Details
-
Bug
-
Status: Done
-
Low
-
Resolution: Fixed
-
5.6.x, 5.7.x, Not 5.5.x
-
None
Description
**Reported in Launchpad by Sveta Smirnova last update 23-01-2018 13:08:07
Originally reported at https://bugs.mysql.com/bug.php?id=86164
Description:
Fulltext search can not find word which contains punctuation marks like "!?,.{}/" in boolean mode BUT it does in natural language mode.
There is one case reported for "." and "," but according our tests it includes much more punctuation marks.
The reported case is https://bugs.mysql.com/bug.php?id=85876
How to repeat:
The problem can be reproduced as follows:
=================================
CREATE TABLE `test` (
`seq` int(11) NOT NULL AUTO_INCREMENT,
`cont` text COLLATE utf8mb4_bin,
PRIMARY KEY (`seq`),
FULLTEXT KEY `fx_txts` (`cont`) /*!50100 WITH PARSER `ngram` */
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_bin;
insert into test (cont) values ('aBc!efg');
insert into test (cont) values ('aBc@efg');
insert into test (cont) values ('aBc#efg');
insert into test (cont) values ('aBc$efg');
insert into test (cont) values ('aBc%efg');
insert into test (cont) values ('abc^efg');
insert into test (cont) values ('abc&efg');
insert into test (cont) values ('abc*efg');
insert into test (cont) values ('abc(efg');
insert into test (cont) values ('abc)efg');
Then try to select whatever value using boolean mode:
mysql [localhost]
Empty set (0.00 sec)
mysql [localhost] {msandbox}
(test) > select * from test where match(cont) against ('"c#e"' in boolean mode);
Empty set (0.00 sec)
mysql [localhost]
{msandbox} (test) > select * from test where match(cont) against ('"c$e"' in boolean mode);Empty set (0.00 sec)
mysql [localhost] {msandbox}
(test) > select * from test where match(cont) against ('"c!e"' in boolean mode);
Empty set (0.00 sec)
If change to natural language mode it works:
mysql [localhost]
----
| seq | cont |
----
| 1 | aBc!efg |
| 11 | abc!efg |
----
2 rows in set (0.00 sec)
The index entries are created as can be seen in INFORMATION_SCHEMA.INNODB_FT_INDEX_CACHE:
mysql [localhost] {msandbox}
(test) > SELECT * FROM INFORMATION_SCHEMA.INNODB_FT_INDEX_CACHE; ----------------------------------------------------+
WORD | FIRST_DOC_ID | LAST_DOC_ID | DOC_COUNT | DOC_ID | POSITION |
----------------------------------------------------+
!e | 2 | 12 | 2 | 2 | 3 |
!e | 2 | 12 | 2 | 12 | 3 |
#e | 4 | 4 | 1 | 4 | 3 |
$e | 5 | 5 | 1 | 5 | 3 |
%e | 6 | 6 | 1 | 6 | 3 |
&e | 8 | 8 | 1 | 8 | 3 |
(e | 10 | 10 | 1 | 10 | 3 |
)e | 11 | 11 | 1 | 11 | 3 |
*e | 9 | 9 | 1 | 9 | 3 |
@e | 3 | 3 | 1 | 3 | 3 |
^e | 7 | 7 | 1 | 7 | 3 |
<...>
c! | 2 | 12 | 2 | 2 | 2 |
c! | 2 | 12 | 2 | 12 | 2 |
c# | 4 | 4 | 1 | 4 | 2 |
c$ | 5 | 5 | 1 | 5 | 2 |
c% | 6 | 6 | 1 | 6 | 2 |
c& | 8 | 8 | 1 | 8 | 2 |
c( | 10 | 10 | 1 | 10 | 2 |
c) | 11 | 11 | 1 | 11 | 2 |
c* | 9 | 9 | 1 | 9 | 2 |
c@ | 3 | 3 | 1 | 3 | 2 |
c^ | 7 | 7 | 1 | 7 | 2 |
<...>
----------------------------------------------------+
55 rows in set (0.00 sec)
Attachments
Issue Links
- duplicates
-
PS-258 lp1689268 - Fulltext search can not find word which contains punctuation marks
-
- Done
-
- is duplicated by
-
PS-258 lp1689268 - Fulltext search can not find word which contains punctuation marks
-
- Done
-
- is triggering
-
PS-4764 ft_query_extra_word_chars variable needs documentation
-
- Done
-
- mentioned in
-
Page Loading...