Question

function csv_data_to_zip_array($csv)
{
    $f = fopen($csv, "r");
    $i= 0;
    while($line= fgets($f)){
        $line = preg_replace("/[^0-9]/", "", $line);
        if(is_numeric($line)&&strlen($line)==5){ // it is a zip code
            $array[] = $line;
        }
        $i++;
    }
    fclose($f);
    return $array;
}

That is my function, it’s reading a large csv with a bunch of zip codes into an array.

Answer 1

It’s what Ignacio said: either the while or if is always returning false, and you should initialize $array as an empty array before that while loop anyway.

Here is a suggestion for maybe getting results faster: If the CSV always contains the ZIP codes in the same column, use fgetcsv instead. You can then just specify the column number to check (is_numeric(substr($line[column_num],0,5))) rather than running a preg_replace on the whole line.

Edit:

Using the sample data you posted from your website (I noticed all data files in there have the zipcode in the first column; some files have a header row and some do not), this function does the trick:

function csv_data_to_zip_array($csv_path) {
  $fd = fopen($csv_path,'r');
  $zipcodes = array();
  while ($columns = fgetcsv($fd)) {
    if(is_numeric($columns[0])) {
      $zipcodes[] = $columns[0];
    }
  }
  return $zipcodes;
}