Handle unicode variant subtags on likelySubtags and removeLikelySubtags
rxaviers opened this issue · comments
Since variant subtags are not being handled on likelySubtags, the minlanguageId of en-US-POSIX
is wrongly en
instead of en-POSIX
leading to wrong bundle lookup. Similarly for maxLanguageId.
@rxaviers Any workarounds for this? I'm getting bitten by this issue since we're loading the full data set, and my en
numbers aren't being formatted correctly. (Note: I am using the setAvailableBundles 'hack' as we've previously discussed.)
This is supposed to be an easy fix... Although, it will take me a while to work on it. Help appreciated. Basically, the files that need fix are likelySubtags and removeLikelySubtags (with special attention to the instruction of the lines I've highlighted). This may also help: subtags. Both likelySubtags and removeLikelySubtags are used in bundle loopup (root of the problem) and also in init.
Tests need to be included to assert this fix.
@patch, FYI this is fixed.
@rxaviers
Released in 0.4.4.