Request for new language packages for Kurdish Sorani (ckb)

Kurdish (Iraq) needs an own locale definition, since Kurdish is an official
language in Iraq since 200x

duo to:
https://bugs.launchpad.net/ubuntu/+source/langpack-locales/+bug/266975

the locale are defined: http://www.zkurd.org/aras/ckb_IQ.txt

Created attachment 3704
localedata for Kurdish Sorani (CKB)

the localedata for Kurdish Sorani (CKB)

Comment on attachment 3704
localedata for Kurdish Sorani (CKB)

comment_char %
escape_char /
% Kurdish (Sorani) language locale for Iraq and Iran.
% Contributed by Aras Noori <email address hidden> and
% Erdal Ronahi<email address hidden>.
% Contact: Aras
% Language: ku
% Date: 2009-01-29
% Distribution and use is free, also
% for commercial purposes.
% History:
%

LC_IDENTIFICATION
title "Kurdish language locale for Sorani dialects"
source ""
address ""
contact "Aras"
email "<email address hidden>, <email address hidden>"
tel ""
fax ""
language "Kurdish"
territory "Iraq"
revision "1.0"
date "2009-01-29"
%
category "ckb_IQ:2000";LC_IDENTIFICATION
category "ckb_IQ:2000";LC_CTYPE
category "ckb_IQ:2000";LC_COLLATE
category "ckb_IQ:2000";LC_TIME
category "ckb_IQ:2000";LC_NUMERIC
category "ckb_IQ:2000";LC_MONETARY
category "ckb_IQ:2000";LC_MESSAGES
category "ckb_IQ:2000";LC_PAPER
category "ckb_IQ:2000";LC_NAME
category "ckb_IQ:2000";LC_ADDRESS
category "ckb_IQ:2000";LC_TELEPHONE
category "ku_IQ:2000";LC_MEASUREMENT

END LC_IDENTIFICATION

LC_CTYPE
copy "i18n"
END LC_CTYPE

LC_COLLATE

% Copy the template from ISO/IEC 14651
copy "iso14651_t1"

END LC_COLLATE

LC_MONETARY
% This is the POSIX Locale definition the LC_MONETARY category.
% These are generated based on XML base Locale difintion file
% for IBM Class for Unicode/Java
%
int_curr_symbol "<U0049><U0051><U0044><U0020>"
currency_symbol "<U062F><U002E><U0639><U002E>"
mon_decimal_point "<U002E>"
mon_thousands_sep "<U002C>"
mon_grouping 3
positive_sign ""
negative_sign "<U002D>"
int_frac_digits 3
frac_digits 3
p_cs_precedes 1
p_sep_by_space 1
n_cs_precedes 1
n_sep_by_space 1
p_sign_posn 1
n_sign_posn 2
%
END LC_MONETARY

LC_NUMERIC
% This is the POSIX Locale definition for the LC_NUMERIC category.
%
decimal_point "<U002E>"
thousands_sep "<U002C>"
grouping 3
%
END LC_NUMERIC

LC_TIME
% This is the POSIX Locale definition for the LC_TIME category.
% These are generated based on XML base Locale difintion file
% for IBM Class for Unicode/Java
%
% Abbreviated weekday names (%a)
abday "<U062D>";"<U0646>";/
     "<U062B>";"<U0631>";/
     "<U062E>";"<U062C>";/
     "<U0633>"
%
% Full weekday names (%A)
day "<U06CC><U06D5><U0643><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U062F><U0648><U0648><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U0633><U06CE><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U0686><U0648><U0624><U0631><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U067E><U06CE><U0646><U062C><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U0647><U06D5><U06CC><U0646><U06CC>";/
     "<U0634><U06D5><U0645><U0645><U06D5>";/
%
% Abbreviated month names (%b)
abmon "<U064A><U0646><U0627>";"<U0641><U0628><U0631>";/
     "<U0645><U0627><U0631>";"<U0623><U0628><U0631>";/
     "<U0645><U0627><U064A>";"<U064A><U0648><U0646>";/
     "<U064A><U0648><U0644>";"<U0623><U063A><U0633>";/
     "<U0633><U0628><U062A>";"<U0623><U0643><U062A>";/
     "<U0646><U0648><U0641>";"<U062F><U064A><U0633>"
%
% Full month names (%B)
mon "<U064A><U0646><U0627><U064A><U0631>";/
  ...

The correspondig bug in the Launchpad bug tracker for Ubuntu is
https://bugs.launchpad.net/ubuntu/+source/langpack-locales/+bug/266975

Aras is the reporter, not the assignee.

The file is ill-formed. The file is named ckb_IQ and you use it in "copy" in
various categories? You also have copy and definitions in categories like
LC_PAPER. You have to fix it up.

