| Bug #58194 | LIKE on indexed column over custom utf8 collation fails on partial match | ||
|---|---|---|---|
| Submitted: | 15 Nov 2010 8:55 | Modified: | 17 Dec 2010 8:51 |
| Reporter: | Nicolas Kratz | Email Updates: | |
| Status: | No Feedback | Impact on me: | |
| Category: | MySQL Server: Charsets | Severity: | S3 (Non-critical) |
| Version: | 5.0 | OS: | Linux |
| Assigned to: | CPU Architecture: | Any | |
| Tags: | utf8 uca collation index | ||
[17 Nov 2010 8:51]
Susanne Ebrecht
As I already pointed out on IRC .. I am not able to repeat this with our 5.0 packages on Ubuntu. This really seems to be a Debian/Ubuntu package problem. Did you already got feedback from Debian?
[18 Dec 2010 0:00]
Bugs System
No feedback was provided for this bug for over a month, so it is being suspended automatically. If you are able to provide the information that was originally requested, please do so and change the status of the bug back to "Open".
Description: Using LIKE '$string%' on a char or varchar column - containing duplicate values - using a custom utf8 collation - having an index which is actually used does not return partial matches. Additional observations: - The actual collation definition does not seem to matter, as long as it is valid and contains at least one rule. - It happens with MyISAM and InnoDB. - Forcing a table scan yields the expected results - Bug observed on versions: 5.0.51a-debug (built from source), 5.0.51a-24+lenny4-log, 5.0.75-0ubuntu10.5, aka all 5.0 versions I could lay my hands on. - Bug not observed on various 5.1 linux and mac desktops scattered across the office. How to repeat: create table t1 ( id int unsigned primary key auto_increment, t varchar(64) charset utf8 collate utf8_custom_ci, index t_idx (t) ); insert into t1 (t) values ('test'),('testcase'),('testcaselonger'); insert into t1 (t) values ('test'),('testcase'),('testcaselonger'); select * from t1 where t like 'testcase%'; Result: +----+----------+ | id | t | +----+----------+ | 2 | testcase | | 5 | testcase | +----+----------+ 2 rows in set (0.00 sec) Expected: +----+----------------+ | id | t | +----+----------------+ | 2 | testcase | | 3 | testcaselonger | | 5 | testcase | | 6 | testcaselonger | +----+----------------+ 4 rows in set (0.00 sec)