[RESOLVED] ICU 4.9 causes a bug with writer (words aren't split correctly)

Caolán McNamara caolanm at redhat.com
Mon May 14 06:37:21 PDT 2012


On Mon, 2012-05-14 at 09:43 +0200, David Tardon wrote:
> Unfortunately this does not work with genbrk from older ICUs (ICU 4.6
> here). 

Well, as always, if a non-default --with-system-foo doesn't work you get
to keep both pieces :-)

> Commit dd49c193de9c4515335ad4a29778ceff225e3c38 attempts to avoid
> the problem (by filtering references to the new character class out).

buildtime regression test should at least ensure that the specific 
Hebrew case works anyway.

>Maybe we should require newer ICU (I hope this does work with 4.8)?

Could make things a little easier with this filtering rules to support multiple
icu versions by keeping multiple rules in the .txt file and tag them
with easier to filter comments e.g. something like...

$AL_FOLLOW_CM   = [$CL $EX $IS...     # ICU 48
$AL_FOLLOW_CM   = [$CL $EX $HL $IS... # ICU 49

and strip them out via their tags

C.



More information about the LibreOffice mailing list