Comment on attachment 3704
localedata for Kurdish Sorani (CKB)

escape_char /
comment_char %
% Kurdish (Sorani) language locale for Iraq and Iran.
% Contributed by Aras Noori <email address hidden> and
% Erdal Ronahi<email address hidden>.
% Contact: Aras Noori
% Language: ku
% Date: 2009-04-14
% Distribution and use is free, also
% for commercial purposes.
% History:
% January 2009: Defining CKB locale
% March 2009: Adding rule for CKB
%

LC_IDENTIFICATION
title "Kurdish language locale for Sorani dialects - Central Kurdish"
source ""
address ""
contact "Aras"
email "<email address hidden>"
tel ""
fax ""
language "Kurdish"
territory "Iraq"
revision "1.1"
date "2009-04-15"
%

category "ckb_IQ:2000";LC_IDENTIFICATION
category "ckb_IQ:2000";LC_CTYPE
category "ckb_IQ:2000";LC_COLLATE
category "ckb_IQ:2000";LC_TIME
category "ckb_IQ:2000";LC_NUMERIC
category "ckb_IQ:2000";LC_MONETARY
category "ckb_IQ:2000";LC_MESSAGES
category "ckb_IQ:2000";LC_PAPER
category "ckb_IQ:2000";LC_NAME
category "ckb_IQ:2000";LC_ADDRESS
category "ckb_IQ:2000";LC_TELEPHONE
category "ckb_IQ:2000";LC_MEASUREMENT

END LC_IDENTIFICATION

LC_CTYPE
copy "i18n"
END LC_CTYPE

LC_COLLATE
% The Sorani Kurdish dialect is mainly written using a modified Arabic-based
alphabet with 33 letters.
% Unlike the regular Arabic alphabet, which is an abjad, Sorani is an alphabet
in which vowels are mandatory, making the script easy to read.
%
% The CKB (Sorani) alphabet order is:
% in Latin: a, b, c, ç, d, e, ê, f, g, h, i, î, j, k, l, ll, m, n, o, p, q,
r, rr, s, sh, t, u, uu, v, w, x, y, z
% ئ، ب، پ، ت، ج، چ، ح، خ، د، ر، ڕ، ز، ژ، س، ش،
ف، ڤ، ق، ع، غ، ك، گ، ل، ڵ، م، ن، و، وو، ۆ، هـ،
ی، ێ
% vowels: A, E, I, O, U, UU
% پیتەبزوێنەكان ئەمانەن: ئ، ا، ە، و، وو، ۆ،
ی، ێ،
%
% Copy the template from ISO/IEC 14651
copy "iso14651_t1"

collating-element <ئا> from <U0626><U0627>
collating-element <وو> from <U0648><U0648>
collating-element <لا> from <U0644><U0627>

collating-symbol <U0628>
collating-symbol <U062C>
collating-symbol <U0631>
collating-symbol <U0632>
collating-symbol <U0641>
collating-symbol <U0643>
collating-symbol <U0644>
collating-symbol <U0648>
collating-symbol <U06CC>

reorder-after <U0628> <U067E>
reorder-after <U062C><U0686>
reorder-after <U0631><U0695>
reorder-after <U0632><U0698>
reorder-after <U0641><U06A4>
reorder-after <U0643><U06AF>
reorder-after <U0644><U06B5>
reorder-after <U0648><U06C6>
reorder-after <U06CC><U06CE>

% Kurdish digits same as Arabic ones: they are the basic forms.
reorder-after <U0660>
<U0660> <0>;<PCL>;<MIN>;IGNORE
<U0661> <1>;<PCL>;<MIN>;IGNORE
<U0662> <2>;<PCL>;<MIN>;IGNORE
<U0663> <3>;<PCL>;<MIN>;IGNORE
<U0664> <4>;<PCL>;<MIN>;IGNORE
<U0665> <5>;<PCL>;<MIN>;IGNORE
<U0666> <6>;<PCL>;<MIN>;IGNORE
<U0667> <7>;<PCL>;<MIN>;IGNORE
<U0668> <8>;<PCL>;<MIN>;IGNORE
<U0669> <9>;<PCL>;<MIN>;IGNORE

reorder-end

END LC_COLLATE

LC_MONETARY
% This is the POSIX Locale definition the LC_MONETARY category.
% These are generated based on XML base Locale difintion file
% for IBM Class for Unicode/Java
%
int_curr_symbol "<U0049><U0051><U0044><U0020>"
currency_symbol "<U062F><U002E><U0639><U002E>"
mon_decimal_point "<U00...

I updated the file. please check if still has a bad format.

Subject: Re: Please add Kurdish locale for Kurdish
 Sorani (CKB)

Hi
I update the locale couple weeks ago, would you please check it again

http://www.sourceware.org/ml/libc-locales/2009-q2/msg00021.html

Its my pleasure to hear from you a feedback.

