[Resolved] file size, an issue?

[Jason_Batten]

Hey all,

Just a general question about the size of PHP files. I have an articles.php which I use to add, edit and delete articles. It's up to 9kb, although I have seen a lot larger .php files I am just wondering when it starts to load a bit too long?

Do some functions/operations slow down others more or less?

Cheers

[bradgrafelman]

Yes, other functions take longer to execute than others. ereg() for example, is a lot slower than strstr() or something similar.

The difference, however, is less than a tenth of a second (in most cases). If your script is taking a long time to execute (i.e. a few seconds or more), a more common problem is that you are being slowed by file/URL opens, MySQL queries, etc. It's unlikely that the PHP interpreter is actually taking 5 seconds just to process your script alone. Now, if you have thousands upon thousdands of lines of code, this may be so. If you can, I would break it up into different categories, i.e. "file_functions.php", "mysql_operations.php", "xml_parser.php", etc., so that you can only load the functions/scripts that you need for a particular action, rather than declaring tons of functions and running tons of ereg()'s when you only need 1 set of instructions.

Clear as mud, right?

[Drakla]

A while back I wrote a simple thing to compare the speed of preg_match v str_replace and came up with some really surprising results related to scalability. It was only done for fun, but try it yourself if you like

<?php
	/*	How much slower is using preg_match compared to str_replace, and is it linear?
	 *	There are 3 tags per line.  When there are
	 *
	 *	1 line:  10 / 1
	 *	2 lines: 23 / 1
	 *  4 lines: 30 / 1
	 *	8 lines: 44 / 1
	 *	16 lines 67 / 1
	 *	64 lines 60 / 1
	 *  256 lines = 40 / 1 again?
	 *	4096 lines - 12 / 1
	 */
set_time_limit(60);

$reps1 = 100;
$reps2 = 10;
$linePow = 12;

echo "Num lines: ", pow(2, $linePow), '<br><br>';

$array = array(
	'name' => 'Foo Jones',
	'address' => 'Wembley Arena',
	'telephone' => 'oh yes, have one'
);

$replace = "Hello my name is {name} and I live at {address} \n\nand my telephone situation is {telephone}";

// make $replace a bit longer

for ($i = 0; $i < $linePow; $i++) $replace = $replace . "<br>\n" . $replace;


$t1 = get_microtime();

for ($i = 0; $i < $reps1; $i++)
{
	$r = $replace;

foreach($array as $tag => $val)
{
	$r = str_replace('{' . $tag . '}', $val, $r);
}
}

$t1 = get_microtime() - $t1;

echo "t1: $t1<br><br>";


$t2 = get_microtime();

for ($i = 0; $i < $reps2; $i++)
{
	$r = $replace;

$r = preg_replace('%{([^}]*)}%e', '$array["\\1"]', $r);
}

$t2 = get_microtime() - $t2;

echo "t2: $t2";


echo '<br>Ratio t2 / t1: ', ($reps1 * $t2) / ($t1 * $reps2);


function get_microtime()
{
	list($sec, $usec) = explode(' ', microtime());

return (float)$sec + (float)$usec;
}
?>

str_replace races ahead, but then the lead peaks and falls back as the amount of data being sifted grows.