Odd entries under Locale's "Available languages"

[humdinger]

This is on alpha4.

There are a few odd entries in the Locale prefs "Available languages" list.

• The first entry is collapsed with no lable. Opening it, "Kamerun" is revealed. Why not a "Kamerun" entry under "K"?
• There's a "Azerbaycanca" entry that's not collapsable; it has no flag. Below it is another "Azerbaycanca (Latin)" with a flag; collapsable.
• There are more examples in that list if you scroll down.

Is that just the way ICU handles stuff, or can it be tweaked?

[pulkomandy]

The data comes from ICU, we just list all the locales and try to build a hierarchical list out of them. The first empty entry is likely missing translation of the language, the others are alternative scripts (cyrillic/latin or so) for the same language, and some special cases such as Esperanto, which has no country. It's not easy to filter them all properly.

A way to improve this is to update to a later version of ICU or CLDR. (CLDR is the locale information used by ICU, it can be updated without changing the lib). The CLDR can be updated separately from the ICU code to avoid too many changes.

[diver]

Seems to be fixed now.

[pulkomandy]

The Kamerun entry is fixed, but there is still the problem of alternative scripts. In the case of Azarbaycan, there is a latin entry and a cyrillic one further down in the list. The "generic" entry would mean "I don't care which script". It could be the parent of all the available scripts, maybe.

[pulkomandy]

Fixed in hrev50091.