Regards
Aras

On Sat, Feb 7, 2009 at 5:53 AM, drepper at redhat dot
com<email address hidden> wrote:
>
> ------- Additional Comments From drepper at redhat dot com  2009-02-07 03:53 -------
> The file is ill-formed.  The file is named ckb_IQ and you use it in "copy" in
> various categories?  You also have copy and definitions in categories like
> LC_PAPER.  You have to fix it up.
>
> --
>           What    |Removed                     |Added
> ----------------------------------------------------------------------------
>             Status|NEW                         |WAITING
>
>
> http://sourceware.org/bugzilla/show_bug.cgi?id=9809
>
> ------- You are receiving this mail because: -------
> You reported the bug, or are watching the reporter.
>

Can you please attach the current version instead of adding it as a comment? The
latter destroys all the non-ASCII characters.

Created attachment 4013
new Locale info for CKB

hier is the new local info for CKB

Created attachment 4168
new fixed Locale info for CKB

fixed bugs in Message category

Created attachment 4228
Locale info for CKB

Info is provided, changing back to NEW

I get a lot of errors when I try to build this locale:

locales/ckb_IQ:62: LC_COLLATE: syntax error
locales/ckb_IQ:63: LC_COLLATE: syntax error
locales/ckb_IQ:64: LC_COLLATE: syntax error
locales/ckb_IQ:66: LC_COLLATE: syntax error
locales/ckb_IQ:67: LC_COLLATE: syntax error
locales/ckb_IQ:68: LC_COLLATE: syntax error
locales/ckb_IQ:69: LC_COLLATE: syntax error
locales/ckb_IQ:70: LC_COLLATE: syntax error
locales/ckb_IQ:71: LC_COLLATE: syntax error
locales/ckb_IQ:72: LC_COLLATE: syntax error
locales/ckb_IQ:73: LC_COLLATE: syntax error
locales/ckb_IQ:74: LC_COLLATE: syntax error
locales/ckb_IQ:76: trailing garbage at end of line
locales/ckb_IQ:77: trailing garbage at end of line
locales/ckb_IQ:78: trailing garbage at end of line
locales/ckb_IQ:79: trailing garbage at end of line
locales/ckb_IQ:80: trailing garbage at end of line
locales/ckb_IQ:81: trailing garbage at end of line
locales/ckb_IQ:82: trailing garbage at end of line
locales/ckb_IQ:83: trailing garbage at end of line
locales/ckb_IQ:84: LC_COLLATE: cannot reorder after U000006CC: symbol not known
locales/ckb_IQ:155: extra trailing semicolon
LC_NAME: invalid escape sequence in field `name_fmt'
LC_ADDRESS: invalid escape `%I' sequence in field `postal_fmt'
LC_ADDRESS: `lang_ab' value does not match `lang_term' value
LC_ADDRESS: `lang_lib' value does not match `lang_term' value
LC_ADDRESS: `country_ab2' value does not match `country_num' value
LC_ADDRESS: `country_ab3' value does not match `country_num' value

I am analyzing the Errors now and try to fix them as soon as I can.

Thanks & Regards
Aras

Created attachment 4357
CKB locale - updated

Hi,
I fixed some errors due to the occured Errors. How can I test it by myself
before release to bugzilla?

Regards

Any progress?

(In reply to comment #16)
> Created an attachment (id=4357)
> CKB locale - updated
>
> Hi,
> I fixed some errors due to the occured Errors. How can I test it by myself
> before release to bugzilla?
>
> Regards
>

Use localedef to compile your locale, and $LOCPATH if you don't want to install
it system-wide in order to test it.

The file isn't usable as-is, there are many problems when compiling it.

First, it must be in UTF-8.

Second, the collation rules seem all pretty bogus since there already are rules for all the characters defined. If needed, you have to redefine the relocation.

Third, there are many syntax errors.

Fourth, all the values for the fields must use the <U....> notation, not real strings.

Fifth, the values for some fields is plain wrong. localedef will tell you.

I did add the language code to localedef now.

Just run localedef like

   localedef -i ./YOURFILE -f UTF-8 ./SOMEDIR

Created attachment 5727
CKB-IQ locale info (Kurdish Sorani)

CKB-IQ locale info (Kurdish Sorani)

Your locale also sorted many control characters and ASCII punctuation characters.

I think there is no reason to deviate from the default for these characters, therefore I removed them.

If you have a good reason why some of these need to be sorted differently for Kurdish, please tell me.

Your locale sorted the Kurdish numbers at the top, i.e. before the
Western numbers. The default order (as you can see in the ckb_IQ.UTF-8.in sorting test file in my patch) sorts these in between the Western numbers. Like this:

    0
    ٠
    1
    ١
    2
    ٢
    3
    ٣
    4
    ٤
    5
    ٥
    6
    ٦
    7
    ٧
    8
    ٨
    9
    ٩

That is reasonably good, isn’t it?

Your locale also resorted all the ASCII letters to make upper case letters come first.

I.e.

