[ wxwindows-Patches-1858926 ] Missing languages in src/common/intl.cpp

Sun Mar 2 15:16:47 PST 2008

Patches item #1858926, was opened at 2007-12-27 14:26
Message generated for change (Comment added) made by vaclavslavik
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=309863&aid=1858926&group_id=9863

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: bug fix
Status: Closed
Resolution: Accepted
Priority: 3
Private: No
Submitted By: Tim Kosse (botg)
Assigned to: Nobody/Anonymous (nobody)
Summary: Missing languages in src/common/intl.cpp

Initial Comment:
The following languages are missing in src/common/intl.cpp:
ka_GE
ku_TR
ne_NP

Patch applies against both WX_2_8_BRANCH as well as trunk.

----------------------------------------------------------------------

>Comment By: Vaclav Slavik (vaclavslavik)
Date: 2008-03-03 00:16

Message:
Logged In: YES 
user_id=80834
Originator: NO

FYI, there's no such thing as LANG_VALENCIAN constant in win32 headers, as
was done in this patch.

----------------------------------------------------------------------

Comment By: Vadim Zeitlin (vadz)
Date: 2008-02-17 21:11

Message:
Logged In: YES 
user_id=71618
Originator: NO

Thank you for your patch, it has been applied to both svn trunk and 2.8
branch!

----------------------------------------------------------------------

Comment By: Tim Kosse (botg)
Date: 2008-02-17 20:27

Message:
Logged In: YES 
user_id=156254
Originator: YES

Another update:

- Keep the Cyrillic entries
- The sr_YU entries is still in for backwards compatibility
- Add entry with SUBLANG_DEFAULT and name "Serbian" to langtabl.txt so
that wxLANGUAGE_SERBIAN gets defined
File Added: intl-v4.patch

----------------------------------------------------------------------

Comment By: Vaclav Slavik (vaclavslavik)
Date: 2008-02-12 12:54

Message:
Logged In: YES 
user_id=80834
Originator: NO

> The issue with Serbian is that there's no Serbian entry with
> SUBLANG_DEFAULT in langtabl.txt

And that's a problem why exactly?

> A language code on its own is not a valid locale. 

Not always, it's up to the implementation. Of course, the fact that recent
glibc versions always use this form is reason enough to update the rest of
the table.

> Furthermore, langtabl.txt currently uses sr_YU for both Cyrillic and
Latin
> script. YU (Yugoslawia) does not exist anymore though.

It should still be handled for older systems (if setting sr_SR fails, try
sr_YU), there's couple of other exceptions like this in intl.cpp.

> Correct would be sr_SR with SUBLANG_DEFAULT and sr_SR at latin with
> SUBLANG_SERBIAN_LATIN.

Current mapping for Windows (wxLANGUAGE_SERBIAN_CYRILLIC ->
SUBLANG_SERBIAN_CYRILLIC and wxLANGUAGE_SERBIAN_LATIN ->
SUBLANG_SERBIAN_LATIN) looks perfectly correct to me, why do you say
SUBLANG_DEFAULT should be used even though we know it's
SUBLANG_SERBIAN_CYRILLIC we want?

----------------------------------------------------------------------

Comment By: Tim Kosse (botg)
Date: 2008-02-12 12:35

Message:
Logged In: YES 
user_id=156254
Originator: YES

> why did we lose wxLANGUAGE_SERBIAN?

The issue with Serbian is that there's no Serbian entry with
SUBLANG_DEFAULT in langtabl.txt

Furthermore, langtabl.txt currently uses sr_YU for both Cyrillic and Latin
script. YU (Yugoslawia) does not exist anymore though.

Correct would be sr_SR with SUBLANG_DEFAULT and sr_SR at latin with
SUBLANG_SERBIAN_LATIN. At least according to the glibc loclales I have
installed.

I've updated the patch accordingly.

> I don't know if "ka_GE" and "ku_TR" should be used instead of just "ka"

A language code on its own is not a valid locale. Just try the internat
sample under wxGTK and try to set language to Georgian, it'll fail.

There are many more languages without country code as well, but I don't
have the time to look up the country codes for all of them.

> the first change to langcodes.tex removes a trailing space -- this
shouldn't be done, tex2rtf needs it.

Fixed
File Added: intl-v3.patch

----------------------------------------------------------------------

Comment By: Vadim Zeitlin (vadz)
Date: 2008-02-10 14:49

Message:
Logged In: YES 
user_id=71618
Originator: NO

Switching to pending because IMO this can't be applied before
wxLANGUAGE_SERBIAN issue is resolved.

----------------------------------------------------------------------

Comment By: Vadim Zeitlin (vadz)
Date: 2008-01-02 21:40

Message:
Logged In: YES 
user_id=71618
Originator: NO

I don't know if "ka_GE" and "ku_TR" should be used instead of just "ka"
and "ku" but why did we lose wxLANGUAGE_SERBIAN? I think we should really
have it and we definitely can't remove it from 2.8 branch...

P.S. To whoever applies the patch: the first change to langcodes.tex
removes a trailing space -- this shouldn't be done, tex2rtf needs it.

----------------------------------------------------------------------

Comment By: Tim Kosse (botg)
Date: 2007-12-30 15:44

Message:
Logged In: YES 
user_id=156254
Originator: YES

Updated patch:

- misc/languages/langtabl.txt updated without the duplicates
- Added Valencian ca_ES at valencia
- fixed misc/languages/genlang.py so that generated stuff matches what's
currently in the generated files (e.g. different syntax in
docs/latex/wx/langcodes.tex)
- Added comments about generated content to intl.h/cpp as it's easy to
miss otherwise.
- Ran genlang.py and pasted the generated code into the target files.

Updated patch tested against both WX_2_8_BRANCH as well as trunk.
File Added: intl-v2.patch

----------------------------------------------------------------------

Comment By: Vaclav Slavik (vaclavslavik)
Date: 2007-12-27 14:35

Message:
Logged In: YES 
user_id=80834
Originator: NO

This is the wrong place to do it, see the comment about this being
generated code in the file (we should probably split generated parts out
into a file of its own...). You have to update the source in misc/languages
instead.

The changes themselves don't look right either. There shouldn't be
duplicate entries for single language -- if e.g. "ne" locale isn't
legitimate (why?!), then only "ne_NP" should be there, not both. If that's
the case, an explanation of the reasons would be useful, too.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=309863&aid=1858926&group_id=9863