Im running a database log and every day I log on a new row. My Mysql query therefore checks if the day (date (the unique key)) already exists, and if so, it tries to increment all the loggable values of the log-row by one. If the date record doesnt eyist yet, it will create a new row.
My SQL query is:
INSERT INTO `log` (`date`,`hits`,`stale`)
VALUES ('2012-03-06',1,1)
ON DUPLICATE KEY
UPDATE `hits`=`hits`+1,`stale`=`stale`+1
WHERE `date`='2012-03-06';"
All columns have 0 as default value, so if this query runs directly after midnight only 'stale' and 'hits' are set to 1. Otherwise 'stale' and 'hits' are both incremented.
I wish! (it doesn't work).
What am I missing? Which separator other then a comma should I use between 'hits' = 'hits' +1 and 'stale'='stale'+1?
Just get rid of the WHERE clause:
INSERT INTO `log` (`date`,`hits`,`stale`)
VALUES ('2012-03-06',1,1)
ON DUPLICATE KEY
UPDATE `hits`=`hits`+1,`stale`=`stale`+1;
Your separator is correct, but the UPDATE has already found the duplicate row to be able to trigger the ON DUPLICATE KEY, so you don't need to try to select it again using WHERE.
INSERT INTO `log` (`date`,`hits`,`stale`)
VALUES ('2012-03-06',1,1)
ON DUPLICATE KEY
UPDATE `hits`=`hits`+1,`stale`=`stale`+1
Demo here.
You shouldn't have the WHERE clause. ON DUPLICATE KEY UPDATE automatically limits the row it affects to the one that has the existing key.
Remove it and your query should work fine.
If you only want to do the update if some specific expression is true, you can do it with two statements:
INSERT IGNORE INTO x VALUES (.....);
UPDATE x SET ..... WHERE .....;
The INSERT will silently fail if there is a duplicate key.
Related
In a PHP script I'm using the MySQL INSERT ... ON DUPLICATE KEY UPDATE command to insert/update a number of records in my DB.
Is there anything that MySQL returns to say which records were inserted or updated?
An 'ugly' way would be to do a SELECT before each INSERT and see if the key exists before each INSERT but I'd like to know if MySQL has this function built in.
In case anyone need further info on what I'm trying to do is, to save the record id's to a log.
You will need another SELECT regardless, specifically SELECT ROW_COUNT().
From http://dev.mysql.com/doc/refman/5.0/en/information-functions.html#function_row-count
For INSERT ... ON DUPLICATE KEY UPDATE statements, the affected-rows value is 1 if the row is inserted as a new row and 2 if an existing row is updated.
No, there is no way to tell which records were inserted and which were updated purely with the means provided by MySQL without additional queries. However, you can add a column to the table where you can keep an indicator that you can use to mark the record as updated instead of inserted, e.g.
INSERT INTO YourTable (Col1, Col2, Updated)
VALUES ("value1", "value2", 0)
ON DUPLICATE KEY UPDATE
Col1 = values(Col1),
Col2 = values(Col2),
Updated = 1
Is there a way of removing record on duplicate key in MySQL?
Say we have a record in the database with the specific primary key and we try to add another one with the same key - ON DUPLICATE KEY UPDATE would simply update the record, but is there an option to remove record if already exists? It is for simple in/out functionality on click of a button.
It's a work-around, but it works:
Create a new column and call it do_delete, or whatever, making it a tiny-int. Then do On Duplicate Key Update do_delete = 1;
Depending on your MySQL version/connection, you can execute multiple queries in the same statement. However, if not, just run a separate query immediately afterwords. Either way, the next query would just be: Delete From [table] Where do_delete = 1;. This way, if its a new entry, it will not delete anything. If it was not a new entry, it will then mark it for deletion then you can delete it.
Use REPLACE INTO:
replace into some_table
select somecolumn from othertable
will either insert new data or if thr same data exist will delete the data and insert the new one
The nearest possible solution for the same is REPLACE statement. Here is the documentation for REPLACE.
A similar question was asked on MySQL Forums and the recommended(and only) answer was to use REPLACE.
to be more clear with mySql:
values can be from same table:
replace into table1 (column1,column2) select (val1,val2) from table1
or
values can be from another table:
replace into table1 (column1,column2) select (val1,val2) from table2
or
values can be from any table with condition:
replace into table1 (column1,column2) select (val1,val2) from table1 where <br>column3=val3 and column4=val4 ...
or
also remember values can be static with table name for namesake:
replace into table1 (column1,column2) select (123,"xyz") from table1
no error will be thrown even if the update results in duplicate entry, as it will be replaced.