A
a

instead of

a
A

Lower case first is what comes from

    copy "iso14651_t1"

When using CLDR for sorting, one can use an option
[caseFirst upper], see for example:

https://github.com/unicode-org/cldr/blob/master/common/collation/da.xml

glibc has no easy option to do that at the moment.

It is *possible* do sort A-Za-z differently in your locale *but*
if you do that you will get a weird order for all Latin characters you forget.
I.e. if you do not include äÄ in your sort order as well, they would still sort
lower case first. It is a lot of work to do this correctly for *all* Latin characters without a convenient option like CLDR’s [caseFirst upper],
I would recommend not doing that if it is not absolutely required.

(In reply to Mike FABIAN from comment #53)
> Your locale also resorted all the ASCII letters to make upper case letters
> come first.
>
> I.e.
>
> A
> a
>
> instead of
>
> a
> A
>
> Lower case first is what comes from
>
> copy "iso14651_t1"
>
> When using CLDR for sorting, one can use an option
> [caseFirst upper], see for example:
>
> https://github.com/unicode-org/cldr/blob/master/common/collation/da.xml
>
> glibc has no easy option to do that at the moment.
>
> It is *possible* do sort A-Za-z differently in your locale *but*
> if you do that you will get a weird order for all Latin characters you
> forget.
> I.e. if you do not include äÄ in your sort order as well, they would still
> sort
> lower case first. It is a lot of work to do this correctly for *all* Latin
> characters without a convenient option like CLDR’s [caseFirst upper],
> I would recommend not doing that if it is not absolutely required.

Hello Fabian,
thanks to your suggestions and notice. You are right with sorting (aA) as well with Numbers, this should be modified.
The kurdish alphabet order is:

ئ
U+0626

ا
U+0627

ب
U+0628

پ
U+067E

ت
U+062A

ج
U+062C

چ
U+0686

ح
U+062D

خ
U+062E

د
U+062F

ر
U+0631

ڕ
U+0695

ز
U+0632

ژ
U+0698

س
U+0633

ش
U+0634

ع
U+0639

غ
U+063A

ف
U+0641

ڤ
U+06A4

ق

U+0642

ک
U+06A9

گ
U+06AF
ل
U+0644
ڵ
U+06B5
م
U+0645
ن
U+0646
و
U+0648
ۆ
U+06C6
ھ
U+0647
ە
U+06D5
ی
U+06CC
ێ
U+06CE

thank you mike you is really appreciated i have pointed all my answers according to your question and suggestion to our locale as follow:

1. For positive sign and negative i agree with you let it be + and - .
2. For regular expression i didn't know how to type it in my language hope to hekp me solve this.
we have "ب" for Y in English and "ن" for N in English .
3.You right we type Iraq in Kurdish(Sorani) now changed.
4.We have Kurdish alphabet as Aras Noori wrote before my reply and i look at iso14651_t1 now all characters which is used in Kurdish are exist, these characters that you did add them are from Arabic language not Kurdish.

Can you send the .dat file with your last changes?

Best Regards

Thank you mike you your help is really appreciated
I have pointed all my answers according to your question and suggestion to our locale as follow:

1. For positive sign and negative i agree with you let it be + and - .
2. For regular expression i didn't know how to type it in my language hope to hekp me solve this.
we have "ب" for Y in English and "ن" for N in English .
3.You right we type Iraq in Kurdish(Sorani) now changed.
4.We have Kurdish alphabet as Aras Noori wrote before my reply and i look at iso14651_t1 now all characters which is used in Kurdish are exist, these characters that you did add them are from Arabic language not Kurdish.

Can you send the .dat file with your last changes?

Best Regards

> thanks to your suggestions and notice. You are right with sorting (aA) as
> well with Numbers, this should be modified.

So sorting

a
A

and

0
٠
1
١
...

is OK? I hope so ...

> The kurdish alphabet order is:

To achieve that order, this is enough:

   copy "iso14651_t1"

   reorder-after <S0631> % ر
   <S0695> % ڕ

   reorder-after <S0646> % ن
   <S0648> % و
   <S06C6> % ۆ

I added the test file ckb_IQ.UTF-8.in in my patch, this file is sorted
using the rules of my patched ckb_IQ locale, the sorted result should
be the same as the original file, otherwise the test fails.

As the test passes, the above collation rules work and achieve the
order as in the ckb_IQ.UTF-8.in test file.

I’ll paste this test file here again for your easy refererence:

0
٠
1
١
2
٢
3
٣
4
٤
5
٥
6
٦
7
٧
8
٨
9
٩
a
A
b
B
c
C
d
D
e
E
f
F
g
G
h
H
i
I
j
J
k
K
l
L
m
M
n
N
o
O
p
P
q
Q
r
R
s
S
t
T
u
U
v
V
w
W
x
X
y
Y
z
Z
ئ
ا
ب
پ
ت
ج
چ
ح
خ
د
ر
ڕ
ز
ژ
س
ش
ع
غ
ف
ڤ
ق
ک
گ
ل
ڵ
م
ن
و
ۆ
ه
ە
ی
ێ

Other characters not in this test file are sorted according to the defaults from

    copy "iso14651_t1"

(In reply to Jwtiyar Nariman from comment #56)
> Thank you mike you your help is really appreciated
> I have pointed all my answers according to your question and suggestion to
> our locale as follow:
>
> 1. For positive sign and negative i agree with you let it be + and - .

