Re: Fw: Debian release-notes: enable PDF build for Japanese ?
On Thu, 15 Jun 2017 22:05:17 +0200
Holger Wansing wrote:
> I just managed to get the PDF build successfully running for Japanese
> (that is currently disabled for the release-notes, but now possible due
> to improvements in dblatex).
> Attached is a patch to introduce this.
> It moves font selection from the Makefile into a separate xetex_param.xsl
> file: there vlgothic fonts are selected for Japanese, and for all other
> languages the same fonts as before are used.
> It also switches ja from xmlroff to dblatex.
> I have attached Japanese variants of the release-notes for amd64, for you
> to check the result: the PDF and also the txt and the html variants, because
> they have also changed (because the whole build process for Japanese is
> switched from xmlroff to dblatex).
> Could you take the time to check, if everything is fine?
actually it could be built on local vm even when it got disabled long time ago;
the cause might be lack of fonts on -master, though,
just a guess as no log for that time remain now;
the repeated logs were completely cipher for me at the time :p
early this year I tried using clean stretch,
but some required packages cannot be co-installed at the time;
I haven't managed to have time to play with it again :p
thanks for taking care of the translations :D
> When building the PDF, I get the following notes:
> Character U+30B1 (ケ) not in font '[lmmono10-italic]'
> Character U+30B8 (ジ) not in font '[lmmono10-italic]'
> Character U+30C3 (ッ) not in font '[lmmono10-italic]'
> Character U+30D1 (パ) not in font '[lmmono10-italic]'
> Character U+30FC (ー) not in font '[lmmono10-italic]'
> What are these characters? Can you see places, where they are missing?
> And even if they are missing somewhere, maybe we can enable the PDF build
> nevertheless, because it's better with the missing characters then having
> no PDF at all?
those are so-called Katakana and
what this log shows is vlgothic is not used for them;
set****font vars are also needed in addition to setCJK****font
as xsl:otherwise clause is not applied there
--
victory
no need to CC me :-)
Reply to: