Quick Regular Expressions tutorial

[RobinTibbs]

I am totally new to regular expressions but think the script i am working on now would benefit from them. basically i know that different browsers/OSs identify themselves differently in the user agent string, and i know the pattern in my head but dont know how to apply that to regular expressions.

for instance i know that in firefox it is pretty much always the last part of the string,followed by a '/' then the version number. with IE its MSIE <version>

how can i use regular expressions with this, as currently i am using some string functions (strpos etc)

many thanks and hope that made sense  :D

[owenjh]

You can use preg_match to match if Firefox or MSIE is in a string like so:

[CODE]
$string = "Firefox /5.1";
if (preg_match("/Firefox(.*?)/", $string)) { echo "True"; }
[/CODE]

or if you want to get the version number of firefox you can use preg_match_all:

[CODE]
$string = "Firefox /5.1";
preg_match_all("/Firefox(.*)/", $string, $var);
print_r($var);
[/CODE]

Hope this helps

[RobinTibbs]

that looks like just what i need. i will give it a try, thanks :)

[RobinTibbs]

using that code i get the following output

[code]
Array ( [0] => Array ( [0] => Firefox /5.1 ) [1] => Array ( [0] => /5.1 ) )
[/code]

maybe i'm doing something wrong hmm. while i'm here,can i get the version and name with one expression or are 2 needed? thanks again

[effigy]

The first item is the entire match, the second the captured portions--those in parenthesis. Try something like:

[tt]/^(\w+)\s+([\d.]+)/[/tt]

Can you provide some examples of their identification strings?

[RobinTibbs]

some examples:

[b]Firefox 2.0[/b] - Mozilla/5.0 (Windows; U; Windows NT 5.2; en-GB; rv:1.8.1) Gecko/20061010 Firefox/2.0
[b]Firefox 1.5.0.7[/b] - Mozilla/5.0 (OS/2; U; Warp 4.5; en-US; rv:1.8.0.7) Gecko/20060915 Firefox/1.5.0.7

[b]IE 6[/b] - Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; {CB1C5177-05C4-4DFE-B687-AD6AF39ACD3E}; .NET CLR 1.1.4322)
[b]IE 5.5[/b] - Mozilla/4.0 (compatible; MSIE 5.5; Windows 98; )

as you can see, the pattern of the BrowserName and version are the same. hope that helps

[effigy]

Thanks. The expression I posted should work with these, capturing the browser name and the version.

[RobinTibbs]

so do i just use the same code but use the expression you gave?

[effigy]

Use[tt] preg_match [/tt]if you're matching a string that should only have one instance, and you only want one instance. Use[tt] preg_match_all [/tt]if you're matching a string that should have multiple instances, and you want them all.

[RobinTibbs]

sorry to be a pain, but could you go through what the expression does and what the symbols mean? :) many many thanks though!

[trochia]

http://www.google.com/search?hl=en&lr=&q=determine+browser+php

[effigy]

See the links in my signature.

^ anchors the expression to the beginning of the line (BOL).
(\w+) captures 1 or more "word" (a-z, 0-9, _) characters. The parens do the actual capturing.
\s+ consumes 1 or more instances of whitespace (spaces, new lines, tabs).
([\d.]+) captures (again, the parens) one or more digits (\d) or periods (.). These are grouped together by a character class ([ ]).

The beginning and ending forward slash delimit the expression.

[RobinTibbs]

just used that expression and it seems to not be capturing any info, the output i get is:

[code]Array ( [0] => Array ( ) [1] => Array ( ) [2] => Array ( ) )[/code]

trying to figure out what i'm doing wrong *goes off and tinkers*

[effigy]

[code]
<pre>
<?php
$tests = array(
'Firefox 2.0 - Mozilla/5.0 (Windows; U; Windows NT 5.2; en-GB; rv:1.8.1) Gecko/20061010 Firefox/2.0',
'Firefox 1.5.0.7 - Mozilla/5.0 (OS/2; U; Warp 4.5; en-US; rv:1.8.0.7) Gecko/20060915 Firefox/1.5.0.7',
'IE 6 - Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; {CB1C5177-05C4-4DFE-B687-AD6AF39ACD3E}; .NET CLR 1.1.4322)',
'IE 5.5 - Mozilla/4.0 (compatible; MSIE 5.5; Windows 98; )'
);
foreach ($tests as $test) {
preg_match('/^(\w+)\s+([\d.]+)/', $test, $matches);
print_r($matches);
}
?>
</pre>

[/code]

[RobinTibbs]

i think perhaps there was a misunderstanding?

the bit preceding the -  in the ID strings isnt actually part of the string, just was used as pretty formatting by me. sorry if that caused confussion

[effigy]

[code]
<pre>
<?php
$tests = array(
### Firefox 2.0
'Mozilla/5.0 (Windows; U; Windows NT 5.2; en-GB; rv:1.8.1) Gecko/20061010 Firefox/2.0',
### Firefox 1.5.0.7
'Mozilla/5.0 (OS/2; U; Warp 4.5; en-US; rv:1.8.0.7) Gecko/20060915 Firefox/1.5.0.7',
### IE 6
'Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; {CB1C5177-05C4-4DFE-B687-AD6AF39ACD3E}; .NET CLR 1.1.4322)',
### IE 5.5
'Mozilla/4.0 (compatible; MSIE 5.5; Windows 98; )'
);
foreach ($tests as $test) {
preg_match('%(MSIE|Firefox)[ /]([\d.]+)%', $test, $matches);
print_r($matches);
}
?>
</pre>

[/code]

[RobinTibbs]

thank you very much, that works perfectly, and goes along way to help me cracking regular expressions.