Your original locale had the positive sign empty.
Probably a mistake. So I’ll make it + now.

> 2. For regular expression i didn't know how to type it in my language hope
> to hekp me solve this.
> we have "ب" for Y in English and "ن" for N in English .

That is what I used:

yesexpr "^[+1yY<U0628>]"
noexpr "^[-0nN<U0646>]"

So these regular expressions except +, 1, y, Y, and ب as a yes answer.
And -, 0, n, N, and ن as a no answer.

> 3.You right we type Iraq in Kurdish(Sorani) now changed.
> 4.We have Kurdish alphabet as Aras Noori wrote before my reply and i look at
> iso14651_t1 now all characters which is used in Kurdish are exist, these
> characters that you did add them are from Arabic language not Kurdish.

I don’t understand. Most of these characters are used both in Arabic *and* Kurdish.

Created attachment 12194
ckb_IQ

> Can you send the .dat file with your last changes?

Here is the latest file with the changes I made.
I just added the + as the positive_sign.

Created attachment 12195
0001-Add-ckb_IQ-locale.patch

Updated patch.

Created attachment 12196
0002-Fix-ckb_IQ-Add-ckb_IQ-to-SUPPORTED-file-Add-ckb_IQ.U.patch

Updated patch.

(In reply to Mike FABIAN from comment #57)
> > thanks to your suggestions and notice. You are right with sorting (aA) as
> > well with Numbers, this should be modified.
>
> So sorting
>
> a
> A
>
> and
>
> 0
> ٠
> 1
> ١
> ...
>
> is OK? I hope so ...
>
> > The kurdish alphabet order is:
>
> To achieve that order, this is enough:
>
> copy "iso14651_t1"
>
> reorder-after <S0631> % ر
> <S0695> % ڕ
>
> reorder-after <S0646> % ن
> <S0648> % و
> <S06C6> % ۆ
>
> I added the test file ckb_IQ.UTF-8.in in my patch, this file is sorted
> using the rules of my patched ckb_IQ locale, the sorted result should
> be the same as the original file, otherwise the test fails.
>
> As the test passes, the above collation rules work and achieve the
> order as in the ckb_IQ.UTF-8.in test file.
>
> I’ll paste this test file here again for your easy refererence:
>
> 0
> ٠
> 1
> ١
> 2
> ٢
> 3
> ٣
> 4
> ٤
> 5
> ٥
> 6
> ٦
> 7
> ٧
> 8
> ٨
> 9
> ٩
> a
> A
> b
> B
> c
> C
> d
> D
> e
> E
> f
> F
> g
> G
> h
> H
> i
> I
> j
> J
> k
> K
> l
> L
> m
> M
> n
> N
> o
> O
> p
> P
> q
> Q
> r
> R
> s
> S
> t
> T
> u
> U
> v
> V
> w
> W
> x
> X
> y
> Y
> z
> Z
> ئ
> ا
> ب
> پ
> ت
> ج
> چ
> ح
> خ
> د
> ر
> ڕ
> ز
> ژ
> س
> ش
> ع
> غ
> ف
> ڤ
> ق
> ک
> گ
> ل
> ڵ
> م
> ن
> و
> ۆ
> ه
> ە
> ی
> ێ
>
> Other characters not in this test file are sorted according to the defaults
> from
>
> copy "iso14651_t1"

Sorting is good now, but adding these
  reorder-after <S0631> % ر
> <S0695> % ڕ
>
> reorder-after <S0646> % ن
> <S0648> % و
> <S06C6> % ۆ
iam not understanding because for example this " <S0695> % ڕ " how you order it?

(In reply to Jwtiyar Nariman from comment #62)

> > Other characters not in this test file are sorted according to the defaults
> > from
> >
> > copy "iso14651_t1"
>
> Sorting is good now, but adding these
> reorder-after <S0631> % ر
> > <S0695> % ڕ
> >
> > reorder-after <S0646> % ن
> > <S0648> % و
> > <S06C6> % ۆ
> iam not understanding because for example this " <S0695> % ڕ " how you
> order it?

copy "iso14651_t1"

contains

copy "iso14651_t1_common"

and some modifications which affect only Chinese and Japanese.

So we look into the iso14651_t1_common file to see what the default sort order is.

