Re: [PATCH] Locales: Cyrillic -> ASCII transliteration table [BZ #2872] v2

[Volodymyr Lisivka <vlisivka@gmail.com>]

чт, 11 жовт. 2018 о 14:05 Rafal Luzynski <digitalfreak@lingonborough.com> пише:
>
> Thank you, Egor.  I am looking at your patch and although I have
> not yet finished, here are some remarks:
>
> First of all, I think that such a large patch should also include
> the tests.  Please see how automatic tests are performed in locale
> data and write your own.
>
> 11.10.2018 00:29 Egor Kobylkin <egor@kobylkin.com> wrote:
> > [...]
> > From this patch I have excluded locales that already mention cyrillic or
> > have a transliteration table for it:
> > az_AZ
> > iso14651_t1_common
> > ky_KG
> > mn_MN
> > sr_RS
> > tg_TJ
> > tk_TM
> > tt_RU
> > uk_UA
> > uz_UZ
> > uz_UZ@cyrillic
> > [...]
>
> I think that eventually we would like to include your translit_cyrillic
> also in these locales because I assume that your rules should work good
> for them as well, also should include more characters than the individual
> language contributors took into account.

It's very good idea. Transliteration in Ukrainian locale predates this
work for about decade. It well tested. I also have automatic test
cases, which I can adapt to current standard. Let's drop Russian
transliteration rules and replace them with Ukrainian transliteration
rules. I assume that Ukrainian rules should work good for them as
well.

Ukrainian language is the oldest and most developed language in Slavic
family - last king of all Slavs named Madzhak/Muzhik (Brave), leader
of Volyniana union, was lived in Western Ukraine in Volyn` region.
After Madzhak capturing of Madzhak, kingdom was split into multiple
western parts and eastern part, where 9 Slavic tribes were united by
Rus` tribe, which abandoned their city, now known as Old Russa,
because of epidemic. IMHO, it's will be fair to use rules of the
oldest Slavic union.

> Similarly to Mike's work on
> collation: a common rules were created and all locales include them adding
> their own language specific modifications.

It's good idea too. In our own locale we prefer that words in our
language will be at top of a sorted list. Currently, in Ukrainian
locale it works as intended, but Russian locale has inverted order.
IMHO, Russian locale should use Ukrainian rules.

$ echo 'один два three four'| tr ' ' '\n' | LANG=uk_UA.utf8 sort
два
один
four
three
$ echo 'один два three four'| tr ' ' '\n' | LANG=ru_RU.utf8 sort
four
three
два
один


