Re: [wgo] i18n strategy and progress



Hi Jens,

--On Freitag, Juni 12, 2009 09:08:21 +0000 "Jens W. Klein"
<jens bluedynamics com> wrote:

just a small update:

We proposed a solution [1] which is simple enough for both sides:
translators get just a GIT with POT/PO files, at Plone side we're using
LinguaPlone as usal and need to add a thin automation-layer, also we need
to add a small webservice for pulling/pushing translations.

[1] http://live.gnome.org/GnomeWeb/Plone/Translation

The posting at gnome-18n [2] list didnt received the amount of response
I'd expect. Translators itself arent interested very much - which I can
understand. The minority of technical people didnt understand the problem
well. Anyway, there wasnt any good result.

[2] http://article.gmane.org/
gmane.comp.gnome.internationalization.general/21913

I also think it's got stuck a bit.
There's the problem Shaun has with escaped html.
IMHO it can be solved inside our handler (e.g. with a small transformation
that uses beautifulsoup or lxml.html) or in the xml2po mode. It is probably
better as we can check if the field contains html or plain text. We can
give the tag and attribute 'xml="1"' then. We should also strip out all
"uninteresting" content like wf states etc. and reconstruct it on upload.
The xml2po mode then doesn't have to know about the fields for different
content types.

I think that's largely what xml2po needs, but we only come to a result if
we get a consensus with a guy that can write the xml2po mode, that will
probably be Shaun.
I know that all the requirement pinning was done ages ago by people not
present anymore - with all the problems it brings IMHO - but one is that
the CMS has to provide xml, and a further integration into gnomes
translation infrastructure (git/svn, damned lies, xml2po) is simply out of
scope.

Does anyone have an ideas who else can do to the connection from plone
(provides and consumes xml) to git (scripting, xml2po)? Wo wrote it for
docbook translations?

I will get back and follow up on the i18n list later or tomorrow continue
discussion.

What I can do now is:
1) to start writing the neccessary code for Plone.
2) Second task is to write the code for pushing/pulling the code.
automatically.
3) Third is to write some kind of xml2po-extension.

I can do (1)+(2). For (3) I really need help or better a person familar
with xml2po and a person familar with GIT.

Yes, I'd start with the plan we have now and solve the xml2po issue in the
next days. There will be no changes to the content sync/web service plan.

Tschö
..Carsten


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]