This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH v2] Sync "language", "lang_name", "territory", "country_name"


Florian Weimer <fw@deneb.enyo.de> さんはかきました:

> * Carlos O'Donell:
>
>> On 10/25/19 5:10 PM, Florian Weimer wrote:
>>> * Carlos O'Donell:
>>> 
>>>> On 9/18/19 11:19 AM, Mike FABIAN wrote:
>>>>> According to feedback from Rafał Lużyński, I made
>>>>>
>>>>>     country_name "Polska"
>>>>>
>>>>> instead of "Polsko" in szl_PL.
>>>>>
>>>>> The rest of the patch is unchanged. New patch attached.
>>>>
>>>> Great work here.
>>>>
>>>> Is there any way we can make the sync automated?
>>> 
>>> I recommend against that because it has introduced regressions in the
>>> past.  See bug 15601.
>>
>> Bug 15601 is an instance where we should be working to fix CLDR, and
>> update a sync scripts to skip that name and document the URL of the
>> upstream CLDR bug tracking the harmonization issue.
>>  
>> I don't want the sync to happen in a cron job. I just want the ability
>> to sync via a script, and have a human review the results for possible
>> regressions and issues. I want to avoid manual updates.

I did it with a script. I was not sure whether that script is ready for
inclusion into glibc. It is here:

https://github.com/mike-fabian/langtable/blob/master/tools/compare_with_glib_source.py

Quite ugly, just a quick hack.

I did run that and looked at the results and then changed in glibc only
what seemed reasonable to me.

>> Does that clarify my position?
>
> Yes, that makes sense.  Thanks.

-- 
Mike FABIAN <mfabian@redhat.com>
睡眠不足はいい仕事の敵だ。


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]