Bugzilla – Bug 5236
Strange behaviour with _some_ non-latin characters
Last modified: 2007-10-26 13:35:27 UTC
I'm seeing strings cut off at some characters when sending through the web server. A Hungarian user reported this issue for my StringEditor, but I was able to reproduce it with the search form as well: sending "Főoldal" doesn't return a valid string, but will be cut off after the F. The string's representation is fine until some unicode magic is applied (S::W::HTTP::processHTTP, line 555: call to Slim::Utils::Unicode::utf8encode_locale). After that the string is crippled. After disabling that line, it returned just fine. But it's way beyond my scope to assume that's a valid fix :-) Although marked as fixed, bug 5196 (https://bugs-archive.lyrion.org/show_bug.cgi?id=5196) could be related to this issue.
Index: /Users/mh/Documents/workspace/slimpy/Slim/Web/HTTP.pm =================================================================== --- /Users/mh/Documents/workspace/slimpy/Slim/Web/HTTP.pm (revision 12557) +++ /Users/mh/Documents/workspace/slimpy/Slim/Web/HTTP.pm (working copy) @@ -551,7 +551,7 @@ if ($value ne '*' && $value ne '') { $value = Slim::Utils::Unicode::utf8on($value); - $value = Slim::Utils::Unicode::utf8encode_locale($value); +# $value = Slim::Utils::Unicode::utf8encode_locale($value); } Disabling that one line did the trick for Norbert (Hungarian translator). And I didn't see a drawback, yet. What is it good for?...
I'm not sure who would know. Do you think it would be too risky to commit and put it in the nightlies?
Change 13097 - let's give this some exposure...
Apparently fixed. Please reopen if it still occurs.