Re: [Help-nano] man-html translated in RO

[Mihai Cristescu]

Dear Mario,

Thanks for your reply and very good material you sent me.

Well, I was already finished translating the man files, after I have created a special filter based on regular expresions in memoq for importing groff files. 
I was almost finished with the reviewing for errors. I also have generated the html files with groff-utf8 command.

But, now after receiving your scripts I made some tests.
Actually they work like a charm. I wish I had them before, they create much better quality of pot file (po4a seems a good tool for the future).

So, I have generated the ro.po file (combined from all text sources) and imported successfully in my CAT system, that means I can use the translation memory already done for this project.

I have installed also Lokalize, Gtranslator and Poedit to have an idea and they all are good but I will keep working in memoq because I have more experience in it for complex tasks.
Maybe for the future I'll start using Lokalize seems to have more features inside.

Question: After I finish translating the file, I send the material to this list or to the http://translationproject.org for revision?

After I finish nano I can have a look also to help2man, but that is already generated pot file isn't it? ... no po4a involved.

And is with this robot thing, for newr versions of the sources, how is that working?

Thanks again!

Mihai

2014-07-26 23:44 GMT+03:00 Mario Blättermann <address@hidden>:

Hi Mihai and Benno,

Am 25.07.2014 11:47, schrieb Benno Schulenberg:

>
> Hi Mihai,
>
> (For Mario: Mihai made Romanian translations of the nano docs, but
> of the HTML pages.  I'm wondering, how are you getting along with
> applying po4a to nano?)
>

Nice to know! New doc translators are always welcome.

> On Fri, Jul 25, 2014, at 11:25, Mihai Cristescu wrote (quoted in full):
>> Thanks for reply and sorry that the files can't be used in this form.
>> I know that the html page are generated from the man groff format, but the
>> thing is I am using a CAT system (memoQ) which uses a translation memory
>> and speed up the translation process significantly.
>
> Ah, very good.  Then, when the nano docs become available in
> POT files, this translation memory will save you a lot of work.
>

Off-line translation tools like Lokalize, Poedit or Gtranslator also have a
translation memory. But as long as your special tools can handle normal po files
(which I strongly assume) then it shouldn't be a problem to keep your software
which your are familiar with.

>> The problem is that I can't import the groff files into memoQ.
>>
>> I have tried several approaches to convert the man groff files to: rtf ,
>> ps, html formats which can be imported in memoQ, omegat , Trados SDL etc,
>> but the issue is that there is no way to convert the translated files to
>> man groff back. It seems that the only productive way to translate the man
>> pages is to translate the files directly, but this is time consuming since
>> you are not using a translation memory and term base.
>>
>> Regarding po4a :
>> I have installed yesterday this package on my arch system and followed the
>> documentation. I succeeded converting the man pages to pot files, import
>> them in memoQ, translated, but I was not able to convert back to man page
>> (I had some strange errors on po4a-translate command) and I had to give up.
>> Maybe I didn't quite understand how the po4a system works.
>
> Mario has some experience with po4a.  Maybe you can work together
> on solving the issues?
>

First, forget about the too difficult process described in the po4a docs, with
the configuration file which allows to accept non-UTF8 files, add extended
translator acknowledgements and some other things which we don't really need.
Just generate the pot file and localized docs with the attached scripts.

Put both scripts in the doc directory. With create-docs-pot.sh you generate then
a merged pot template for all groff and texinfo sources.

Then use the pot file to create your po file as usual, put the file ro.po (note
the file name, no other naming scheme will be accepted) into the doc directory
and launch translate-docs.sh. The translated groff files will be created in
subdirectories, named according to the locale. The texinfo file gets a suffix
(ro in your case, because there's currently no default way to install localized
texinfo files on a system).

Note, per default po4a needs at least 80% of the po strings translated for each
source file, otherwise the creation of that file will be skipped. For testing
purposes you can screw down ${threshold}, currently it is set to zero. But in my
mind, users won't benefit of translations lower than 80%, who wants to read a
documentation with more holes than real info in his/her native language...?

>> Nevertheless I shall contact Laurentiu to have a chat with him about the
>> process.
>>

Would be nice in any case to complete (or at least almost complete) the UI
translation first. Besides that, I don't see any other examples of Romanian man
pages, so it would be a good idea to translate the 49 messages from help2man [1]
first and let the terms confirm by a Romanian reviewer. Would be a good base for
future man page translations.

Moreover, there are already French plain text translations of the manpages. I
will try to import those into po files, once someone with Autotools skills has
setup the installation chain and the TP domain has been initialized.

[1] http://translationproject.org/domain/help2man.html

Best Regards,
Mario