Is there a Tajik locale file for OpenOffice.org already available?

Created attachment 34464 [details]
Tajik locale data tg-TJ.XML file

Yes, we have Tajik locale data file. Please find attached.

Thanks,
Murod

Ok, the steps you should be following are outlined at
http://translate.sourceforge.net/doc/project-openoffice.html

This report is marked as an ENHANCEMENT and now it includes a tg-TJ.xml locale
file, which is good.

By the way, add your details at
http://l10n.openoffice.org/languages.html
According to the page, send an e-mail to Pavel Janik. This way, if anyone else
is interested in Tajik OpenOffice, they will be contacting you.

Created attachment 34894 [details]
Patch for Tajik language inclusion (langtab.src)

reassigning...

Sorry, but the attached locale data file doesn't even pass an XML validation,
also after changing all "TRUE" to "true" and "FALSE" to "false" there remain
about 20 errors. Please try with the service available at
http://www.validome.org/ and correct the errors and attach the corrected file to
this issue. When done, please reassign it to me ('er').

Btw, Mircha, what is your full name and did you sign the JCA? Without that we
can't include your contribution, please see
http://contributing.openoffice.org/programming.html#jca

Thanks
  Eike

Created attachment 34918 [details]
Tajik locale file (corrected)

Hello,
Yes, I have submitted JCA much earlier. I suddenly discovered that locale file
is not valid. I actually generated it using tool from
http://www.it46.se/localegen and submitted without checking. 
Thanks,

My full name is Murod Latifov A.

Hi Murod,

> I suddenly discovered that locale file is not valid. I actually
> generated it using tool from http://www.it46.se/localegen and
> submitted without checking.

Which actually makes me wonder because http://www.it46.se/localegen
lists a tg_TJ locale
http://www.it46.se/localegen/locale/1140753715_tg_TJ.xml created on
Feb24 by you that is completely different from your's attached to this
issue, it has different separators and different data, passes XML
validation and compiles. Your data now passes validation, but building
it results in:

parsing document tg_TJ started
Warning: ThousandSeparator is an ' ' ordinary space, this should be a
non-breaking space U+00A0 instead.Warning: Don't forget to adapt FormatCode
elements when changing separators.
Error: ThousandSeparator not present in FormatCode formatindex="4".
Warning: formatindex="4","44","45" are the only FormatCode elements checked,
there may be others that have errors.
Error: in data for tg_TJ: 1
parsing document tg_TJ finished

The error is because the ThousandSeparator defined is the ordinary
space, but used in the number formats is the correct non-breaking space.

Please clarify what the correct data would be.

  Eike

Created attachment 34946 [details]
Thousands/decimal separator corrected

Now the format codes don't match the separators defined:

parsing document tg_TJ started
Warning: DecimalSeparator is different from Time100SecSeparator, this may be
correct or not.
Warning: Don't forget to adapt FormatCode elements when changing separators.
Error: DecimalSeparator not present in FormatCode formatindex="4".
Error: Ordering of ThousandSeparator and DecimalSeparator not correct in
formatindex="4".
Warning: formatindex="4","44","45" are the only FormatCode elements checked,
there may be others that have errors.
Error: in data for tg_TJ: 2
parsing document tg_TJ finished

Furthermore, some of the currency format codes use [$р.-419] while some
others use [$с.-419]. The currency symbol used there and the one defined
with <Currency usedInCompatibleFormatCodes="true"> must be identical.
And the LCID number is wrong now, instead of 419 it must be 428.
All that was correct in the first version you submitted.

There are two Currency elements defined, both with CurrencyID TJS, the
default and usedInCompatibleFormatCodes="true" element says сом. would
be the CurrencySymbol, but that is not used in the FormatCode elements.
The other Currency element has с. as the CurrencySymbol. So which one is
it? There should be only one Currency element for a given currency. Btw,
BankSymbol should be identical to the CurrencyID ISO code.

trueWord and falseWord don't seem to be correct, they are
<trueWord>?а??он?</trueWord>
<falseWord>?албак?</falseWord>
and the question marks seem to result from some encoding error an editor
produced, probably not capable of UTF-8. Same for the two
<FollowPageWord>са?.</FollowPageWord>
elements. These encoding errors were not present in the first version, but
are already in the previous version.
Please also check the other elements like day and month names.

