From: Darren S. <li...@yo...> - 2005-07-21 22:45:07
|
There are some README files on the website which aren't encoded using ISO8859-1 and, consequently, are brokenly rendered by the common or garden web browser. I've just converted the test website to use UTF-8 and to convert these files on the fly, the source encoding being determined according to filename. Hopefully, these encodings are correct: cs: ISO8859-2 pl: ISO8859-2 uk: KOI8-U everything else: ISO8859-1 (There's also a test which is applied to determine whether the conversion is required. I'd rather run a very small risk of incorrectly not recoding some text than outputting doubly-encoded UTF-8...) Ideally, the documentation would already be encoded using UTF-8. and once a group of documents on the web site is fully UTF-8, the corresponding keywords file entry should be altered to reflect this ("charset=utf8" in the flags will do it). "charset=guess" invokes the above encoding determination heuristic. (In the absence of a charset flag, the default is ISO8859-1.) -- | Darren Salt | nr. Ashington, | d youmustbejoking,demon,co,uk | Debian, | Northumberland | s zap,tartarus,org | RISC OS | Toon Army | @ Say NO to UK ID cards | http://www.no2id.net/ See that abacus, that's your hard drive, that is. |