XPost: linux.debian.www

This is a MIME-formatted message. If you see this text it means that your E-mail software does not support MIME-formatted messages.

Hello,
On Mon, Nov 20, 2023 at 01:41:04PM +0100, Laura Arjona Reina wrote:

El 20/11/23 a las 10:43, Thomas Lange escribió:

I'm still not sure which of the languages we need or which are just
bugs in packages. Or do we have bug in the scripts, that generat this language list?
What about the AA_BB and AA@somestring languages?

For e.g. I wonder why we have international/l10n/po/man_DE
which links only to this po file: https://i18n.debian.org/material/po/unstable/main/i/i2p/installer/resources/locale-man/i2p_0.9.48-1.1_man_de.po.gz

This po file cleary says
"Language: de\n"
Why is the language then called man_DE and not just "de"? Is this a
bug in our scripts?

I think this is a bug, because those po files are named man_ because they
are translations of the manual pages, not because of the Mandingo language.

"Similar" thing happens with https://www.debian.org/international/l10n/po/bos_DE (and all other bos_XX links in https://www.debian.org/international/l10n/po/): they are translations files of the boswars package, having two translations
templates, one named     boswars_version_xx.po.gz and another one named  boswars_version_bos_xx.po.gz, so that "bos" is misdetected as Bosnian language.

Ideally these would be fixed upstream, but 2nd best option would be to
filter them, i.e. if it is invalid combination (as shown) then simply
simply present it unter "DE" (in your example).

This, howeve, would require a list of all valid combination. I don't
know if such a list exists.

Greetings

Helge

--
Dr. Helge Kreutzmann debian@helgefjell.de
Dipl.-Phys. http://www.helgefjell.de/debian.php
64bit GNU powered gpg signed mail preferred
Help keep free software "libre": http://www.ffii.de/

-----BEGIN PGP SIGNATURE-----

iQIzBAABCAAdFiEEbZZfteMW0gNUynuwQbqlJmgq5nAFAmVcOZ0ACgkQQbqlJmgq 5nAoHQ//Tq4Z7akJeSNDynAUxdmfPgvnnwf9itrBLr3fYgkLc8C4318HfI36cINA yqUjfR6dElDQ3gTyOYk9bf3ri2VvJEGmta1CjjNiwFS6aPtCk97HDGnaV7pNpask 3i+hdKVTOeWtS6DilbcfUj3xNQ8bPDjJk5Am4Mrk6xjx5E6Tb2FgX6cBmDvqifAn gSvqHQFkjmxssg80IVGRqAyoHja4vaEoDal3iZhqnKbOVcOsqor3zm5Nu779cN4l C7ngikRKGuBGDA3v0hukBPGeuiY0Yx7SeBNboCfyS+Q5Z0LcJ8zF5LUWanz+VJbT knWD1f7Swf6ZXSzxrPKIUkAT21VRktZlNi7m2h1tTof1Gsc0kDBhvGdT4ie53SFq kYPB0NNUQltrLTu2YkVueVi3pM4kvQ9jVgr6p2BEIbaqBtbABRe0O20A6QJTjEq0 OhraX5WE2y1oupVENJd7SG9bWjKfEpjqwxuSFKf2axEYfGNThi74dSQHPJeNCHNW F8t26W1cOQN0ys8ET+haebhilc64fvnnQesLtuUJR5v5+Sxr04oa4tfSBFCPISZQ 614OJpBjyv92IuBDQEw4yho7u5ew0OEAT+AJdkG

XPost: linux.debian.www

Hi all,

thanks for all the feedback. It seems that there's more work to do
than I've expected. Currently I will not work on fixing the po
statistics, because I want to concentrate on the security pages.
So, if anyone else likes to work on this, feel free to do so.

--
regards Thomas

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

XPost: linux.debian.www

Hi,

Thomas Lange <lange@cs.uni-koeln.de> wrote (Thu, 23 Nov 2023 13:39:16 +0100):

thanks for all the feedback. It seems that there's more work to do
than I've expected. Currently I will not work on fixing the po
statistics, because I want to concentrate on the security pages.
So, if anyone else likes to work on this, feel free to do so.

I have a tested patch here, which allows to strip out some of the language entries out of the 'langs' list, so those entries do no longer show up at https://www.debian.org/international/l10n/po/ and the respective html
pages are not built as well.

For the first step I have chosen to delete most of the "Unknown language" entries and some more, which are false-positives because of wrongly
named files/directories in packages.
It workes fine here locally.

The functionality is as the following:
When make is executed in the ./english/international/l10n/po directory,
the gen-files.sh script is called, which builds the list of languages
out of the material from https://i18n.debian.org/material/.
This list of languages is written to ./english/international/l10n/data/langs.

I have now created a script named strip-langs.sh, which removes entries
by choice from this langs file (you can define in this script, which
entries to remove).
The strip-langs.sh is called after the gen-files.sh run.
And then the html files for all the left languages are built.

With this, we would have an infrastructure, to remove single language
entries from the langs list choice by choice.

In additional steps, we could also remove entries like de_AT, de_CH, de_DE
and only leave 'de', as mentioned by Thomas.
But that's another discussion.


I have filed this as a merge request: https://salsa.debian.org/webmaster-team/webwml/-/merge_requests/947

Holger


--
Holger Wansing <hwansing@mailbox.org>
PGP-Fingerprint: 496A C6E8 1442 4B34 8508 3529 59F1 87CA 156E B076

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

XPost: linux.debian.www

Hi Holger,

this script looks good to me. Thanks a lot for your work on this.
I think you can merge it.

On Mon, 1 Jan 2024 23:39:09 +0100, Holger Wansing <hwansing@mailbox.org> said:

> In additional steps, we could also remove entries like de_AT, de_CH, de_DE
> and only leave 'de', as mentioned by Thomas.
> But that's another discussion.
I guess we do not want to remove those languages, but merge the
variants into de. But maybe this need to be done in a different
script, maybe in gen-files.pl.

--
regards Thomas

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)