[wp-polyglots] Issue with curly quotes.

[Wacław Jacek]

Xavier Borderie pisze:
> I'd much rather have a complete and correct translation, and have WP
> do all the tedious regexp work, rather than leave these fields
> untranslated and have to work out the regexp-magic to make it happen
> :)
> To me this is a WP bug, and as such it should be corrected.
> 
> -x.
> 
> 2009/8/13 Wacław Jacek <mail at waclawjacek.com>:
>> Xavier Borderie pisze:
>>> Since 2.8, we translators can adapt the curly quotes to the local
>>> usage. This is great, but there seems to be an issue in some case.
>>>
>>> In fr_FR, we applied the necessary entities in order to respect French
>>> typographical rules: there has to be a space between the quote word
>>> and the quote.
>>>
>>> #. translators: opening curly quote
>>> #: wp-includes/formatting.php:37
>>> msgctxt "opening curly quote"
>>> msgid "&#8220;"
>>> msgstr "&laquo;&nbsp;"
>>>
>>> #. translators: closing curly quote
>>> #: wp-includes/formatting.php:39
>>> msgctxt "closing curly quote"
>>> msgid "&#8221;"
>>> msgstr "&nbsp;&raquo;"
>>>
>>> It works great... except when the quoted section combines these two facts:
>>>  - it also serves entirely as (or ends with) a link ;
>>>  - it is immediately followed by punctuation.
>>>
>>> Here is a sample post on my trunk test-blog (2.9-rare, with latest 2.8
>>> PO/MO).
>>>
>>> I'm guessing I should file a ticket in Trac for this bug, but maybe
>>> it's not actually a bug, so if anyone here knows better...
>>>
>> This is just a wild proposition and I doubt it will be implemented, but one
>> idea is to keep the regular quotation marks (' " ') in the translation files
>> and then have a separate file parse them (i.e. eg. make them into curly
>> quotes using a regexp).
>>
>> W. J.
>> _______________________________________________
>> wp-polyglots mailing list
>> wp-polyglots at lists.automattic.com
>> http://lists.automattic.com/mailman/listinfo/wp-polyglots
>>
> 
> 
> 
The point is to avoid the use of &#8220; and &#8221;, so translators 
could simply use the " symbol instead, and a script would parse that to 
a "start" and "end" string of their choice, i.e. "some text" would be 
replaced with [starting " replacement]some text[ending " replacement], 
but that could just be me wanting to automate everything possible. It 
might be worth considering, though.

W. J.