Home

str_replace vs. preg_replace

2008, May 9 - 13:27 — müzso
Today I faced the problem of removing all occurances of a fixed number of characters from a string. I was curious how str_replace performed against preg_replace in this case.

Here's an example code that tries to simulate running both functions on a text of 100 sentences, each having 5-10 words and each word consisting of 10-20 characters. The sentences are separated by newlines and words are separated by spaces. The code runs both methods a 100000 times removing newlines, carriage return and tab characters from the random generated text.
<?php
set_time_limit(240);
$input = '';
$ord_a = ord('a');
$max = ord('z') - $ord_a;
for ($i = 0; $i < 100; $i++) {
  $s_len = rand(5, 10);
  for ($j = 0; $j < $s_len; $j++) {
    $w_len = rand(10, 20);
    $word = '';
    for ($j = 0; $j < $w_len; $j++) {
      $word .= chr($ord_a + rand(0, $max));
    }
    $input .= $word . ' ';
  }
  $input .= $word . "\n";
}

$count = 100000;
$start = time();
for ($i = 0; $i < $count; $i++) {
  $output = str_replace(array("\n", "\r", "\t"), array('', '', ''), $input);
}
print('time1: ' . (time() - $start));
$start = time();
for ($i = 0; $i < $count; $i++) {
  $output = preg_replace('/[\n\r\t]/', '', $input);
}
print('time2: ' . (time() - $start));
?>

The difference is quite significant: str_replace performed almost a magnitude faster than preg_replace!

Syndicate content