Skip to content

Updated Lithuanian 8-dots table, added a test for it.#218

Merged
egli merged 1 commit into
liblouis:masterfrom
rimas-kudelis:master
Nov 3, 2016
Merged

Updated Lithuanian 8-dots table, added a test for it.#218
egli merged 1 commit into
liblouis:masterfrom
rimas-kudelis:master

Conversation

@rimas-kudelis

Copy link
Copy Markdown
Contributor

No description provided.

@egli egli left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I do not know Lithuanian, but the patch seems of fantastic quality: tests are there, well commented, Makefiles are updated. The only thing that is missing is a NEWS entry :-). But I can take care of that.
Thanks for your contribution

@egli egli merged commit 00b995e into liblouis:master Nov 3, 2016
@egli egli added this to the 3.1 milestone Nov 3, 2016
@rimas-kudelis

Copy link
Copy Markdown
Contributor Author

Glad you liked it. :)
Do you guys share your patches with brltty, or should I submit a pull request for them as well?

@rimas-kudelis

Copy link
Copy Markdown
Contributor Author

Oh, by the way...
There also exists a Lithuanian hyphenation dictionary, but it is licensed under LaTeX Project Public Licence. Is it acceptable to include it in liblouis' repository?
On the other hand, does it even make sense for liblouis to host own copies of these files if you don't modify them?

@egli

egli commented Nov 3, 2016

Copy link
Copy Markdown
Member

To my knowledge we do not share any patches with brltty. They might watch our repo, but to be safe I'd submit a PR to them as well

@egli

egli commented Nov 3, 2016

Copy link
Copy Markdown
Member

The hyphenation library is another question. Generally I think we'd like to stick to LGPL code and tables. As far as I know the existing hyphenation tables are either created specifically for liblouis, as is the case for the danish hyphenation table, or they originate from OpenOffice and are licensed under LGPL.

The licensing of the hyphenation tables is a bit murky, see the wiki page on table licensing. On the other hand according to the Free Software Foundation including hyphenation tables constitutes "mere aggregation", see #26 (comment), which means that we could legally include hyphenation tables with a different license as long as this license allows for redistribution.

So ideally we should extend liblouis to use standard hyphenation tables. Failing that I guess we could add the Lithuanian hyphenation dictionary.

@bertfrees

Copy link
Copy Markdown
Member

Note that Liblouis can't make use of LaTeX tables, only OpenOffice.org tables. (I assume you are talking about a LaTeX table because you mentioned their license). The two type of tables have a similar format, and OpenOffice.org tables can be generated from LaTeX tables, but Liblouis can't directly process the LaTeX format.

@rimas-kudelis

rimas-kudelis commented Nov 3, 2016

Copy link
Copy Markdown
Contributor Author

@bertfrees:

...and OpenOffice.org tables can be generated from LaTeX tables...

That's exactly why the table I'm talking about is licensed like that. :)

This hyphenation table was generated from LPPL-licensed TeX pattern files more than 10 years ago, and had't been changed until the last year when it was re-converted from the same original source following newer instructions. I suppose that means it is nearly perfect. :)

This table (or its older version) is or was used by LibreOffice, Mozilla, OpenOffice.org and likely many other open-source projects. The project above (not LibO or AOO) is as close as it gets to being upstream source of this data, the only higher upstream being the original TeX patterns, which, it seems, haven't been changed since 2004 or perhaps even earlier either.

EDIT: i lied a little. Current upstream of our hyphenation table is http://tug.org/svn/texhyphen/trunk/hyph-utf8/tex/generic/hyph-utf8/patterns/txt/, from where it was taken and converted to OpenOffice.org/LibreOffice format. Interestingly, this source seems to quote a different, much more permissive license for the hyphenation data.

@egli:
I think that the less duplication of code (or data in this case), the better. If you can extend liblouis so that hyphenation tables don't have to be explicitly included in Braille conversion tables, I guess that would be the first step towards turning them into a completely external resource. In the long run, I think it would be best if these were treated by everybody just like ispell/aspell/myspell/hunspell dictionaries are – as data for libhyphen, not something every application or library should ship own copy of.

@rimas-kudelis

Copy link
Copy Markdown
Contributor Author

By the way, when preparing my test, I discovered that during test runs liblouis does not translate empty cells to an appropriate Unicode character (U+2800 BRAILLE PATTERN BLANK), but instead uses U+0020 SPACE. @egli, is this an oversight?

@egli

egli commented Nov 24, 2016

Copy link
Copy Markdown
Member

@rimas-kudelis I don't know if this an oversight. Might be by design. I don't know the code well enough

@bertfrees

bertfrees commented Nov 24, 2016

Copy link
Copy Markdown
Member

It is by design. (see also #208 (comment))

@rimas-kudelis

rimas-kudelis commented Nov 24, 2016

Copy link
Copy Markdown
Contributor Author

Well that confirms the status quo, but doesn't really explain the reasons behind the design decision, or even confirm that there was a decision at all.

@bertfrees

bertfrees commented Nov 24, 2016

Copy link
Copy Markdown
Member

I'm sorry I don't have an explanation for it. (I didn't design it).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants