Hi Søren,

On Sat, May 13, 2006 at 11:16:56 +0200, Søren Thing Pedersen wrote:

> I have uploaded a patch to 
> http://qa.openoffice.org/issues/show_bug.cgi?id=65342

Thanks.

> I have no idea where to look for the sorting order in ICU (CLDR?). I saw 
> some syntax I didn't understand in core.zip. Can you point me to where I 
> should look?

In CLDR's core.zip the collation rules are expressed as XML elements.
A representation that fits with ICU's collation rules is available at
http://www.unicode.org/cldr/data/diff/collation/da_DK.html

For a description of ICU's collation rules see
http://icu.sourceforge.net/userguide/Collate_Concepts.html
http://icu.sourceforge.net/userguide/Collate_ServiceArchitecture.html

While ICU now uses CLDR to generate it's locale data, the ICU 2.6 data
still lives on its own and is under icu/source/data/locales of the
unpacked icu archive, in the OOo build tree it is
icu/$INPATH/misc/build/icu/source/data/locales, with
$INPATH==unxlngi6.pro,wntmsci10.pro,...

The Danish locale in ICU 2.6 is icu/source/data/locales/da.txt, where
you see "& V<<< w<<< W" in the CollationElements. That's the "culprit".
Funny enough, the documentation at
http://icu.sourceforge.net/userguide/Collate_Customization.html
lists Danish as an example:

| Danish: &V <<< w <<< W
| The letter 'W' is sorted after 'V', but is treated as a tertiary difference
| similar to the difference between 'v' and 'V'. 

However, according to the latest ICU locale explorer using ICU 3.4 this
treatment of W was removed, you may see and try out at
http://www.icu-project.org/cgi-bin/locexp/?_=da&x=col
So when upgrading to ICU 3.4 for OOo2.0.4 this should get automatically
corrected.

  Eike

-- 
 PGP/OpenPGP/GnuPG encrypted mail preferred in all private communication.
 Key ID: 0x293C05FD - 997A 4C60 CE41 0149 0DB3  9E96 2F1A D073 293C 05FD

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to