Update mysql table if row excist if not add new

Question

First here I have js/jquery/ajax code:

$('#sacuvajMapu').click(function (e) {
 $.ajax({
            url: "insertMap.php",
            type: "POST",
            async: true, 
            data: { ajdi:ajdi, mapData:$('#mapData').val();}, //your form data to post goes here as a json object
            dataType: "html",

            success: function(data) {
            console.log(data);
            } 
            error: function(data) {
            console.log(data);
            }
        });
        });


        });

also I have mysql database:

Now I need to insert data with my ajax code to database so I write:

      try {        
            $STH = $db->prepare("INSERT INTO map (id_parcele, json, user_id)
VALUES (:1,:2,:3);");

            $STH->bindParam(':1', $_POST['ajdi']);
            $STH->bindParam(':2', $_POST['mapData']);
            $STH->bindParam(':3', $user_id);

and all is OK.

Now I have a problem. I must check if value $_POST['ajdi'] is in table column id_parcele and if it is then just to update table column json with values from $_POST['mapData'] and if POST['ajdi'] does not exist in table column id_parcele then I need to add new row with data above...

How I can do this? What is the right way for that?

Answer 1

Assuming that the record already existing will throw a primary key violation, you can do:

INSERT INTO map (id_parcele, json, user_id)
VALUES (:1,:2,:3)
on duplicate key update json=values(json), user_id=values(user_id)

This syntax is MySQL specific. http://dev.mysql.com/doc/refman/5.0/en/insert-on-duplicate.html

Answer 2

You can do this with a single query (as long as you put a unique key on id_parcele) with the following:

http://dev.mysql.com/doc/refman/5.0/en/insert-on-duplicate.html

INSERT INTO map (id_parcele, json, user_id)
VALUES (:1,:2,:3)
on duplicate key update json=// etc etc

You can also use the Replace Into syntax if you simply want to replace the values each and every time:

REPLACE works exactly like INSERT, except that if an old row in the table has the same value as a new row for a PRIMARY KEY or a UNIQUE index, the old row is deleted before the new row is inserted. See Section 13.2.5, “INSERT Syntax”.

replace INTO map (id_parcele, json, user_id)
VALUES (:1,:2,:3)

Again however, you will need to have a primary key or unique index on id_parcele.

The difference is that the replace into syntax will simply delete the old row first, then insert whatever you have in your query values - while the on duplicate... syntax allows you do do something like adding a counter to see how many updates have happened to that row by using the values in it already:

Assuming your original row was created with 0 in a field called parcel_updates (and there was such a column) you could do the following:

INSERT INTO map (id_parcele, json, user_id, parcel_updates)
VALUES (:1,:2,:3,0)
on duplicate key update parcel_updates=parcel_updates+1

This would keep the original values intact, but increase the count by one in the parcel_updates field.

Edit: Indexes

To add an index to a table that already exists, you need to run something like this (only once) before these types of commands will work:

CREATE unique INDEX parcelID ON map (id_parcele);

This will then not allow more than one row with the same value for id_parcele. You should however make sure that you don't already have the same value in there more than once. You can also define a column as unique when creating the table (for future reference):

create table map(
id_parcele int(10) unique key,
// etc etc