[Libreoffice] [REVIEW] cherry-picking fix for fdo#38095, character classification for font selection

[Caolán McNamara]

On Tue, 2011-07-19 at 01:14 +0900, Takeshi Abe wrote:
> Hi all,
> 
> According to a discussion on the Japanese local mailing list,
> the following issue may prevent some users from migrating
> to 3.4.x series:
> https://bugs.freedesktop.org/show_bug.cgi?id=38095

Looking for ack-s for 3-4 and 3-4-2.

The goal is to reset the classification of the entire full+half width
range back to ASIAN, i.e. consider the 0-9 and A-z full-width forms as
ASIAN despite their UAX24 classification as Latin, because that's the
way it used to be done and documents change otherwise, and because the
full-width Latin/Common Numerals exist in order to be the same cell
width as full-width CJK chars so they should be rendered with a CJK
font, which is all we want this particular classification to determine
for us.

I also reset back the entire number-forms range to WEAK as well, i.e.
the single-glyph roman-numerals "i", "IV", etc. seeing as they also
really only exist in order to be nicely rendered in CJK vertical text
mode.

http://cgit.freedesktop.org/libreoffice/libs-gui/commit/?id=e76c8d80009c8e29abf0447b7edc157eb42c9e56

looks big, but its basically just if "old explicitly assigned range use
old assignment" else "do probably technically correct thing"

C.