MySQL update statement match only the first row - mysql

Question

Here is the my table :
mysql> select * from t1;
+------+-------+
| id | value |
+------+-------+
| 1 | 1 |
+------+-------+
1 row in set (0.00 sec)
mysql> select * from t2;
+------+-------+
| id | value |
+------+-------+
| 1 | 2 |
| 1 | 1 |
| 1 | 2 |
| 1 | 3 |
+------+-------+
4 rows in set (0.00 sec)
Then ,I run a sql to update the date in table t1 for some purpose:
mysql> update t1 join t2 on t1.id=t2.id set t1.value=t2.value ;
Query OK, 0 rows affected (0.00 sec)
Rows matched: 1 Changed: 0 Warnings: 0
And now,see the changes:
mysql> select * from t1;
+------+-------+
| id | value |
+------+-------+
| 1 | 2 |
+------+-------+
1 row in set (0.00 sec)
I wonder that why the rows matched count is 1 ,and it's hardly understand that the column value of t1 has a value 2 where the id=1 rather than 3.Is that update stops when it matches the first row ?
I think it will do a full data match across t1 and t2 in this case.
Any help is appreciated!
update
Thanks,here is the situation that I'm dealing with actully:
For the values in t2 ,concat them seperated by ',' and the merge into the value in table t1 group by each id ,But ,all the element in t1's value should be distinct.For example: as table t1 and t2 list above , after the update operation,the t1's value should be :"1,2,3",neither 2 nor 3 .
if I use the function groupconcat(),It's will be hard to make values to be distinct for t1's value.
Agin,I don't think it's clever to update only on row as in this case.If a update across multi tables ,all the rows matched by the join condition should be updated one by one in a loop.

Answer 1

Based on your update to your question you can do it like this
UPDATE t1 JOIN
(
SELECT id, GROUP_CONCAT(DISTINCT value ORDER BY value) value
FROM t2
GROUP BY id
) q
ON t1.id = q.id
SET t1.value = q.value
Outcome:
+------+-------+
| id | value |
+------+-------+
| 1 | 1,2,3 |
+------+-------+
Here is SQLFiddle demo
UPDATE: Based on your comments which changed your question again. To be able to update a delimited string of values in t1 based on values in t2 you you'll need help of a numbers(tally) table to split t1.value on the fly.
You can easily create such table like this
CREATE TABLE tally(n INT NOT NULL PRIMARY KEY);
INSERT INTO tally (n)
SELECT a.N + b.N * 10 + 1 n
FROM
(SELECT 0 AS N UNION ALL SELECT 1 UNION ALL SELECT 2 UNION ALL SELECT 3 UNION ALL SELECT 4 UNION ALL SELECT 5 UNION ALL SELECT 6 UNION ALL SELECT 7 UNION ALL SELECT 8 UNION ALL SELECT 9) a
,(SELECT 0 AS N UNION ALL SELECT 1 UNION ALL SELECT 2 UNION ALL SELECT 3 UNION ALL SELECT 4 UNION ALL SELECT 5 UNION ALL SELECT 6 UNION ALL SELECT 7 UNION ALL SELECT 8 UNION ALL SELECT 9) b
ORDER BY n
That script creates a table with a sequence of numbers from 1 to 100 which will allow to effectively split up to 100 delimited values. If you need more or less you can easily adjust the script.
Now to update t1.value you can do
UPDATE t1 JOIN
(
SELECT id, GROUP_CONCAT(value ORDER BY value) value
FROM
(
SELECT id, SUBSTRING_INDEX(SUBSTRING_INDEX(t1.value, ',', n.n), ',', -1) value
FROM t1 CROSS JOIN tally n
WHERE n.n <= 1 + (LENGTH(t1.value) - LENGTH(REPLACE(t1.value, ',', '')))
UNION
SELECT id, value
FROM t2
) v
GROUP BY id
) q
ON t1.id = q.id
SET t1.value = q.value
Assuming that you have in t1
| ID | VALUE |
|----|-------|
| 1 | 1,4 |
outcome of the update will be
| ID | VALUE |
|----|---------|
| 1 | 1,2,3,4 |
Here is SQLFiddle demo
That all being said in the long run you better reconsider your db schema and normalize your data. That will pay off big time by allowing normally maintain and query your data.

Answer 2

What a weird query! It seems that MySQL is clever enough to not update the same row 4 times. But besides that, on any database the result (the new value for t1.value) is undefined. You should always make sure that you update with a value of one row, or use an aggregate function (like min, max, ...)