[whatwg] Charset sniffing from XML prolog
Boris Zbarsky
bzbarsky at MIT.EDU
Wed Oct 7 19:10:33 PDT 2009
On 10/7/09 9:52 PM, Kartikaya Gupta wrote:
> Anything else that might be affecting this?
In general, yes. Charset info can come from the HTTP cache, from user
bookmarks, etc, etc.
In this case, though, it's totally my fault: I just forgot that I had
the HTML5 parser turned on locally. Turning that off, I do get UTF-8,
because of
http://hg.mozilla.org/mozilla-central/file/603759afc77a/parser/htmlparser/src/nsParser.cpp#l2553
and following. That code is just bogus, in my somewhat biased
opinion.... ;)
-Boris
More information about the whatwg
mailing list