I have an issue with PDO that I'd really like to get an answer for after being plagued by it for quite some time.
Take this example:
I am binding an array of ID's to a PDO statement for use in a MySQL IN statement.
The array would be say: $values = array(1,2,3,4,5,6,7,8);
The database-safe variable would be $products = implode(',' $values);
So, $products would then be a STRING with a value of: '1,2,3,4,5,6,7,8'
The statement would look like:
SELECT users.id
FROM users
JOIN products
ON products.user_id = users.id
WHERE products IN (:products)
Of course, $products would be bound to the statement as :products.
However, when the statement is compiled and values bound, it would actually look like this:
SELECT users.id
FROM users
JOIN products
ON products.user_id = users.id
WHERE products IN ('1,2,3,4,5,6,7,8')
The problem is it is executing everything inside of the IN statement as a single string, given that I've prepared it as comma-separated values to bind to the statement.
What I actually need is:
SELECT users.id
FROM users
JOIN products
ON products.user_id = users.id
WHERE products IN (1,2,3,4,5,6,7,8)
The only way I can actually do this is by placing the values within the string itself without binding them, however I know for certain there has to be an easier way to do this.
This is the same thing as was asked in this question: Can I bind an array to an IN() condition?
The answer there was that, for a variable sized list in the in
clause, you'll need to construct the query yourself.
However, you can use the quoted, comma-separated list using find_in_set
, though for large data sets, this would have considerable performance impact, since every value in the table has to be cast to a char type.
For example:
select users.id
from users
join products
on products.user_id = users.id
where find_in_set(cast(products.id as char), :products)
Or, as a third option, you could create a user defined function that splits the comma-separated list for you (cf. http://www.slickdev.com/2008/09/15/mysql-query-real-values-from-delimiter-separated-string-ids/). This is probably the best option of the three, especially if you have a lot of queries that rely on in(...)
clauses.