Re: Foreign Languages / interface translation

["Giuseppe Bilotta (Oblomov)" <oblomov@freemail.it>]

----- Original Message -----
From: Hans Hagen <pragma@wxs.nl>
To: Giuseppe Bilotta (Oblomov) <oblomov@freemail.it>
Cc: <ntg-context@ntg.nl>
Sent: Tuesday, September 12, 2000 10:38 PM
Subject: Re: Foreign Languages / interface translation

> At 05:30 PM 9/12/00 +0200, you wrote:
>
> >> (1) mult-com.tex
> >
> >How do I change this files? The format is somewhat strange, and for some
> >of them I don't know the correct translation, or it's quite longish. Does
> >this pose any problem?
>
> Do not change the files yet. Just prepare a list first or make sure that
we
> can use a perl script to merge the files. We must not get out of tune.
>
> There is no problem in the length, but keep an eye on what users want. If
> you don't know the translation yet, keep it english for the moment.
>

The problem with length is that the list is too long :-); and while for some
words I can guess the meaning (or the context in which to use them), this is
not true for all of them; more over: when can I use accented letters, and
when can I not?

If my hypothesis is correct, I cannot use accented letters in mult-com nor
in mult-con; this poses some problem, since quite a lot of words actually
need the accent.

> >> (2) mult-con.tex
>
> Czech also has that problem. The 'one' is used as a stand alone number, in
> mult-con there are only keywords and values. Although we may some day
> consider a gender specific label prefix mechanism (when needed).

By stand alone number, do you mean something like [columns=one] ? in this
case, the gender should anyway be the one of the word it refers to (columns,
in this case)

>
> >For both the previous files, is it possible to do something in such a way
> >that italian keys are defined, e.g., like english?
>
> yes:
>
> dutch: dutchkey    englishkey
>        germankey   czechkey
>        italiankey

Sorry, I didn't express myself clearly: do I have to copy all of the
definitions, or do I just add the ones for which the word is different from
the english one?
(never mind, reading further I saw I have to do them all *sigh* it's a long
list!)

>
> >> (3) grep on \startmessages
> >
> >I translated them. This was the easy part :-)
>
> All of them? That's fast!

It's not that fast ... the real problems are mult-com and mult-con:
translating a list of words is much longer than translating a sentence ...

> You can test that with the file cont-it.tex i'll
> send you.
>
>   texexec --make --alone it
>
> will generate you a version with an english interface and italian
messages.

It *seems* to work (that is, with the "original" mult-com and mult-con, but
with all the messages translated).

>
> >Does this mean that all I have to do is add the italian translation and
then
> >send you the appropriate files?
>
> messages    : isolate them in one file

Uh-oh: I put them all in the corrispective files ... can you grep them out?
(I can't do it easily, being in a Win32 environment). I guess that you can
easily find them by looking for "italian" ...

> keys/values : add a header 'italian' and add keywords, use - for unknown
> entries

That's the hard part ... oh my! is there an "easy" way to do it?

And what about the corrected lang-ita.tex?

--
Giuseppe Bilotta