richard.wordingham at ntlworld.com
Thu Mar 6 12:58:53 CST 2014
On Thu, 6 Mar 2014 14:31:38 +0700
Martin Hosken <martin_hosken at sil.org> wrote:
> How would I derive a case mapping from LDML. For example, how would I
> use tr.xml to derive that lc(I)<>dotless i and uc(i)<>dotted cap I? I
> realise there is something deep a mysterious going on in 2.5 level
> collation that is described rather opaquely. Is this where the
> information is? Or do I need to look somewhere else?
I don't believe one is intended to derive this from collation. The
full Lithuanian rules are not derivable from the Lithuanian collation
The simple answer appears to be that the transforms can be found, at
least for CLDR Version 24, in the files:
This is based on looking for the data. I can't work out how to derive
the file names from the LDML Version 24 Part 2 Section 10 or
http://cldr.unicode.org/#TOC-How-to-Use-, which are the locations where
I would look.
http://cldr.unicode.org/index/downloads does give the hint that the
data might be in common/transforms. 'Lower', 'Title' and 'Upper'
appear to be undocumented targets.
More information about the CLDR-Users