HashAndUTF8CharactersTranslator::translate() calls convertUTF8ToUTF16() to turn the UTF8 source into character data. It should use the optional "is all ASCII" flag parameter to convertUTF8ToUTF16() and create an 8 bit string if the source is all ASCII.
Created attachment 172880 [details] Patch
Comment on attachment 172880 [details] Patch Changes like this are better if the change log or comments state why we are doing this. I know why, but I don’t think it’s obvious why this is the right thing to do. You can imagine a different approach where we actually allocated a UTF-16 buffer and scanned it and determined the whole thing was Latin-1 even for non-ASCII cases. Presumably that’s not important, but this optimization is important. Some brief remark about how you decided to do this is needed. The change log you have here is entirely mechanical, and it says what the code does, but what people are interested in more often is why.
Comment on attachment 172880 [details] Patch Clearing flags on attachment: 172880 Committed r133918: <http://trac.webkit.org/changeset/133918>
All reviewed patches have been landed. Closing bug.