Issue 13157 - Wrong sorting of some words
Summary: Wrong sorting of some words
Status: CLOSED FIXED
Alias: None
Product: Calc
Classification: Application
Component: code (show other issues)
Version: OOo 1.1 Beta
Hardware: PC All
: P3 Trivial (vote)
Target Milestone: ---
Assignee: frank
QA Contact: issues@sc
URL:
Keywords:
Depends on:
Blocks: 12943
  Show dependency tree
 
Reported: 2003-04-08 11:42 UTC by josef.cacek
Modified: 2013-08-07 15:12 UTC (History)
1 user (show)

See Also:
Issue Type: DEFECT
Latest Confirmation in: ---
Developer Difficulty: ---

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this issue.
Description josef.cacek 2003-04-08 11:42:00 UTC 
Data->Sort function runs incorrectly on some localized strings:
Results of sorting 2 words: 
Ša, Sa => correct Sa, Ša
Ša, Sb => incorrect Ša, Sb
(it was tested for document language english and czech)
With these strings is bug only in versions 1.0.X
version 1.1Beta runs correctly.

Problem strings in 1.1Beta are Ťa,Ta
Comment 1 frank 2003-04-08 12:23:36 UTC 
Hi Eike,

please have a look at this one.

Frank
Comment 2 ooo 2003-04-08 14:01:26 UTC 
@Frank: Since when do we reassign issues with status unconfirmed?

Collation (sort) algorithms in 1.0.x are not properly
internationalized. Will not be fixed. Use 1.1 instead.

As for kwart's last sentence: I don't even get the meaning.
Comment 3 ooo 2003-04-08 14:41:20 UTC 
close wontfix
Comment 4 josef.cacek 2003-04-09 08:23:23 UTC 
Still the problem in 1.1Beta: T with hatchek and T are not sorted
correctly.
Comment 5 pavel 2003-04-09 09:03:53 UTC 
I confirm this issue.

Try sorting this:

Ta
Tb
T~b
T~a

where T~ is T with hook.
Comment 6 pavel 2003-04-09 09:06:21 UTC 
Changing version and Target.
Comment 7 ooo 2003-04-09 10:57:43 UTC 
Now I get it..

What language is to be sorted? Which language/algorithm did you chose
in the Data.Sort.Options dialog? If Default language, what is the
system configured to?
Comment 8 josef.cacek 2003-04-09 11:05:16 UTC 
Language: default (system: Czech)
tested also with language: Czech
Comment 9 ooo 2003-04-09 12:37:03 UTC 
Hi Karl,

This seems to be one for you. ICU collator for cs_CZ (in fact the
standard Western collator as inherited from en_US) treats \u0164 as
being equal to normal 'T'.

If this turns out to be a bug in ICU itself, please also file a bug
under http://www.jtcsv.com/cgi-bin/icu-bugs

Thanks
  Eike
Comment 10 ooo 2003-04-09 12:39:54 UTC 
reassign to Karl
Comment 11 karl.hong 2003-04-15 01:31:02 UTC 
ICU has multiple collation strength setting, Primary, Secondary, 
Tertiary and so on. We have only two setting, case sensitive and 
insensitive or ignore. We have mapped case sensitive to Tertiary, 
which distinguishs the Upper/Lower case and is correct. But we 
wrongly mapped case insensitive to Primary, which not only ignore 
case, but also ignore accent, so 'T' and 'T-caron' are treated 
equally when user select case insensitive. I have changed it to 
Secondary, which only ignore case.
Comment 12 frank 2003-04-25 13:30:14 UTC 
AFAIK found fixed on cws Apps61Beta2 using Windows and Solaris
Comment 13 karl.hong 2003-04-29 18:37:35 UTC 
Karl: Send the bug to QA.
Comment 14 frank 2003-04-30 12:19:58 UTC 
reatoring fixed state
Comment 15 frank 2003-04-30 12:21:02 UTC 
restoring verified state
Comment 16 frank 2003-04-30 13:38:42 UTC 
found integrated on Linux, Solaris and Windows