#23721 [Bgs]: TRUE and FALSE constants slower than keywords in previous versions
| From: | sniper@php.net | Date: | Wed, 21 May 2003 01:08:24 +0000 |
| Subject: | #23721 [Bgs]: TRUE and FALSE constants slower than keywords in previous versions | ||
| References: | 1 | Groups: | php.bugs |
| Request: | Send a blank email to php-bugs+get-40109@lists.php.net to get a copy of this message | ||
ID: 23721
Updated by: sniper@php.net
Reported By: max at fuck dot org
Status: Bogus
Bug Type: Performance problem
Operating System: FreeBSD
PHP Version: 4.3.1
New Comment:
In real life applications you really won't notice any difference. And
the lowercase version of any constant defined
as case-insensitive will always be looked up faster.
(ie. true is looked for first, then if not found, TRUE)
Previous Comments:
------------------------------------------------------------------------
[2003-05-20 19:12:25] max at fuck dot org
no, i think it's great the performance is better, but all of a sudden
there is a performance difference between using true and TRUE. if true
is (as it would seem) the faster and more efficient way of using the
keyword, maybe someone should go through the documentation and make a
note of that?
------------------------------------------------------------------------
[2003-05-20 18:00:24] sniper@php.net
If better performance is a problem for you, maybe you should consider
using some slower language? :)
There is no bug.
------------------------------------------------------------------------
[2003-05-20 11:43:18] max at fuck dot org
-bash-2.05b$ cat true_test.php
<?php /* -*- C++ -*- */
function getmicrotime(){
list($usec, $sec) = explode(" ",microtime());
return ((float)$usec + (float)$sec);
}
$time_start = getmicrotime();
for ($i = 0; $i <= 1000000; $i++) {
if ($i % 2) {
$n = true;
}
else {
$n = false;
}
if ($n == true) {
}
else if ($n == false) {
}
}
$time_end = getmicrotime();
$time = $time_end - $time_start;
echo "true/false:\t $time seconds.\n";
$time_start = getmicrotime();
for ($i = 0; $i <= 1000000; $i++) {
if ($i % 2) {
$n = TRUE;
}
else {
$n = FALSE;
}
if ($n == TRUE) {
}
else if ($n == FALSE) {
}
}
$time_end = getmicrotime();
$time = $time_end - $time_start;
echo "TRUE/FALSE:\t $time seconds.\n";
?>
-bash-2.05b$ php -v
4.2.3
-bash-2.05b$ uname -a
FreeBSD 4.8-RELEASE FreeBSD 4.8-RELEASE #0: Thu Apr 3 10:53:38 GMT
2003
-bash-2.05b$ php true_test.php
X-Powered-By: PHP/4.2.3
Content-type: text/html
true/false: 5.9379440546036 seconds.
TRUE/FALSE: 5.9383209943771 seconds.
<configure/make/make install-cli 4.3.1 (no special config options)>
-bash-2.05b$ php -v
PHP 4.3.1 (cli) (built: May 20 2003 11:15:26)
Copyright (c) 1997-2002 The PHP Group
Zend Engine v1.3.0, Copyright (c) 1998-2002 Zend Technologies
-bash-2.05b$ php true_test.php
true/false: 4.2005170583725 seconds.
TRUE/FALSE: 5.1550530195236 seconds.
<configure/make/make install-cli latest snapshot (no special config
options)>
PHP 4.3.2RC4-dev (cli) (built: May 20 2003 11:41:28)
Copyright (c) 1997-2003 The PHP Group
Zend Engine v1.3.0, Copyright (c) 1998-2003 Zend Technologies
-bash-2.05b$ php true_test.php
true/false: 4.2775070667267 seconds.
TRUE/FALSE: 5.1591459512711 seconds.
I've tested this on multiple machines, seems that anything newer than
4.3.0 has this issue. Although I haven't looked myself I was told that
recently TRUE and FALSE are now defined as constants rather than
keywords. While these are both faster than the previous version(s), if
this difference in speed was intended, it wasn't documented anywhere.
------------------------------------------------------------------------
--
Edit this bug report at http://bugs.php.net/?id=23721&edit=1