Summary: | JS parse errors on microsoft.com (UTF-8) | ||
---|---|---|---|
Product: | [Applications] konqueror | Reporter: | George Staikos <staikos> |
Component: | kjs | Assignee: | Konqueror Developers <konq-bugs> |
Status: | RESOLVED FIXED | ||
Severity: | normal | ||
Priority: | NOR | ||
Version: | SVN | ||
Target Milestone: | --- | ||
Platform: | Compiled Sources | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: |
Description
George Staikos
2003-11-27 03:24:52 UTC
The msdnsearchcontrol.js starts with a hexadecimal efbbbf sequence. A marker for UTF-8. Which library should do the decoding ? UNIX doesn't use the Byte Order Mark encoded in UTF-8 to identify UTF-8 text files. That's a Microsoft invention. The charset should be provided in the HTTP header and it's not: kio_http: (3197) "Content-Type: application/x-javascript" That is, unless the standard says JS scripts are supposed to be encoded in UTF-8. *** This bug has been marked as a duplicate of 63498 *** its not really a duplicate Subject: kdelibs/khtml [POSSIBLY UNSAFE] CVS commit by mueller: * misc/loader.cpp (CachedCSSStyleSheet/CachedScript::data): add some utf8/utf16 encoding handling. Workaround Qt bugs (#69115). Initial investigation and patch by Harri. CCMAIL: 69115-done@bugs.kde.org M +5 -0 ChangeLog 1.125 M +45 -28 misc/loader.cpp 1.160 [POSSIBLY UNSAFE: qDebug] M +4 -2 misc/loader.h 1.63 |