Re: Current status on migration

["Ulf-D. Ehlert" <ude88@xxxxxx>]

Hi Julien,

julien (Freitag, 5. September 2008, 20:48):
> I ran
> xml2po-support> python tools/split-xml-multi-lang.py --debug -l fr -f
> oldsrc/menus/file-new.xml -d DEST
>
> I got two directories, 'en' and 'fr' with file-new.xml in both. OK
>
> I noticed in the 'fr' file:
> - in <note> and <sect>, all language attributes are shown
> (lang=en;de;es...) even though there are fr para only.

Yes, 'xml2po' won't care about the "lang" attribute, so I didn't remove 
or replace them. And they may be useful for finding bugs:

    <para>English text</para>

may be ok in a French XML file, but e.g.

    <para lang="de;fr">English text</para>

would obviously indicate a serious bug in our script.

> - The tip is in english because it is not translated to 'fr'.

Yes, that's intended. We need two XML files of the very same structure 
to use xml2po's "reuse" option. Try

    xml2po --output=french.po --reuse=french.xml --language=fr
    english.xml

xml2po should replace any "translation" which is the same as the English 
source string with an empty string. (Check this!)

> - In the grayscale section, there are 4 'en' para  and 2 'fr' para in
> the source. In the destination 'fr' file, "missing" para are in
> english.

Hey, nobody is perfect. ;-)
What should the script do? Are the the missing 'fr' (and 'de') paras 
intentionally not translated? The 'en' paras may be added just 
yesterday, how can the script know?

But maybe it's a good idea to detect these cases and emit a warning.

> What is the use of the 'seqnum=' added?

They are internally used to mark a node as "seen". The current script 
version has been changed to remove these attributes when constructing 
the destination document.

Ulf
_______________________________________________
Gimp-docs mailing list
Gimp-docs@xxxxxxxxxxxxxxxxxxxxxx
https://lists.XCF.Berkeley.EDU/mailman/listinfo/gimp-docs