Adding Juraj to CC:.

target OOo1.1Beta2 to align with issue 7043

Reminder: if issue 7043 is to be incorporated for 1.1Beta2 we'd need
the corrected sk_SK.xml until 2003-04-07

I'd like to remind, too: either somebody takes care of this issue now,
or we'll re-target this issue very soon to a later OOo version.

Dan knows that and he also know April 7th is his deadline.
Thank you for reminder.

We already have almost good version. Only small problems are left.

OK here is example of patch between cs_CZ.xml and original in OOo 1.1
beta.

If You will accept it, the Slovak version (sk_SK) will be done in a
few hours. Slovaks have the same typografic rules, only the words are
another.

A few comments:

In Czech You shloud write date like this: 31. 12. 2003, so there are
spaces after dots. This format seems be unparsable by OOo. In
Microsoft products is not posible to insert the space after dot, so
the people are used to write date like this 31.12.2003, without the
spaces. My patch goes now in this direction.

Second one, we never use roman numbers for date. But we use many forms
of the monts, we dont have prepositions like english, we have another
words for this. So there are no longer roman characters and the space
is used for second recent form of months name. Its probably not clean,
but is very usefull for all czechs.

The patch is tested on OOo 1.1 beta and seems work for all normal
operations.

If there are still problems with the date formats, please accept at
least the quotes characters. In OOo are now very bad charactes used
for quotes from our point of view. Its the first part of diff. They
are same for Czech and Slovak language.

Created attachment 5471 [details]
Patch for Czech locales

grabbing issue

Created attachment 5519 [details]
patch against previous patch

I attached another patch. Note that I only changed codes that used
NNNN as that already includes the LongDateDayOfWeekSeparator, and
reordered the date formats merely to ease comparison with the format
indexes lined out in offapi/com/sun/star/i18n/NumberFormatIndex.idl
(about those indexes see also
i18npool/source/localedata/data/locale.dtd rev.1.9.28.1). Also changed
the DefaultName entries to have proper content.

Please try to find matches for the date format codes as close as
possible to their meaning in
offapi/com/sun/star/i18n/NumberFormatIndex.idl, however, you don't
need an exact match, just imagine the same index being used with
another locale shouldn't change meaning significantly.

You used the D.M.YYYY code now at least three times, but there is no
D.M.YY anymore, nor DD.MM.YYYY. Is this really intentional? The
type="short" and type="medium" defaults are now identical with that
change. Note that there is "the one default" calculated by the
application, which is one of types medium, short, or long defaults,
whatever was encountered first in that particular order.

Btw: no, you're right, blanks in between numeric dates aren't always
parseable at the moment. And I really don't know why there were roman
numbers as month abbreviations.

Deadline for changes is 9th April.

Eike,

Dan left us for Japan and will be there for about 14 days. I think we
could stick with the version you have for Beta 2. And Dan will fix it
after he come back.

Is it OK?

That's fine with me. However, we'll close this issue as soon as
integration of this patch will have been verified on the master
branch, to be able to track release issues properly. We should reopen
it or create a 1.1Final issue as soon as a corresponding milestone
target will be available.

I changed the summary to not contain sk_SK anymore, as that wasn't
covered at all.

Fixed on branch cws_srx644_apps61beta2:

i18npool/source/localedata/data/cs_CZ.xml  1.7.44.1

status resolved fixed

reassign to QA

AFAIK found fixed on cws apps61beta2

set to verified

Found integrated on Solaris, Windows and Linux using srx644m13

Created attachment 72210