We find for example:

...
<S0631> % ARABIC LETTER REH
<S0632> % ARABIC LETTER ZAIN
<S0691> % ARABIC LETTER RREH
<S0692> % ARABIC LETTER REH WITH SMALL V
<S0693> % ARABIC LETTER REH WITH RING
<S0694> % ARABIC LETTER REH WITH DOT BELOW
<S0695> % ARABIC LETTER REH WITH SMALL V BELOW
<S0696> % ARABIC LETTER REH WITH DOT BELOW AND DOT ABOVE
...

Looking at this you see that ڕ U+0695 ARABIC LETTER REH WITH SMALL V BELOW
is sorted right after ڔ U+0694 ARABIC LETTER REH WITH DOT BELOW by default.
That is not what you want for Kurdish. For Kurdish, you want
ڕ U+0695 ARABIC LETTER REH WITH SMALL V BELOW to be sorted right after
ر U+0631 ARABIC LETTER REH.

This is achieved by the rule:

reorder-after <S0631> % ر
<S0695> % ڕ

Which removes U+0695 from its default position in the sort order
and inserts it again after U+0631.

reorder-after <S0646> % ن
<S0648> % و
<S06C6> % ۆ

does a similar thing to change the sorting of U+0648 and U+06C6.

To find out which of these rules I need, I created the ckb_IQ.UTF-8.in
test file first and wrote the Kurdish characters in the order you wanted
into that file.

Then I ran a test sort using a ckb_IQ locale which had *only*

LC_COLLATE
copy "iso14651_t1"
END LC_COLLATE

and *nothing* else.

The test sort showed that only U+0695, U+0648, and U+06C6 were sorted incorrectly.
All other characters from your list of Kurdish characters were sorted correctly
already. So I needed only to add rules to fix the sort order for these 3 characters.

You can see the same by just reading the iso14651_t1_common and find out which
of the Kurdish characters are already in the correct order in that file and which are not.
You have to do nothing for the characters which are already in correct order.
For the characters which are in a wrong position in iso14651_t1_common, you add
rules like

reorder-after <... collating-symbol after which to reorder ...>
<... the collating-symbol which should be reordered ...>

I found writing the test file and checking which characters are sorted
wrongly by default easier than staring at iso14651_t1_common. And it
is a good idea to have the test file anyway to make sure that the
Kurdish sort order always stays correct when something is changed in
glibc. If we have the test file, we will notice when some change causes a problem.

Thank you very much dear mike i got it, you made a great job, thanks again.
So now every thing is ready to be accepted in glibc.

Best Regards (In reply to Mike FABIAN from comment #63)
> (In reply to Jwtiyar Nariman from comment #62)
>
> > > Other characters not in this test file are sorted according to the defaults
> > > from
> > >
> > > copy "iso14651_t1"
> >
> > Sorting is good now, but adding these
> > reorder-after <S0631> % ر
> > > <S0695> % ڕ
> > >
> > > reorder-after <S0646> % ن
> > > <S0648> % و
> > > <S06C6> % ۆ
> > iam not understanding because for example this " <S0695> % ڕ " how you
> > order it?
>
> copy "iso14651_t1"
>
> contains
>
> copy "iso14651_t1_common"
>
> and some modifications which affect only Chinese and Japanese.
>
> So we look into the iso14651_t1_common file to see what the default sort
> order is.
>
> We find for example:
>
> ...
> <S0631> % ARABIC LETTER REH
> <S0632> % ARABIC LETTER ZAIN
> <S0691> % ARABIC LETTER RREH
> <S0692> % ARABIC LETTER REH WITH SMALL V
> <S0693> % ARABIC LETTER REH WITH RING
> <S0694> % ARABIC LETTER REH WITH DOT BELOW
> <S0695> % ARABIC LETTER REH WITH SMALL V BELOW
> <S0696> % ARABIC LETTER REH WITH DOT BELOW AND DOT ABOVE
> ...
>
> Looking at this you see that ڕ U+0695 ARABIC LETTER REH WITH SMALL V BELOW
> is sorted right after ڔ U+0694 ARABIC LETTER REH WITH DOT BELOW by default.
> That is not what you want for Kurdish. For Kurdish, you want
> ڕ U+0695 ARABIC LETTER REH WITH SMALL V BELOW to be sorted right after
> ر U+0631 ARABIC LETTER REH.
>
> This is achieved by the rule:
>
> reorder-after <S0631> % ر
> <S0695> % ڕ
>
> Which removes U+0695 from its default position in the sort order
> and inserts it again after U+0631.
>
> reorder-after <S0646> % ن
> <S0648> % و
> <S06C6> % ۆ
>
> does a similar thing to change the sorting of U+0648 and U+06C6.
>
> To find out which of these rules I need, I created the ckb_IQ.UTF-8.in
> test file first and wrote the Kurdish characters in the order you wanted
> into that file.
>
> Then I ran a test sort using a ckb_IQ locale which had *only*
>
> LC_COLLATE
> copy "iso14651_t1"
> END LC_COLLATE
>
> and *nothing* else.
>
> The test sort showed that only U+0695, U+0648, and U+06C6 were sorted
> incorrectly.
> All other characters from your list of Kurdish characters were sorted
> correctly
> already. So I needed only to add rules to fix the sort order for these 3
> characters.
>
> You can see the same by just reading the iso14651_t1_common and find out
> which
> of the Kurdish characters are already in the correct order in that file and
> which are not.
> You have to do nothing for the characters which are already in correct order.
> For the characters which are in a wrong position in iso14651_t1_common, you
> add
> rules like
>
> reorder-after <... collating-symbol after which to reorder ...>
> <... the collating-symbol which should be reordered ...>
>
> I found writing the test file and checking which characters are sorted
> wrongly by default easier than staring at iso14651_t1_common. And it
> is a good idea to have the test file anyway to make sure that the
> Kurdish sort order always stays c...

