-archive. I had a brief look at the patches and I am afraid I can't use either. HTML::HeadParser seems to fail for HTML4 documents with an object element in <head>, and it might change the behavior of the module when multiple Content-Type meta elements are specified (though I didn't check carefully). Looking just for the body start tag is likely the best solution here, but the performance impact rather low, so I've not yet done this. This problem has always been noted in the source though. The decode-last-chunk-only.patch is incorrect, the purpose of doing it this way is precisely to handle sequences that overlap two chunks in the right way, forgetting the previous chunk would break that. It should also be ineffective, since $data is modified by Encode::decode, but it seems the wrapper broke that. I've fixed that bug instead. A new version of the module is on its way to CPAN. -- Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de Weinh. Str. 22 · Telefon: +49(0)621/4309674 · http://www.bjoernsworld.de 68309 Mannheim · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/
Received on Wednesday, 9 April 2008 18:33:19 UTC