MySQL Bugs: #6611: Wrong sorting order and upper to lower and lower to upper case

Bug #6611	Wrong sorting order and upper to lower and lower to upper case
Submitted:	14 Nov 2004 17:55	Modified:	7 May 2005 8:12
Reporter:	Dubravko Penezic	Email Updates:
Status:	Duplicate	Impact on me:	None
Category:	MySQL Server	Severity:	S2 (Serious)
Version:	4.1.7 and before	OS:	Any (All)
Assigned to:	Assigned Account	CPU Architecture:	Any

Description:
Wrong sorting order for latin2_croatian_ci collation in latin2.xml
Wrong Upper to lower, and lower to upper case in latin2 croatian character SsZz.
Mising collation for croatian character in cp1250.xml

How to repeat:
set table with characetr and order by

Suggested fix:
I have patche files please contact me or ask Arjen G. Lentz for they.

Upper case to lower (using LOWER or LCASE ) in case of a table field simply fails to return correct output.
When used on a string constant, the function returns a correct result.

reproduced using:

select LOWER( 'BOOM' ); -> works
select LOWER( Deal_name ) from produts; -> fails, where Deal_name is TINYTEXT type.

can you please upload the patch in the "Files" section of this bug 
and add a reproducable test case that shows both current and
expected results?

Patche for MySQL version 4.0.x croat.conf & win1250.conf

Attachment: mysql_hr_cs_4.0.x_v101.zip (application/zip, text), 3.16 KiB.

Patch for MySQL 4.1.x latin2.xml, Index.xml & cp1250.xml

Attachment: mysql_hr_cs_4.1.x_v10.zip (application/zip, text), 13.16 KiB.

I put patche for MySQL 4.0.x and for MySQL 4.1.x in Files.

It is character sets related problem do You have any system how in this case create showcase ? Becouse it is some how ASCII set with diferent location and character, and if you don have corect font you wouldnt see some character ?
Picture ? 

I may create web page with show case using 2 DB engine with original files and with thous with implemented patche.

On more version of MySQL and you didnt fix that bug ... I am very dispointed with you ... you ar not able even read document with standard.

The attached files (latin2.xml, cp1250.xml) suggested by Dubravko P. have the correct "latin2_croatian_ci" and "cp1250_croatian_ci" collation mapping. Could you please use them in the upcoming releases.

Here are some links I've found on google for possible help/verification purposes

Alphabet only:
http://developer.mimer.com/collations/charts/croatian.htm

Everything:
http://developer.mimer.com/collations/charts/EOR_croatian.htm

duplicate with 6505