Question

I have a csv file with 26,000 rows which i’m looping through rows and updating records (sometimes multiple) in a table with 250,000+ records. At the moment, its taken ages! I was wondering if there was an alternative way to do this quicker (in code or mysql/etc)

$row = 1;
if (($handle = fopen("zip-codes-database-DELUXE-BUSINESS2.csv", "r")) !== FALSE) {
    while (($data = fgetcsv($handle, 1000, ",")) !== FALSE) {
            if($row> 1){


                # GET THE AREACODE FROM data 20

                # name is: 21

                $insert = "UPDATE ".TBLPREFIX."cities SET data = '".escape(serialize($data))."' WHERE area_code = ".$data[20]." AND title = '".trim(strtoupper($data[21]))."'";
                mysql_query($insert) or die(mysql_error());
            }
            $row++;
    }
    fclose($handle);
}

Answer 1

Based on nothing I might try:

1. get the csv into a table via cmd line or ‘load data infile’
2. update the records into a temp table using a ‘insert … select’ where you join the old and new
3. move the temp table back onto the original (delete/rename)

Seems like it would be faster.. if a bit kludgy.