Return Affected Rows about insert-on-duplicate-key HOT 3 CLOSED

Sure. Do a PR and I'll merge.

Note that affected rows counts as 2 for duplicates.

http://stackoverflow.com/questions/3747314/why-are-2-rows-affected-in-my-insert-on-duplicate-key-update

from insert-on-duplicate-key.

@votemike I'm not sure how to use this feature when inserting multiple rows at once, and I believe there is at least an imprecision in the docs.
The current documentation says:

/*
* @return int 0 if row is not changed, 1 if row is inserted, 2 if row is updated
*/
public static function insertOnDuplicateKey(array $data, array $updateColumns = null)

however with multiple inserts/updates the output is the sum of the single affected-rows values.
Now I get that for a single operation 0-1-2 values help to discriminate what happened, but here, from the sum, how do I define how many rows have been inserted and how many updated?

It would be nice to return an array like

[ 'updates' => 2, 'inserts' => 17 ]

but is it even possible, I mean does MySQL even allow to discriminate between insert/update in a big single statement like the one the method creates?

Thanks, and thank you @yadakhov for the great tool! 👍

from insert-on-duplicate-key.

Wow, it's been a while since I did this work. @vifer do you still work with this? Any idea about @stebogit 's question?

from insert-on-duplicate-key.