[XeTeX] Xetex in lao (another try)

[Mojca Miklavec]

On Mon, Sep 23, 2013 at 6:01 AM, Robert M Campbell wrote:
> Mojca,
>
> As for patching ICU, here is what they did for me:
> http://bugs.icu-project.org/trac/changeset/34229
>
> It should be 'copy-and-paste' for many parts, but might be wise to have
> someone more experienced in programming than myself double check and compile
> things...
>
> Or better yet, is there someone who would be willing to mentor me through
> the steps, so that I can get some experience in this field? (I'd like to get
> a better feel for programming, but have minimal experience in C/C++) If so,
> though it may take more work, could help me in the future submit patches
> more easily.

If this patch is all that is needed and if it applies cleanly on 51.1
(I didn't check), Peter or Khaled might be willing to apply it to TeX
Live repository and there isn't much effort involved (and since you
don't have commit rights there is not much that you can do). If it
doesn't apply cleanly, it might need some manual intervention (in that
case you could help with creating a new patch against the version in
TeX Live's SVN repository). Hopefully this patch doesn't interfere
with other code and doesn't require additional functionality that
wasn't present in 51.1.

If you want you can checkout
     svn co svn://tug.org/texlive/trunk/Build
and test if the patch downloaded from the url mentioned above
(http://bugs.icu-project.org/trac/changeset/34229?format=diff&new=34229)
can be applied to source/libs/icu/ (using the command "patch").

> I'd like to do something similar for firefox, inkscape,
> openoffice, libreoffice, etc. The more I can do for them upfront, the
> better.

I don't know anything about Inkscape and I don't know how
Open/LibreOffice and Firefox deal with word splitting, but TeX
hyphenation patterns for syllabification should "automatically" work
with Open/LibreOffice and Firefox for example (that is: once they are
created and included upstream; no special patches are needed).

> ICU does various boundary detections, but if only the word break iterator is
> being used, then a solution for syllables would be needed for hyphenation...

Are you suggesting that the patch at
    http://bugs.icu-project.org/trac/changeset/34229
is sufficient for syllabification? I don't quite understand the code,
but it seems to simple to me to implement syllabification. It looks to
me like it does dictionary-based word splitting only.

Of course it's a different question of how to use syllabification from
ICU in XeTeX if one is/was available, but at the moment I'm not sure
if it is implemented at all.

If syllabification is still needed, I have a semi-working patterns,
but they need some fixes and last time when I wanted to finish them, I
didn't get sufficient feedback/explanation. I'm still willing to
finish them, but I don't have time to work on it right now.

> Thankfully, I have a Lao Syllable List attached (ods spreadsheet), as well
> as a document (pdf) describing how things work (if a regex method would be
> preferred). Could this be useful for hyphenation?

Yes, it could. I didn't open the syllable list yet and I'm familiar
with the other PDF document, but I had a bunch of questions a while
back.

> What is the best approach?

I don't know if ICU supports syllabification, but if it does, it makes
sense to make sure that it will work in ICU.

I'm willing to finish the work on patterns if it turns out to be
useful/needed, but probably not in the next few months. (You can use
the partial/imperfect patterns until then, or try to finish the work
yourself.)

If/when I start working on patterns again, it would be really helpful
to have a free utility to do syllabification on an arbitrary text. And
if that is not possible at least a long properly split text and
someone to answer my questions.

Mojca