Posted to the mailing list for review:

https://sourceware.org/ml/libc-alpha/2020-01/msg00255.html
https://sourceware.org/ml/libc-alpha/2020-01/msg00256.html

https://github.com/mike-fabian/langtable/releases/tag/0.0.51

I added ckb_IQ.UTF-8 to langtable to make it usuable for installation on Fedora as soon as it is included in glibc.

By the way, how do you input Kurdish Sorani? Do you use a keyboard layout? Or do you need an input method?

(In reply to Mike FABIAN from comment #67)
> By the way, how do you input Kurdish Sorani? Do you use a keyboard layout?
> Or do you need an input method?

Yes we have and its available.

(In reply to Mike FABIAN from comment #66)
> https://github.com/mike-fabian/langtable/releases/tag/0.0.51
>
> I added ckb_IQ.UTF-8 to langtable to make it usuable for installation on
> Fedora as soon as it is included in glibc.

Our focus now is on Ubuntu because too much users in Ubuntu we have.

Created attachment 12212
0001-Add-new-locale-ckb_IQ-Kurdish-Sorani-spoken-in-Iraq-.patch

git log message changed according to Rafał Lużyński’s review.

Created attachment 12213
0002-Fix-ckb_IQ-BZ-9809.patch

Fixed according to Rafał Lużyński’s review.

(In reply to Mike FABIAN from comment #71)
> Created attachment 12213 [details]
> 0002-Fix-ckb_IQ-BZ-9809.patch
>
> Fixed according to Rafał Lużyński’s review.

Changed to this according to Rafał Lużyński’s suggestion:

d_t_fmt "%A %d %b %Y, %I:%M:%S %p"

date_fmt "%A %d %B %Y, %Z %I:%M:%S %p"

All otherchanges are just whitespace and formatting.

Rafał Lużyński’s review:

https://sourceware.org/ml/libc-alpha/2020-01/msg00281.html

(In reply to Mike FABIAN from comment #73)
> Rafał Lużyński’s review:
>
> https://sourceware.org/ml/libc-alpha/2020-01/msg00281.html

Thanks to your efforts, the locale is now ripe to join to the lib.

We have to wait until the release of glibc 2.31:

https://www.gnu.org/software/libc/
The current development version of glibc 2.31, releasing on or around February 1st, 2020.

The master branch has been updated by Mike Fabian <email address hidden>:

https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=4267522f5e0309f7606a8d1da5d436a166a719e2

commit 4267522f5e0309f7606a8d1da5d436a166a719e2
Author: Jwtiyar Nariman <email address hidden>
Date: Mon Jan 13 10:06:06 2020 +0100

    Add new locale: ckb_IQ (Kurdish/Sorani spoken in Iraq) [BZ #9809]

The master branch has been updated by Mike Fabian <email address hidden>:

https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=ae199e7d6423ed3bd0c8669381966ca4c58f4f49

commit ae199e7d6423ed3bd0c8669381966ca4c58f4f49
Author: Mike FABIAN <email address hidden>
Date: Mon Jan 13 10:12:07 2020 +0100

    Fix ckb_IQ [BZ #9809]

    Add ckb_IQ to SUPPORTED file.
    Add ckb_IQ.UTF-8.in collation test file.
    Mention new ckb_IQ locale in NEWS.

Hey dear Mike
I have downloaded new glibc 2.31 release but couldn't find ckb_iq localedata there? it was not planned to be there?

best regards.

Created attachment 12261
fixed typo in wednesday name in kurdish

Just a typo now fixed, replaced U+0624 with U+0627 in the name of Wednesday in kurdish.

Best Regards.

Created attachment 12262
Added reorder-end command which missing

Adding reorder-end because couldn't compile it with this error exist.

(In reply to Jwtiyar Nariman from comment #78)
> Hey dear Mike
> I have downloaded new glibc 2.31 release but couldn't find ckb_iq localedata
> there? it was not planned to be there?

Yes, of course, that’s what I wrote in

https://sourceware.org/bugzilla/show_bug.cgi?id=9809#c75

2.31 was already in code freeze, I could push this only *after* 2.31 was released. Therefore, the target milesstone of this bug is set to 2.32.

