Migrating Pology to Python 3

[Karl Ove Hufthammer]

I think keeping the original order is best. It results in cleaner diffs.

Sorting the tags may of course *sometimes* be useful, but that can 
always go into a separate sieve, just like we have for ‘normalize_header’.


Adrian Chaves skreiv 08.10.2022 13:28:
>
> What should we do here? Keep the original order, or enforce some 
> specific sorting? (alphabetical?)
>
> On 2022-10-08 12:50, Karl Ove Hufthammer wrote:
>
>> Johnny Jazeix skreiv 05.10.2022 22:14:
>>> It causes multiple changes in the formatting of the po:
>>> -#, kde-kuit-format, no-c-format
>>> +#, no-c-format, kde-kuit-format
>>
>> This is actually a bug (I think) in *both* the Python 2 and Python 3 
>> versions. They change the order of the tags, but in different ways. 
>> Example .po file with two entries:
>>
>> #, no-c-format, kde-format
>> msgid "Search..."
>> msgstr "Søk …"
>>
>> #, kde-format, no-c-format
>> msgid "Find..."
>> msgstr "Finn …"
>>
>> If I run the Python 2 porewrap script, I get:
>>
>> #, kde-format, no-c-format
>> msgid "Search..."
>> msgstr "Søk …"
>>
>> #, kde-format, no-c-format
>> msgid "Find..."
>> msgstr "Finn …"
>>
>> If I run the Pyhon 3 porewrap script, I get:
>>
>> #, no-c-format, kde-format
>> msgid "Search..."
>> msgstr "Søk …"
>>
>> #, no-c-format, kde-format
>> msgid "Find..."
>> msgstr "Finn …"
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-i18n-doc/attachments/20221008/69ca4878/attachment.htm>