The best probably would be to start from scratch again using the locale
generator, which also takes care that the FormatCode elements match the
separators defined.

Btw, I'll be on vacation for two weeks, I may check back for this issue,
but don't expect answers before April.

Created attachment 34956 [details]
Corrected tg-TJ.xml

Hi,

We decided to shift back to space, this will remove current error, ignore error
for "space" in thousandsseparator. The rest is checked, and changed as was
instructed by you. '?' in са'ҳ' is - U+04b7, the other two '?'s are U+049b and
U+04e3. These are pure tajik characters, not supported by most unicode fonts. 

Please change <DateSeparator>/</DateSeparator> to
<DateSeparator>.</DateSeparator>. If validator gives error for space, then it
somehow should be 'space', maybe quotation marks with space inside? In case
space is not valid separator we will change it. 


thanks

Hi Murod,

> We decided to shift back to space, this will remove current error,
> ignore error for "space" in thousandsseparator. The rest is checked, and
> changed as was instructed by you.

This is not a matter of removing an error, but what the correct data
would be. If the ThousandSeparator used with Tajik in Tajikistan is a
comma, then use that and adapt the format codes. If it is a space
indeed, fine.

> '?' in 'ҳ' is - U+04b7, the other two '?'s are U+049b and U+04e3.
> These are pure tajik characters, not supported by most unicode fonts.

This is not a font problem. The question marks I mentioned are literal
0x3f question marks, it is broken in the data submitted. Btw, also the
IndexKey element has a broken character there, and it occurs also in day
names, and also in the CurrencyName.

> Please change <DateSeparator>/</DateSeparator> to
> <DateSeparator>.</DateSeparator>.

Well, yes, I can do that, but why didn't you submit data that has it?

> If validator gives error for space, then it somehow should be 'space',
> maybe quotation marks with space inside? In case space is not valid
> separator we will change it.

I don't quite understand. Space where? The group (AKA thousand)
separator? Again: what is the correct separator that is used with Tajik
in Tajikistan? If it is a space, fine. Then it just should be the
non-breaking space U+00A0 instead of the ordinary space. That is already
used in the format codes, so just should also be defined as
ThousandSeparator element. That is what the warning is about.

Here the current compilation warnings and errors of the last data file
you submitted:

Warning: DecimalSeparator is different from Time100SecSeparator, this may be
correct or not.
Warning: ThousandSeparator is an ' ' ordinary space, this should be a
non-breaking space U+00A0 instead.
Warning: Don't forget to adapt FormatCode elements when changing separators.
Error: DecimalSeparator not present in FormatCode formatindex="4".
Error: ThousandSeparator not present in FormatCode formatindex="4".
Error: Ordering of ThousandSeparator and DecimalSeparator not correct in
formatindex="4".
Warning: formatindex="4","44","45" are the only FormatCode elements checked,
there may be others that have errors.
Error: in data for tg_TJ: 3

The DecimalSeparator is defined as dot '.' but the format codes use
comma ',' instead.

What about the Time100SecSeparator? It is defined as ',' and the format
code MM:SS,00 uses it, but should it really be different from
DecimalSeparator or is it just because also all other formats use ',' as
a decimal separator?

Hello,

I made new tg_TJ.xml from scratch. It is OK now, i made local build with no
error reported. I hope it works this time. This attachment will be named
050406tg_TJ.xml.

Created attachment 35503 [details]
Corrections for tg_TJ.xml

Ah, yes, that looks much better and at least is syntactically correct.

Thanks
  Eike

Hello,

We submitted Tajik SDF file that was accepted for OOo 2.0.3, can we change
locale file target as well? 

Thanks,

Murod

No, sorry, it's too late for 2.0.3. We already closed feature development
integration for this release. I didn't notice you had a Tajik UI localization
ready for this release and this issue could be related.

In CWS locales204

i18npool/source/localedata/data/Attic/tg_TJ.xml  1.1.2.1
i18npool/source/localedata/data/localedata_others.map  1.10.10.4
i18npool/source/localedata/data/makefile.mk  1.33.10.4
i18npool/source/localedata/localedata.cxx  1.40.8.4
svx/source/dialog/langtab.src  1.67.112.8

Reassigning to QA.

re-open issue and reassign to oc@openoffice.org

reassign to oc@openoffice.org

reset resolution to FIXED

verified in internal build cws_locales204

closed because fix available in src680_m176