(In reply to Jwtiyar Nariman from comment #80)
> Created attachment 12262 [details]
> Added reorder-end command which missing
>
> Adding reorder-end because couldn't compile it with this error exist.

Not needed anymore, I rewrote the whole LC_COLLATE section, see

https://sourceware.org/bugzilla/show_bug.cgi?id=9809#c49

If you want to do further changes, please look at what is in current
git master:

https://sourceware.org/git/?p=glibc.git;a=blob_plain;f=localedata/locales/ckb_IQ;hb=refs/heads/master

And then sent a *patch* not the complete new file.

(In reply to Jwtiyar Nariman from comment #79)
> Created attachment 12261 [details]
> fixed typo in wednesday name in kurdish
>
> Just a typo now fixed, replaced U+0624 with U+0627 in the name of Wednesday
> in kurdish.
>
> Best Regards.

$ git diff
diff --git a/localedata/locales/ckb_IQ b/localedata/locales/ckb_IQ
index a18ff69cb7..238c381edf 100644
--- a/localedata/locales/ckb_IQ
+++ b/localedata/locales/ckb_IQ
@@ -124,7 +124,7 @@ abday "<U0634><U06D5><U0645>";/
 day "<U06CC><U06D5><U0643><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U062F><U0648><U0648><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U0633><U06CE><U0634><U06D5><U0645><U0645><U06D5>";/
- "<U0686><U0648><U0624><U0631><U0634><U06D5><U0645><U0645><U06D5>";/
+ "<U0686><U0648><U0627><U0631><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U067E><U06CE><U0646><U062C><U0634><U06D5><U0645><U0645><U06D5>";/
     "<U0647><U06D5><U06CC><U0646><U06CC>";/
     "<U0634><U06D5><U0645><U0645><U06D5>"
lines 1-13/13 (END)

The master branch has been updated by Mike Fabian <email address hidden>:

https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=eb948facd894e66429e2e170043b7d36fe445a8d

commit eb948facd894e66429e2e170043b7d36fe445a8d
Author: Mike FABIAN <email address hidden>
Date: Tue Feb 11 10:17:12 2020 +0100

    Fix typo in the name for Wednesday in Kurdish [BZ #9809]

Fixed in current master.

(In reply to Mike FABIAN from comment #85)
> Fixed in current master.

Thank you dear mike for everything, your help really appreciated.

Best Regards.

HEY dear Mike
Does ckb_IQ will be available in 2.32?
I think there is a problem with re-order as i mentioned before, Due to this commit:
https://sourceware.org/git/?p=glibc.git;a=commit;h=3404def00a1b332080fa51044733f6ead0eae5f3

Best Rgeards

(In reply to Jwtiyar Nariman from comment #87)
> HEY dear Mike
> Does ckb_IQ will be available in 2.32?

Yes.

> I think there is a problem with re-order as i mentioned before, Due to this
> commit:
> https://sourceware.org/git/?p=glibc.git;a=commit;
> h=3404def00a1b332080fa51044733f6ead0eae5f3
>
> Best Rgeards

This is fixed by the mentioned commit.

And even before the commit, it worked, it just printed a warning at build time.

Thank you

On Wed, Jul 22, 2020, 04:39 Maiku-fabian <email address hidden> wrote:

> (In reply to Jwtiyar Nariman from comment #87)
> > HEY dear Mike
> > Does ckb_IQ will be available in 2.32?
>
> Yes.
>
> > I think there is a problem with re-order as i mentioned before, Due to
> this
> > commit:
> > https://sourceware.org/git/?p=glibc.git;a=commit;
> > h=3404def00a1b332080fa51044733f6ead0eae5f3
> >
> > Best Rgeards
>
> This is fixed by the mentioned commit.
>
> And even before the commit, it worked, it just printed a warning at
> build time.
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/266975
>
> Title:
> Request for new language packages for Kurdish Sorani (ckb)
>
> Status in GLibC:
> Fix Released
> Status in Ubuntu Translations:
> Fix Released
> Status in langpack-locales package in Ubuntu:
> Fix Released
>
> Bug description:
> The Kurdish Sorani (ckb) team has made some progress in translating
> packages for Ubuntu in Launchpad. Now they are wondering what is
> necessary to get language packs in that language for Ubuntu.
>
> Is there any chance of this happening in Intrepid, or what has to be
> done to get packages in Intrepid +1?
>
> Here is a link to the page in Launchpad:
> https://translations.launchpad.net/ubuntu/hardy/+lang/ckb
>
> By the way, Kurdish Sorani (ckb) is a Right-to-Left script language
> like Arabic.
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/glibc/+bug/266975/+subscriptions
>

Hey
2.32 is released and ckb is existed, thank you for everyone specially Gunnar and Mike, I wonder to know how Ubuntu will update it to latest 2.32?