[Po4a-devel]Winedocs and Po4a dependencies

Martin Quinson martin.quinson@loria.fr
Fri, 27 May 2005 09:44:42 +0200


--n8g4imXOkfNTN/H1
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable


On Thu, May 26, 2005 at 11:40:27PM +0200, Denis Barbier wrote:
> On Thu, May 26, 2005 at 07:49:44PM +0200, Martin Quinson wrote:
> > On Thu, May 26, 2005 at 05:52:39PM +0200, Francois Gouget wrote:
> [...]
> > >  * Text::WrapI18N (perl module)
> > >    Pure perl (so easy to check in) but depends on Text::CharWidth whi=
ch=20
> > > is not pure perl.
> > >    Provided by libtext-wrapi18n-perl on Debian. Found no RPM packages=
=20
> > > providing it.
> > >=20
> > >    Text::WrapI18N was not used in po4a 0.16.2. I initially thought it=
=20
> > > was used to wrap the text being output to the .po and .sgml files but=
 in=20
> > > fact it seems to only be used to print messages, warnings and errors.=
=20
> > > Why is it needed? Doesn't a simple print work fine?
> >=20
> > This module becomes important when you want to wrap CKJ languages
> > (japaneese, corean), which don't have any spaces, if I understood well.=
 So
> > finding where you can cut the sentence properly is not as easy as in, s=
ay,
> > french.
> >
> > So, actually, we *ought* to use it all around the place. Maybe through a
> > wrapper such as the one you propose for gettext...
>=20
> AFAICT this module is only useful to print error messages, so this is
> all about cosmetics and a wrapper (no, this one does not wrap lines ;))
> should be fine.  When this module is not present, no wrapping is performe=
d.

s/ought/should/

I know we don't use it very much, but I persist to say that it would be
better if we did so.=20

I now remember why it's not used in the Po.pm module (but there is no excuse
why it's not used in Sgml.pm and such). We have to preserve the space at the
end of the line here, and all wrapping modules keep eating it. They change:
 "toto tutu toto tutu "
to
 "toto tutu"
 "toto tutu"
But to preserve the po file semantic, we need:
 "toto tutu "
 "toto tutu "

Explaining why we keep using a wrapping algorithm which is known to be
broken for some languages. I wanted for ages to fill a wishlist bug report
asking Text::WrapI18N to provide an "space preservation" option. But I
didn't yet, my bad. I'll try to do so soon.

Now, about using it through a wrapper, I think this is a good idea. The
wrapper could even be smart enough to detect that you're using thai (or
another language making this module mandatory) encoding (or chars, if UTF8),
you're going into problem and emit a warning.

> > >  * nsgmls (command line tool)
> > >    Needed by po4a to parse the Sgml files.
> > >    Provided by sp on Debian, sgml-tools or openjade on Mandrake and=
=20
> > > Fedora Core, opensp on SUSE.
> > >=20
> > >    This is an essential part of po4a. So it will have to remain as a=
=20
> > > dependency which is reasonable.
> >=20
> > Actually, I'd like since a long time to kill this one, changing it to an
> > home-made parser such as the one used in xml or so...
>=20
> I thought you had no time ;)

Sure. Once again, I was unclear, sorry. I'd love to see disapear, but won't
do anything for this myself in the next months. I just wanted to note that
the only dependencies Francois were ok with are the ones I'd like the most
to kill :)

Bye, Mt.

--n8g4imXOkfNTN/H1
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFCls/qIiC/MeFF8zQRAgE2AJ9wbGraCAbkpIDjw2ybqHE+/2ZqAwCeIB/a
jjOSx4vEAOtBgXkbYbM9iNE=
=ltf3
-----END PGP SIGNATURE-----

--n8g4imXOkfNTN/H1--