Re: Texidoc strings

[John Mandereau]

On 2008/07/15 23:16 +0200, Francisco Vila wrote:
> 2008/7/15 Graham Percival <address@hidden>:
> > This is trickier than you think, because input/lsr/ stuff is
> > automatically junked and updated from LSR.  If the snippet in LSR
> > is updated, that new description would need to get put into
> > texidoc/, and the translation infrastructure would need to mark it
> > as modified.
> 
> Alternatively, given that I am first copying lsr/thesnippet.ly into
> texidocs/thesnippet.texidoc and then overwriting the English string
> with the Spanish translation, I could simply NOT to overwrite it but
> copy it aside into the file, from now on (many of them still left).

I could make makelsr.py grab latest texidocs in English from input/new
and LSR and update them in existing .texidocs and create .texidoc files
if necessary.  This is not so tricky, but this requires parsing header
fields like texidoc = "String with possible escaped \"quotes\"... ".

I hope I can do this soon, but I can give no warranty: I'm busy in real
life to move to Paris this summer, and I'm starting learning Perl to
help texi2html replace makeinfo.  This last bit has not been announced
officially yet, but discussion with Karl Berry and Patrice Dumas has got
further enough so that no other reasonable solution has been found to
boost GNU Texinfo development as needed, and after what I did to
integrate docs translations in LilyPond with naive, ad-hoc Python
wrappers, it's obvious I should get involved into this.  This doesn't
mean I'm going to do less for LilyPond -- i.e. mainly maintaining
technical details of the docs, makefiles and Co. -- but this means I
won't be able to do more in the short term.


> What moves me is the sensation that I am deleting some useful
> reference data for translators other than me, whenever I overwrite the
> English strings.

This is a wrong feeling, because the strings in English you overwrite
didn't exist previously in input/texidocs :-p

Cheers,
John