Ashutosh Dhundhara a écrit :

Hi Marc, I think if we will not use "<> 'NEW_VALUE'" in SELECT statement, then it will also return the rows which match the WHERE clause but already have the corresponding column's value = 'NEW_VALUE'. But the UPDATE statement won't count those rows in affected rows where there is already value = 'NEW_VALUE'.

Lets say:

`table_1` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

UPDATE query: UPDATE table_1 SET value = 'value_10' WHERE id = 10; Result: Query OK, 0 rows affected (0.04 sec) Rows matched: 1 Changed: 0 Warnings: 0

SELECT query: SELECT * FROM table_1 WHERE id = 10; Result: 1 row in set.

But this query will yield same result: SELECT * FROM table_1 WHERE id = 10 AND value <> 'value_10';

Result: Empty set (0.00 sec)

Please correct me if I am wrong.

Regards, Ashutosh Dhundhara

Hi Ashutosh, (Please use bottom-posting on this list).

I see what you mean. In your original question, do you have a problem only when UPDATE or SELECT statement involves two tables?

On Friday, 13 June 2014 10:42 PM, Marc Delisle marc@infomarc.info wrote:

Ashutosh Dhundhara a écrit :

...

Hi, I was working on RFE #861 (Simulate UPDATE query). Lets say I have two tables:

`table_1` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

`table_2` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

UPDATE Query: UPDATE table_1, table_2 SET table_1.value = 'NEW_VALUE', table_2.value='NEW_VALUE' WHERE table_1.id > 10 AND table_2.id > 10;

This will affect 2 rows.

How to simulate this query using SELECT statement?

I was trying:

SELECT DISTINCT table_1.value, table_2.value FROM table_1, table_2 WHERE table_1.value <> 'NEW_VALUE' AND table_2.value <> 'NEW_VAUE' AND table_1.id > 10 AND table_2.id > 10;

but this only returns 1 row. Can this be done in a single query only?

Regards, Ashutosh Dhundhara

Hi Ashutosh, any reason why in your SELECT statement, you are using "<> 'NEW_VALUE'" but in your UPDATE statement, you are using "='NEW_VALUE'" ?

---

---

HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions Find What Matters Most in Your Big Data with HPCC Systems Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. Leverages Graph Analysis for Fast Processing & Easy Data Exploration http://p.sf.net/sfu/hpccsystems

---

_______________________________________________ Phpmyadmin-devel mailing list Phpmyadmin-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/phpmyadmin-devel

Le 2014-06-14 12:05, Ashutosh Dhundhara a écrit :

On Saturday, 14 June 2014 12:13 AM, Marc Delisle marc@infomarc.info wrote:

Ashutosh Dhundhara a écrit :

...

Hi Marc, I think if we will not use "<> 'NEW_VALUE'" in SELECT statement, then it will also return the rows which match the WHERE clause but already have the corresponding column's value = 'NEW_VALUE'. But the UPDATE statement won't count those rows in affected rows where there is already value = 'NEW_VALUE'.

Lets say:

`table_1` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

UPDATE query: UPDATE table_1 SET value = 'value_10' WHERE id = 10; Result: Query OK, 0 rows affected (0.04 sec) Rows matched: 1 Changed: 0 Warnings: 0

SELECT query: SELECT * FROM table_1 WHERE id = 10; Result: 1 row in set.

But this query will yield same result: SELECT * FROM table_1 WHERE id = 10 AND value <> 'value_10';

Result: Empty set (0.00 sec)

Please correct me if I am wrong.

Regards, Ashutosh Dhundhara

Hi Ashutosh, (Please use bottom-posting on this list).

I see what you mean. In your original question, do you have a problem only when UPDATE or SELECT statement involves two tables?

Hi Marc,

yes this problem arises when there is a JOIN or more than one table in UPDATE statement. I was wondering if we can tweak the SELECT statement to get the desired results.

Hi Ashutosh, I have no solution for this; maybe this means that the simulation will have to be offered only for one-table queries.

...

On Friday, 13 June 2014 10:42 PM, Marc Delisle marc@infomarc.info wrote:

Ashutosh Dhundhara a écrit :

...

Hi, I was working on RFE #861 (Simulate UPDATE query). Lets say I have two tables:

`table_1` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

`table_2` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

UPDATE Query: UPDATE table_1, table_2 SET table_1.value = 'NEW_VALUE', table_2.value='NEW_VALUE' WHERE table_1.id > 10 AND table_2.id > 10;

This will affect 2 rows.

How to simulate this query using SELECT statement?

I was trying:

SELECT DISTINCT table_1.value, table_2.value FROM table_1, table_2 WHERE table_1.value <> 'NEW_VALUE' AND table_2.value <> 'NEW_VAUE' AND table_1.id > 10 AND table_2.id > 10;

but this only returns 1 row. Can this be done in a single query only?

Regards, Ashutosh Dhundhara

Hi Ashutosh, any reason why in your SELECT statement, you are using "<> 'NEW_VALUE'" but in your UPDATE statement, you are using "='NEW_VALUE'" ?

---

------------------------------------------------------------------------------

...

HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions Find What Matters Most in Your Big Data with HPCC Systems Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. Leverages Graph Analysis for Fast Processing & Easy Data Exploration http://p.sf.net/sfu/hpccsystems

---

_______________________________________________ Phpmyadmin-devel

...

mailing list Phpmyadmin-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/phpmyadmin-devel

On Mon, Jun 16, 2014 at 9:15 PM, Hugues Peccatte hugues.peccatte@gmail.com wrote:

2014-06-13 19:00 GMT+02:00 Ashutosh Dhundhara <ashutoshdhundhara@yahoo.com

...

:

...

Hi, I was working on RFE #861 https://sourceforge.net/p/phpmyadmin/feature-requests/861/ (Simulate UPDATE query). Lets say I have two tables:

`table_1` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

`table_2` +----+----------+ | id | value | +----+----------+ | 10 | value_10 | | 20 | value_20 | +----+----------+

UPDATE Query: UPDATE table_1, table_2 SET table_1.value = 'NEW_VALUE', table_2.value='NEW_VALUE' WHERE table_1.id > 10 AND table_2.id > 10;

This will affect 2 rows.

How to simulate this query using SELECT statement?

I was trying:

SELECT DISTINCT table_1.value, table_2.value FROM table_1, table_2 WHERE table_1.value <> 'NEW_VALUE' AND table_2.value <> 'NEW_VAUE' AND table_1.id > 10 AND table_2.id > 10;

but this only returns 1 row. Can this be done in a single query only?

Regards, Ashutosh Dhundhara

Hi,

Isn't it possible to run the query with set AUTOCOMMIT=0; and a rollback at the end?

Hi,

I doubt that it will work for non innodb databases.

Hugues.