>
> > [...]
> > COMMIT MESSAGE:
> > [...]
> > I am excluding these locales from this proposed patch. I have written
> > directly to locale maintainer emails listed in the files. Volodymyr
> > Lisivka <vlisivka@gmail.com>, Max Kutny <mkutny@gmail.com> (uk_UA),
> > Данило Шеган <danilo@gnome.org> (sr_YU, sr_CS) have confirmed the
>
> I am not sure if we want Cyrillic text in the commit message.  Shouldn't
> it be, uhm, tranlisterated? :-)
>
> "sr_CS" - I guess you meant "sr_RS".
>
> "sr_YU" has been dropped, do we want to mention it?
>
> > [...]
> > [BZ #2872]
> > * localedata/locales/translit_cyrillic: add ISO 9.1995, GOST 7.79
>
> Please start "Add" with an uppercase.  BTW, shouldn't it be "New file"
> instead?
>
> > System A transliteration System B transcription table from Cyrillic to
> > Latin/ASCII.
> > * localedata/locales/C: add include "translit_cyrillic";"" to LC_CTYPE
> > translit section.
>
> Same, "Add" here.
>
> > * localedata/locales/aa_DJ: Likewise.
>
> Good (here and everywhere below).
>
> > [...]
> > diff -uNr a/localedata/locales/translit_cyrillic
> > b/localedata/locales/translit_cyrillic
> > --- a/localedata/locales/translit_cyrillic 1970-01-01 00:00:00.000000000
> > +0000
> > +++ b/localedata/locales/translit_cyrillic 2018-10-09 19:02:54.000000000
> > +0000
> > @@ -0,0 +1,383 @@
> > +escape_char /
> > +comment_char %
> > +
> > +% This file is part of the GNU C Library and contains locale data.
> > +% The Free Software Foundation does not claim any copyright interest
> > +% in the locale data contained in this file. The foregoing does not
> > +% affect the license of the GNU C Library as a whole. It does not
> > +% exempt you from the conditions of the license if your use would
> > +% otherwise be governed by that license.
> > +
> > +% Transliterations of cyrillic letters to latin and/or ascii symbols.
>
> "cyrillic" -> "Cyrillic"; "latin" -> "Latin"; "ascii" -> "ASCII".
>
> > +% Inspired by ISO 9.1995 / GOST 7.79-2000.
> > +% Covers Unicode Range https://www.unicode.org/charts/PDF/U0400.pdf
> > +% i.e [U4001-U4F9, U2019] but only the letters covered by ISO 9.1995
>
> Typos:
>
> "i.e" -> "i.e.," (somebody please fix me if I'm wrong here)
> "U4001" - I guess you meant "U0401"
> "U4F9" -> "U04F9".  I think that "U4F9" is not definitely bad but
> let's be consistent.
>
> Also I can see some gaps in the range.  Are you going to fill them
> or maybe for now just mention that they exist?
>
> > +% It implements the GOST_7.79 System A (Latin Script) as a first
> > +% option and System B Cyrillic (ASCII) as a second option. Check
> > +% https://en.wikipedia.org/wiki/ISO_9 for reference.
> > +% The System B is extended from GOST_7.79-Russian using open sources
> > +% of the transliteration mappings and the "h/`" diacritics logic.
>
> What is "h/`" diacritics logic?
>
> > +
> > +% Usage examples:
> > +% iconv -f UTF-8 -t ISO-8859-15//TRANSLIT \
> > +% | iconv -f ISO-8859-15 -t UTF-8 # System A
> > +% iconv -f UTF-8 -t ASCII//TRANSLIT # System B.
> > +
> > +% Contributions welcome for the rest of Cyrillic script in Unicode
>
> Sure, I'm not going to stop you from pushing these changes just because
> there are missing characters.  I will consider adding them later.
>
> > +% https://en.wikipedia.org/wiki/Cyrillic_script_in_Unicode.
> > +% Bugfix for https://sourceware.org/bugzilla/show_bug.cgi?id=2872.
> > +% Generated from UnicodeData.txt with
> > +% https://sourceware.org/bugzilla/attachment.cgi?id=11301.
>
> 1. Is the file really generated with a script and not modified later?
> If yes then maybe you should contribute the script instead?  In that case,
> you should also not post this file to libc-locale, maintainers and
> developers should be able to regenerate it.
> 2. The link leads to a LibreOffice spreadsheet.
>
> > +LC_CTYPE
> > +
> > +translit_start
> > +
>
> <U0400> is missing here.  Are you going to leave it for now?
>
> > +% CYRILLIC CAPITAL LETTER IO
> > +<U0401> <U00CB>;"<U0059><U004F>"
> > [...]
> > +% CYRILLIC CAPITAL LETTER KJE
> > +<U040C> <U1E30>;"<U004B><U0060>"
>
> <U040D> is missing here.  Can we add it already?
>
> > +% CYRILLIC CAPITAL LETTER SHORT U
> > +<U040E> <U016C>;"<U0055><U0060>"
> > [...]
> > +% CYRILLIC CAPITAL LETTER U
> > +<U0423> <U0055>
> > +% CYRILLIC UNDEFINED
> > +<U0423><U0301> <U00DA>;"<U0055><U0060>"
>
> This still makes me wonder.
>
> Does it work at all?
> What if we remove this rule, won't it be transliterated as
> <U0423> => "U", <U0301> - left unchanged, so "U" + <U0301>"
> will eventually produce "Ú"?
> Why is it called "UNDEFINED"?
> Do we need similar rules for other characters?
>
> > [...]
> > +% CYRILLIC SMALL LETTER U
> > +<U0443> <U0075>
> > +% CYRILLIC UNDEFINED
> > +<U0443><U0301> <U00FA>;"<U0075><U0060>"
>
> Same here.
>
> > [...]
> > +% CYRILLIC SMALL LETTER YA
> > +<U044F> <U00E2>;"<U0079><U0061>"
>
> Again <U0450> missing (because it is lowercase variant of <U0400>).
>
> > +% CYRILLIC SMALL LETTER IO
> > +<U0451> <U00EB>;"<U0079><U006F>"
> > [...]
> > +% CYRILLIC SMALL LETTER KJE
> > +<U045C> <U1E31>;"<U006B><U0060>"
>
> <U045D> missing (same reason as <U040D>).
>
> > +% CYRILLIC SMALL LETTER SHORT U
> > +<U045E> <U016D>;"<U0075><U0060>"
> > +% CYRILLIC SMALL LETTER DZHE
> > +<U045F> "<U0064><U0302>";"<U0064><U0068>"
>
> More letters missing here.  Is this because they are historic so we
> don't want to include them now?  Well, but "YUS" is also historic.
> (Please, do not remove YUS for consistency).
>
> > +% CYRILLIC CAPITAL LETTER BIG YUS
> > +<U046A> <U01CD>;"<U004F><U0060>"
> > +% CYRILLIC SMALL LETTER BIG YUS
> > +<U046B> <U01CE>;"<U006F><U0060>"
> > [...]
>
> I will continue but, again, I don't give any ETA so other reviewers
> are welcome here.
>
> Regards,
>
> Rafal