(remember) only autoincrement value will be increased;
and
if you have column with data-type "TIMESTAMP" with "on update CURRENT_TIMESTAMP", it will have no effect;
Yes of course there is a solutions in MySQL for your problem.
If you want to delete or skip the new inserting record if there already a duplicate record exists in the key column of the table, you can use IGNORE like this:
insert ignore into mytbl(id,name) values(6,'ron'),(7,'son');
Here id column is primary key in the table mytbl. This will insert multiple values in the table by deleting or skipping the new duplicate records.
Hope this will fulfill your requirement.
In my table I have two fields: v_id and ip_address. I want to insert some data into this table only if the IP address doesn't already exist.
After Google'ing I came across the INSERT IGNORE INTO statement, and this is my code:
public function update_visits($ip_address)
{
$sql = 'INSERT IGNORE INTO `24h_visits` (ip_address) VALUES (?)';
if($this->db->query($sql, array($ip_address)))
{
return TRUE;
}
return FALSE;
}
It runs fine and without errors, but duplicate rows are still being made, even if the same IP is passed in as a parameter.
Anyone got a clue? Thanks.
You have to create a UNIQUE index on ip_address for INSERT IGNORE to work:
ALTER TABLE 24h_visits
ADD UNIQUE INDEX(ip_address)
However, I haven't seen the entirety of your schema, but I would assume that there's a column that stores a timestamp of the last visit. It's the only way this would make sense (so you can purge visits older than 24 hours every now and then).
In this case, you actually don't want INSERT IGNORE, but INSERT ... ON DUPLICATE KEY UPDATE instead. Assuming you have a column called last_visit:
INSERT INTO 24h_visits (ip_address, last_visit)
VALUES ('$ip_address', NOW())
ON DUPLICATE KEY UPDATE last_visit = NOW();
With INSERT IGNORE, the new row is never inserted, and thus you would always have the first value ever inserted on last_visit, which (the way I see it) is not entirely correct.
Add the UNIQUE constraint to your ip_address column.
Then your query would fail if it attempts to add a duplicate ip_address row (unless you use INSERT IGNORE).
The other answers don't actually answer the question: Creating a unique index prevents the duplicate from being inserted, which is a good idea, but it doesn't answer "how to insert if not already there".
This is how you do it:
INSERT IGNORE INTO 24h_visits (ip_address)
select ?
where not exists (select * from 24h_visits where ip_address = ?)
Additionally, this approach does not require any changes to schema.
I have a query like so:
INSERT INTO table1 (field1,field2) VALUES ('$value1','$value2') ON DUPLICATE KEY UPDATE field1 = '$value1'
I then want to get the last insert id if it does the insert, how can I do this? If the query ends up doing an update I dont want the last insert id. Is there a way to determine if it did an update or a insert?
I guess I should of searched the site before posting. Basically adding this worked:
id=LAST_INSERT_ID(id)
On the on duplicate update. I found that answer here:
Duplicate Key Last Insert ID
According to this MySQL Manual Page:
If a table contains an AUTO_INCREMENT
column and INSERT ... ON DUPLICATE KEY
UPDATE inserts or updates a row, the
LAST_INSERT_ID() function returns the
AUTO_INCREMENT value.
I want to update a record which may or may not be present in a table. If it is not present in the database then it will be inserted.
To prevent from select I am using UPDATE statement first and checking affected_rows > 0 if not then I am inserting this record into the table.
I was wondering if there is a better way to do this?
You could use INSERT ... ON DUPLICATE KEY UPDATE syntax:
INSERT INTO table (a,b,c) VALUES (1,2,3)
ON DUPLICATE KEY UPDATE c=c+1;
http://dev.mysql.com/doc/refman/4.1/en/insert-on-duplicate.html
The difference between this and REPLACE (Femaref's answer) is that REPLACE will delete the old row and then insert a new row if a key is duplicated, while this will update the existing row.
Use Replace instead of Insert.
http://dev.mysql.com/doc/refman/5.0/en/replace.html