Search the Community

In MySQL, I want to create a trigger with AFTER UPDATE triggering event for my "user" table. In next, I have a table named "user_log" which is use to store the modifications that occurred on the parent table "user" after any update commands. So, data in "user_log" table need be as follows: select * from user_log; +--------+---------+----------------------------+---------------+----------------+---------------------+------+ | log_id | user_id | action | old_data | new_data | changed_date | read | +--------+---------+----------------------------+---------------+----------------+---------------------+------+ | 1 | 10 | Changed yyy's name | yyy | xxx | 2022-06-20 14:06:56 | no | | 2 | 10 | Changed xxx's address | No.111, | No.112, | 2022-06-20 19:07:38 | no | | 3 | 10 | Changed xxx's city | Old City Name | New City Name | 2022-06-20 19:07:38 | no | | 4 | 10 | Changed xxx's phone number | 011-5000000 | 011-4000000 | 2022-06-20 19:07:38 | no | +--------+---------+----------------------------+---------------+----------------+---------------------+------+ As you can see from the data in the table above, it will update several columns at once. So I created my triger as follows, and its working for me. DELIMITER $$ DROP TRIGGER IF EXISTS `user_log` ; $$ CREATE TRIGGER `user_log` AFTER UPDATE ON `user` FOR EACH ROW BEGIN IF OLD.name <> NEW.name THEN INSERT INTO user_log (user_id,action,old_data,new_data) VALUES( NEW.user_id , CASE WHEN (NEW.name <> OLD.name) THEN CONCAT('Changed ', OLD.name, "'s ", 'name') ELSE '' END , CASE WHEN (NEW.name <> OLD.name) THEN OLD.name ELSE '' END , CASE WHEN (NEW.name <> OLD.name) THEN NEW.name ELSE '' END ); END IF; IF OLD.address <> NEW.address THEN INSERT INTO user_log (user_id,action,old_data,new_data) VALUES( NEW.user_id , CASE WHEN (NEW.address <> OLD.address) THEN CONCAT('Changed ', OLD.name, "'s ", 'address') ELSE '' END , CASE WHEN (NEW.address <> OLD.address) THEN OLD.address ELSE '' END , CASE WHEN (NEW.address <> OLD.address) THEN NEW.address ELSE '' END ); END IF; IF OLD.city <> NEW.city THEN INSERT INTO user_log (user_id,action,old_data,new_data) VALUES( NEW.user_id , CASE WHEN (NEW.city <> OLD.city) THEN CONCAT('Changed ', OLD.name, "'s ", 'city') ELSE '' END , CASE WHEN (NEW.city <> OLD.city) THEN OLD.city ELSE '' END , CASE WHEN (NEW.city <> OLD.city) THEN NEW.city ELSE '' END ); END IF; IF OLD.phone <> NEW.phone THEN INSERT INTO user_log (user_id,action,old_data,new_data) VALUES( NEW.user_id , CASE WHEN (NEW.phone <> OLD.phone) THEN CONCAT('Changed ', OLD.name, "'s ", 'phone number') ELSE '' END , CASE WHEN (NEW.phone <> OLD.phone) THEN OLD.phone ELSE '' END , CASE WHEN (NEW.phone <> OLD.phone) THEN NEW.phone ELSE '' END ); END IF; END$$ DELIMITER ; My problem is, I have a lot more columns in the user table. Like I said, all columns or several of them are updated at once. In that case I have to add a large amount of INSERT query to my trigger. So here I would like to know if there is another suitable way to do this. I also tried it in this way. But its working only for one column. DROP TRIGGER IF EXISTS `user_log`; CREATE TRIGGER IF NOT EXISTS `user_log` AFTER UPDATE ON user FOR EACH ROW INSERT INTO user_log (user_id,action,old_data,new_data) VALUES ( NEW.user_id , CASE WHEN (NEW.name <> OLD.name) THEN CONCAT('Changed ', OLD.name, "'s ", 'name') WHEN (NEW.address <> OLD.address) THEN CONCAT('Changed ', OLD.name, "'s ", 'address') WHEN (NEW.city <> OLD.city) THEN CONCAT('Changed ', OLD.name, "'s ", 'city') WHEN (NEW.phone <> OLD.phone) THEN CONCAT('Changed ', OLD.name, "'s ", 'phone number') ELSE '' END , CASE WHEN (NEW.name <> OLD.name) THEN OLD.name WHEN (NEW.address <> OLD.address) THEN OLD.address WHEN (NEW.city <> OLD.city) THEN OLD.city WHEN (NEW.phone <> OLD.phone) THEN OLD.phone ELSE '' END , CASE WHEN (NEW.name <> OLD.name) THEN NEW.name WHEN (NEW.address <> OLD.address) THEN NEW.address WHEN (NEW.city <> OLD.city) THEN NEW.city WHEN (NEW.phone <> OLD.phone) THEN NEW.phone ELSE '' END ); Thank you.

I'm new to using triggers and really not sure how to proceed. I have an access table that a row is inserted whenever someone opens the website or downloads a file with the following structure: `ID` int(5) NOT NULL autoincrement, `LOG_TIME` datetime NOT NULL DEFAULT current_timestamp(), `IP_ADDRESS` int(64) unsigned COLLATE utf8_general_mysql500_ci NOT NULL, `FILENAME` varchar(50) COLLATE utf8_general_mysql500_ci DEFAULT NULL, `country` varchar(50) COLLATE utf8_general_mysql500_ci DEFAULT NULL, `area` varchar(50) COLLATE utf8_general_mysql500_ci DEFAULT NULL, `city` varchar(50) COLLATE utf8_general_mysql500_ci DEFAULT NULL ) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_general_mysql500_ci; When the site is accessed, a row is inserted with only the first 3 fields are filled. When a file is downloaded another row is inserted with the first 4 fields filled. I want to create a trigger such that when a download row is inserted, the IP_ADDRESS is compared to another table to update the country, area, and city fields. I can currently do that for the whole table in mysql shell using the following code: UPDATE access t2, ip_lookup t1 SET t2.country = t1.country, t2.area = t1.area, t2.city = t1.city WHERE ((t2.IP_ADDRESS) BETWEEN (t1.start_ip) AND (t1.end_ip)) AND t2.FILENAME is not null and t2.country is null; How would I write an "after insert" trigger to update the last 3 fields based on the ip of the row that was inserted because of a download? Thanks in advance, Larry

I have a trigger, CREATE TRIGGER `TRIGGER_SOMETABLE_UID` BEFORE INSERT ON `SOMETABLE` FOR EACH ROW SET NEW.Uid = UUID( ) Now, I thought this would fire FOR EACH ROW, but if I'm doing an insert like this... insert into sometable ( col1, col2 ) ( select col1, col2 from SOME_OTHER_tABLE ) ... and that inserts two rows, the result is that Uid in both new rows is identical. This is not the desired behavior. Can you tell me how to get a unique Uid in each row?