Update language data for Sinitic languages

[winstonsung]

Bug: #10488
Bug: openstreetmap/id-tagging-schema#1356
Change-Id: I0ca3cd8b2a82a7edeff70b4be7678e0699214ccc

[1ec5]

This looks like a good set of changes to me at a glance. Have you verified that the associated name:* tags are well-established, so that we don’t end up promoting tags that the language community is unaware of? Also, we may need to update our copy of wmf-sitematrix so that the Wikipedia field’s language selector and Wikidata field’s label functionality reflects recent changes to Wikimedia’s language list.

[winstonsung]

Have you verified that the associated name:* tags are well-established, so that we don’t end up promoting tags that the language community is unaware of?

The language tags were well-established and being used in MediaWiki and on Wikimedia wikis.

• The variant subtag pehoeji, tailo were supported by the Hokkien community, proposed and added to Language Subtag Registry on 2024-03-04.
• The variant subtag pinyin was already existed in Language Subtag Registry since 2008-10-14, but not using the standard BCP 47 language tag on OpenStreetMap (so we need to fix it now).

https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry

Also, we may need to update our copy of wmf-sitematrix so that the Wikipedia field’s language selector and Wikidata field’s label functionality reflects recent changes to Wikimedia’s language list.

I'll take a look into it.

[winstonsung]

Also, we may need to update our copy of wmf-sitematrix so that the Wikipedia field’s language selector and Wikidata field’s label functionality reflects recent changes to Wikimedia’s language list.

The list is up-to-date as far as I can see.

[1ec5]

The language tags were well-established and being used in MediaWiki and on Wikimedia wikis.

I was referring to the OSM database. In theory, the name:* tagging scheme relies on BCP 47, but there have been exceptions in the past, not necessarily the same exceptions as in Wikimedia’s wiki farm. I see that name:zh_pinyin=* has already been mass-replaced with name:zh-Latn-pinyin=*, and name:zh-Latn-pinyin=* is now documented. So that’s good – the momentum is already on the side of the correct code, and iD is just responding to that change. openstreetmap/id-tagging-schema#1356 is another good step.

The list is up-to-date as far as I can see.

It looks like iD is currently using wmf-sitematrix v0.1.x, which means it’ll pull in v0.1.5. We’re overdue for a new release of that package, which will probably be versioned as v0.2.0 due to build system changes, so a small tweak to this code will be necessary once that happens:

iD/modules/core/file_fetcher.js

Line 42 in 72643ba

'wmf_sitematrix': wmfSitematrixCdnUrl.replace('{version}', '0.1') + 'wikipedia.min.json'

It looks like iD is no longer using wmf-sitematrix for anything that also uses the built-in language data, so we can take care of that change in a separate PR.

[1ec5]

It looks like iD is no longer using wmf-sitematrix for anything that also uses the built-in language data, so we can take care of that change in a separate PR.

#10489