You may want to use a case sensitive collation. I believe the default is case insensitive. Example:
CREATE TABLE my_table (
id int,
name varchar(50)
) CHARACTER SET latin1 COLLATE latin1_general_cs;
INSERT INTO my_table VALUES (1, 'SomeThing');
INSERT INTO my_table VALUES (2, 'something');
INSERT INTO my_table VALUES (3, 'SOMETHING');
INSERT INTO my_table VALUES (4, 'SOME4THING');
Then:
SELECT * FROM my_table WHERE name REGEXP '^[A-Z]+$';
+------+-----------+
| id | name |
+------+-----------+
| 3 | SOMETHING |
+------+-----------+
1 row in set (0.00 sec)
If you don't want to use a case sensitive collation for the whole table, you can also use the COLLATE
clause as @kchau suggested in the other answer.
Let's try with a table using a case insensitive collation:
CREATE TABLE my_table (
id int,
name varchar(50)
) CHARACTER SET latin1 COLLATE latin1_general_ci;
INSERT INTO my_table VALUES (1, 'SomeThing');
INSERT INTO my_table VALUES (2, 'something');
INSERT INTO my_table VALUES (3, 'SOMETHING');
INSERT INTO my_table VALUES (4, 'SOME4THING');
This won't work very well:
SELECT * FROM my_table WHERE name REGEXP '^[A-Z]+$';
+------+-----------+
| id | name |
+------+-----------+
| 1 | SomeThing |
| 2 | something |
| 3 | SOMETHING |
+------+-----------+
3 rows in set (0.00 sec)
But we can use the COLLATE
clause to collate the name field to a case sensitive collation:
SELECT * FROM my_table WHERE (name COLLATE latin1_general_cs) REGEXP '^[A-Z]+$';
+------+-----------+
| id | name |
+------+-----------+
| 3 | SOMETHING |
+------+-----------+
1 row in set (0.00 sec)
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…