← Back to team overview

ubuntu-manual team mailing list archive

Re: Not happy at all


On Wed, Jan 12, 2011 at 4:37 PM, lafeber-dumoleyn2
<lafeber-dumoleyn2@xxxxxxxxx> wrote:
> Op 12-01-11 14:44, Ask Hjorth Larsen schreef:
>> On Wed, Jan 12, 2011 at 1:05 PM, lafeber-dumoleyn2
>> <lafeber-dumoleyn2@xxxxxxxxx>  wrote:
>>> Op 12-01-11 11:34, Ask Hjorth Larsen schreef:
>>>> Download the two po-files and run the command:
>>>>   msgmerge lucid-po-file.po maverick-po-file.po>    merged.po
>>>> This will move the translated messages over to the new po-file using
>>>> fuzzy matching.  After this you'll still have to review the fuzzies,
>>>> but this is much easier if you install a po diffing tool such as
>>>> podiff from e.g. pyg3t[1] along with wdiff, which can be used to
>>>> highlight the differences.  After that the po-file can be uploaded and
>>>> completed e.g. on launchpad.
>>>> [1] https://launchpad.net/pyg3t
>>> Hi Ask,
>>> Thank you for your tip. I have done what you described above, and now I
>>> have
>>> a merged file (1661 strings, 400 not ready, 62 untranslated). This may be
>>> the right starting point to continue with the translations of new
>>> versions,
>>> providing that the maverick version is based on the Lucid version + extra
>>> strings. I'm not sure if strings from lucid have been removed in the
>>> maverick version. I will contact the manual guys about this.
>>> I will not upload the merged file unless I am absolutely sure that the
>>> result is right.
>>> Regards,
>>> Hannie
> Hello Ask,
>> It's unimportant whether strings have been removed, the main issue is
>> that the existing ones are transferred.  If msgmerge produces a
>> po-file with more entries in it than it should (I don't know why that
>> sometimes happens), merge the msgmerged file into the maverick po-file
>> again to get something with exactly the correct entries.
> This I will do.
>> I confirm that yes, this *is* the right way to continue with the
>> translations (as was mentioned in the very first reply in this thread,
>> so this should not come as a surprise by now)
> I'm sorry that I haven't done it at that time, I probable didn't quite know
> how to do it.
> One more question. I have installed wdiff, gave the command wdiff fileA
> fileB. This resulted in a lot of messages (about 500 lines) in the terminal.
> I copied and pasted those in my text editor. Is this how it works?
> Regards,
> Hannie
>> , and indeed the exact
>> reason why msgmerge exists.  It's the standard procedure when
>> transferring translations when programmes are updated.

I would msgmerge the lucid file into the maverick file, producing a
merged file.

Then I would wdiff the lucid file with the maverick file, producing
word-wise highlighting of all changes, including those in msgids.  I
would read through all the msgid changes and adapt the corresponding
fuzzies in the merged file accordingly (in a separate window with a
po-editor).  After this, all transferable changes have been
transferred, so at this point the translation can be uploaded to
Launchpad and finished there (or finished off-line).

The wdiff may be more readable if you use podiff from pyg3t, e.g.
`podiff -r oldfile newfile | wdiff -d > differences.txt' to get a
nice, scrollable output.  Since there are quite a few strings I would
probably look for a diffing tool which can highlight with colours,
before starting (let me know what you decide since I haven't done any
of this yet, but will have to at some time).