How to know if when using “on duplicate key update” a row was inserted or updated?

Question

We have a database that gets updated everyday at midnight with a cronjob, we get new data from an external XML.

What we do is that we insert all the new content and

Answer 1

You can get this information at the time of the insert/update by examining the number of affected rows in the result set.

MySQL documentation states:

With ON DUPLICATE KEY UPDATE, the affected-rows value per row is 1 if the row is inserted as a new row and 2 if an existing row is updated.

You'll need to combine ROW_COUNT with LAST_INSERT_ID to get your answer and insert one row at a time.

Answer 2

I can say How I did in PHP:

1) Simple query SELECT MAX(id) and remember it to $max_id from table before Insert On Duplicate.

2) Then during the update process collect ID of affected rows (no mater new or existed): $ids[] = mysql_insert_id();

3) Then $inserted_rows = max($ids)-$max_id;

4) Updated rows = count($ids_srt)-$inserted_rows

$max_id = mysql_query("SELECT MAX(id) from table");
$max_id = mysql_result($max_id, 0);

// !!! prepare here 'insert on duplicate' query in a cycle

$result=mysql_query($query);
$ids[] = mysql_insert_id();

// finish inserting and collecting affected ids and close cycle

$inserted_rows = max($ids)- $max_id;
$updated_rows = count($ids)- $inserted_rows

Answer 3

Add an update_count INT NOT NULL DEFAULT 1 column and change your query:

INSERT
INTO    table (id, col1, col2, col3)
VALUES
(id_value, val1, val2, val3),
(id_value, val1, val2, val3,),
(id_value, val1, val2, val3),
(id_value, val1, val2, val3),
ON DUPLICATE KEY
UPDATE 
        col1 = VALUES (col1), 
        col2 = VALUES (col2), 
        col3 = VALUES (col3),
        update_count = update_count + 1;

You can also increment it in a BEFORE UPDATE trigger which will allow you to keep the query as is.