vidur%netscape.com 599c0d0fe9 Fix for bug 81253. We now sniff upto the first 2k of the first buffer of any HTML stream, looking for a META tag with charset information. If charset information is found, we use it for unicode conversion. This deals with the bulk of cases where we used to do a reload based on charset information in the document. In the worst case (if charset information exists but isn't found during sniffing) we fall back to the reload case. This fix improves initial page load performance for pages with a charset. Degradation in performance for pages loaded out of the cache is still being investigated. r=harishd, sr=waterson ...

git-svn-id: svn://10.0.0.236/trunk@98320 18797224-902f-48f8-a5cc-f745e15eee43

2001-06-29 22:56:58 +00:00

15 KiB

Raw Blame History

View Raw