I am having a difficult time forming a conditional INSERT
I have x_table with columns (instance, user, item) where instance ID is unique. I want to insert a new row
What you want is INSERT INTO table (...) SELECT ... WHERE ...
. from MySQL 5.6 manual.
In you case it's:
INSERT INTO x_table (instance, user, item) SELECT 919191, 123, 456
WHERE (SELECT COUNT(*) FROM x_table WHERE user=123 AND item=456) = 0
Or maybe since you're not using any complicated logic to determiante whether to run the INSERT
or not you could just set a UNIQUE
key on the combination of these two columns and then use INSERT IGNORE
.
Insert into x_table (instance, user, item) values (919191, 123, 456)
where ((select count(*) from x_table where user=123 and item=456) = 0);
The syntax may vary depending on your DB...
So this one stands for PostgreSQL
INSERT INTO x_table
SELECT NewRow.*
FROM (SELECT 919191 as instance, 123 as user, 456 as item) AS NewRow
LEFT JOIN x_table
ON x_table.user = NewRow.user AND x_table.item = NewRow.item
WHERE x_table.instance IS NULL
If your DBMS does not impose limitations on which table you select from when you execute an insert, try:
INSERT INTO x_table(instance, user, item)
SELECT 919191, 123, 456
FROM dual
WHERE NOT EXISTS (SELECT * FROM x_table
WHERE user = 123
AND item = 456)
In this, dual
is a table with one row only (found originally in Oracle, now in mysql too). The logic is that the SELECT statement generates a single row of data with the required values, but only when the values are not already found.
Alternatively, look at the MERGE statement.
With a UNIQUE(user, item)
, do:
Insert into x_table (instance, user, item) values (919191, 123, 456)
ON DUPLICATE KEY UPDATE user=123
the user=123
bit is a "no-op" to match the syntax of the ON DUPLICATE
clause without actually doing anything when there are duplicates.
In case you don't want to set a unique constraint, this works like a charm :
INSERT INTO `table` (`column1`, `column2`) SELECT 'value1', 'value2' FROM `table` WHERE `column1` = 'value1' AND `column2` = 'value2' HAVING COUNT(`column1`) = 0
Hope it helps !