For non-latin languages it's problem to use URL of MediaWiki pages. For example, article with name "Список языков с алфавитами на основе кириллицы" (List of languages using Cyrillic alphabet) has ugly URL:
http://ru.wikipedia.org/wiki/%D0%A1%D0%BF%D0%B8%D1%81%D0%BE%D0%BA_%D1%8F%D0%...
Opera browser auto decodes this URL to readable link: http://ru.wikipedia.org/wiki/%D0%A1%D0%BF%D0%B8%D1%81%D0%BE%D0%BA_%D1%8F%D0%...
In Firefox it's able to use HumanURL extention, but in IE it's problem.
So can I add option "wgEncodeUrlsInUtf8" to LocalSettings.php and check it in function wfUrlencode in GlobalFunctions.php? (to turn off PHP urlencode function)
See also http://en.wikipedia.org/wiki/Internationalized_Resource_Identifier
you can see demo of the hack at http://mediawiki.jino-net.ru/index.php/Main_page
Discussion in Russian: http://community.livejournal.com/ru_wikipedia/24808.html
-- Alexander Sigachov http://meta.wikimedia.org/wiki/user:ajvol
"áÌÅËÓÁÎÄÒ óÉÇÁÞ£×" alexander.sigachov@gmail.com wrote in message news:3ba413350603100720g5f643ea5p@mail.gmail.com...
For non-latin languages it's problem to use URL of MediaWiki pages.
Is there any particular reason we can't use [[punycode]]?
Phil Boswell wrote:
"áÌÅËÓÁÎÄÒ óÉÇÁÞ£×" alexander.sigachov@gmail.com wrote in message news:3ba413350603100720g5f643ea5p@mail.gmail.com...
For non-latin languages it's problem to use URL of MediaWiki pages.
Is there any particular reason we can't use [[punycode]]?
Yes. Punycode only works in domain names, not URL paths.
-- Neil
áÌÅËÓÁÎÄÒ óÉÇÁÞ£× wrote:
For non-latin languages it's problem to use URL of MediaWiki pages. For example, article with name "óÐÉÓÏË ÑÚÙËÏ× Ó ÁÌÆÁ×ÉÔÁÍÉ ÎÁ ÏÓÎÏ×Å ËÉÒÉÌÌÉÃÙ" (List of languages using Cyrillic alphabet) has ugly URL:
http://ru.wikipedia.org/wiki/%D0%A1%D0%BF%D0%B8%D1%81%D0%BE%D0%BA_%D1%8F%D0%...
Opera browser auto decodes this URL to readable link: http://ru.wikipedia.org/wiki/%C3%B3%C3%90%C3%89%C3%93%C3%8F%C3%8B_%C3%91%C3%...
Ironically, in my Mozilla Thunderbird, the "ugly" URL works and even shown the proper Cyrillic title in the status bar, whereas the "readable" version looks like aplhabet soup and takes me to a "no such article" page. I can get it to work if I manually view the message with a koi8-r encoding, but Thunderbird refuses to display it as such by default even though I've unchecked the "always use my encoding" box.
Александр Сигачёв wrote:
For non-latin languages it's problem to use URL of MediaWiki pages. For example, article with name "Список языков с алфавитами на основе кириллицы" (List of languages using Cyrillic alphabet) has ugly URL:
We use IRIs (Internationalized Resource Identifiers) in the URI encoding, per spec.
Unencoded URLs aren't reliable, as various browsers treat things in various exciting ways.
-- brion vibber (brion @ pobox.com)
2006/3/10, Brion Vibber brion@pobox.com:
Unencoded URLs aren't reliable, as various browsers treat things in various exciting ways.
If URL is in UTF8 then IE5, IE6, Opera and Firefox works fine. And, I think, it's useful at least for non-Wikimedia projects in cyrillic languages.
-- Alexander Sigachov
On 3/11/06, Александр Сигачёв alexander.sigachov@gmail.com wrote:
2006/3/10, Brion Vibber brion@pobox.com:
Unencoded URLs aren't reliable, as various browsers treat things in various exciting ways.
If URL is in UTF8 then IE5, IE6, Opera and Firefox works fine. And, I think, it's useful at least for non-Wikimedia projects in cyrillic languages.
To elaborate on Brion's comment: - I believe the "Always send URLs as UTF-8" option is off by default in IE5, and on in IE6. - In Firefox, the default was flipped as recently as a year ago: https://bugzilla.mozilla.org/show_bug.cgi?id=261929
wikitech-l@lists.wikimedia.org