Match the path of a URL, minus the filename extension

Question

What would be the best regular expression for this scenario?

Given this URL:

http://php.net/manual/en/function.preg-match.php

How should I go about selecting everything between (but not including) http://php.net and .php:

/manual/en/function.preg-match

This is for an Nginx configuration file.

`(?:http:[\/]{2}.+?[.])[^\/]+(.+)[.]+.+` – gaussblurinc Sep 01 '12 at 14:35 — gaussblurinc, Sep 01 '12 at 14:35

score 20 · Answer 1 · 2012-01-14T16:53:44.163

20

A regular expression might not be the most effective tool for this job.

Try using parse_url(), combined with pathinfo():

$url      = 'http://php.net/manual/en/function.preg-match.php';
$path     = parse_url($url, PHP_URL_PATH);
$pathinfo = pathinfo($path);

echo $pathinfo['dirname'], '/', $pathinfo['filename'];

The above code outputs:

/manual/en/function.preg-match

edited Jan 14 '12 at 16:53

answered Nov 29 '11 at 16:16

Thanks phoenix, that works too, but i was looking for a reg ex solution for this. – silkAdmin Nov 29 '11 at 16:43
1

@silkAdmin That's curious; why must the solution be a regular expression? – Jan 05 '12 at 16:14
1

Sorry i shouldn't have used the PHP tag that was confusing, i needed the regex for an Nginx config file where your solution is not an option – silkAdmin Dec 13 '12 at 22:51
Ajá. Thanks for the info; I went ahead and updated your question to make it clearer for future visitors. – Dec 14 '12 at 01:19

score 8 · Accepted Answer · edited Oct 16 '13 at 00:37

8

Like this:

if (preg_match('/(?<=net).*(?=\.php)/', $subject, $regs)) {
    $result = $regs[0];
}

Explanation:

"
(?<=      # Assert that the regex below can be matched, with the match ending at this position (positive lookbehind)
   net       # Match the characters “net” literally
)
.         # Match any single character that is not a line break character
   *         # Between zero and unlimited times, as many times as possible, giving back as needed (greedy)
(?=       # Assert that the regex below can be matched, starting at this position (positive lookahead)
   \.        # Match the character “.” literally
   php       # Match the characters “php” literally
)
"

edited Oct 16 '13 at 00:37

answered Nov 29 '11 at 16:12

FailedDev

26,680
9
53
73

Thanks FailedDev for the detail explanation, that work perfectly – silkAdmin Nov 29 '11 at 16:17
3

`'/net(.*)\.php/'`, simpler, shorter (and possibly better performing) version of the same expression. (I prefer expressions without wastefully unnecessary lookarounds.) – Qtax Nov 29 '11 at 16:55

morja · Answer 3 · 2011-11-29T16:19:56.013

3

Try this:

preg_match("/net(.*)\.php$/","http://php.net/manual/en/function.preg-match.php", $matches);
echo $matches[1];
// prints /manual/en/function.preg-match

edited Nov 29 '11 at 16:19

answered Nov 29 '11 at 16:12

morja

8,297
2
39
59

score 3 · Answer 4 · edited Jun 30 '14 at 17:47

3

There's no need to use a regular expression to dissect a URL. PHP has built-in functions for this, pathinfo() and parse_url().

edited Jun 30 '14 at 17:47

Peter Mortensen

30,738
21
105
131

answered Nov 29 '11 at 16:18

CrayonViolent

32,111
5
56
79

Ja͢ck · Answer 5 · 2012-09-02T09:39:15.873

Just for the fun of it, here are two ways that have not been explored:

substr($url, strpos($s, '/', 8), -4)

Or:

substr($s, strpos($s, '/', 8), -strlen($s) + strrpos($s, '.'))

Based on the idea that HTTP schemes http:// and https:// are at most 8 characters, so typically it suffices to find the first slash from the 9th position onwards. If the extension is always .php the first code will work, otherwise the other one is required.

For a pure regular expression solution you can break the string down like this:

~^(?:[^:/?#]+:)?(?://[^/?#]*)?([^?#]*)~
                              ^

The path portion would be inside the first memory group (i.e. index 1), indicated by the ^ in the line underneath the expression. Removing the extension can be done using pathinfo():

$parts = pathinfo($matches[1]);
echo $parts['dirname'] . '/' . $parts['filename'];

You can also tweak the expression to this:

([^?#]*?)(?:\.[^?#]*)?(?:\?|$)

This expression is not very optimal though, because it has some back tracking in it. In the end I would go for something less custom:

$parts = pathinfo(parse_url($url, PHP_URL_PATH));
echo $parts['dirname'] . '/' . $parts['filename'];

score 0 · Answer 6 · edited Aug 27 '12 at 18:58

0

Here's a regex solution better than what most have provided so far, if you ask me: http://regex101.com/r/nQ8rH5

/http:\/\/[^\/]+\K.*(?=\.[^.]+$)/i

edited Aug 27 '12 at 18:58

answered Aug 27 '12 at 06:33

Firas Dib

2,743
19
38

score 0 · Answer 7 · edited Aug 27 '12 at 22:03

0

Simple:

$url = "http://php.net/manual/en/function.preg-match.php";
preg_match("/http:\/\/php\.net(.+)\.php/", $url, $matches);
echo $matches[1];

$matches[0] is your full URL, $matches[1] is the part you want.

See yourself: http://codepad.viper-7.com/hHmwI2

edited Aug 27 '12 at 22:03

sam-w

7,478
1
47
77

answered Aug 27 '12 at 18:48

user1626664

313
1
6

score 0 · Answer 8 · answered Sep 01 '12 at 09:51

|(?<=\w)/.+(?=\.\w+$)|

select everything from the first literal '/' preceded by
look behind a Word(\w) character
until followed by a look ahead
- literal '.' appended by
- one or more Word(\w) characters
- before the end $

  re> |(?<=\w)/.+(?=\.\w+$)|
Compile time 0.0011 milliseconds
Memory allocation (code space): 32
  Study time 0.0002 milliseconds
Capturing subpattern count = 0
No options
First char = '/'
No need char
Max lookbehind = 1
Subject length lower bound = 2
No set of starting bytes
data> http://php.net/manual/en/function.preg-match.php
Execute time 0.0007 milliseconds
 0: /manual/en/function.preg-match

|//[^/](.)\.\w+$|

find two literal '//' followed by anything but a literal '/'
select everything until
find literal '.' followed by only Word \w characters before the end $

  re> |//[^/]*(.*)\.\w+$|
Compile time 0.0010 milliseconds
Memory allocation (code space): 28
  Study time 0.0002 milliseconds
Capturing subpattern count = 1
No options
First char = '/'
Need char = '.'
Subject length lower bound = 4
No set of starting bytes
data> http://php.net/manual/en/function.preg-match.php
Execute time 0.0005 milliseconds
 0: //php.net/manual/en/function.preg-match.php
 1: /manual/en/function.preg-match

|/[^/]+(.*)\.|

find literal '/' followed by at least 1 or more non literal '/'
aggressive select everything before the last literal '.'

  re> |/[^/]+(.*)\.|
Compile time 0.0008 milliseconds
Memory allocation (code space): 23
  Study time 0.0002 milliseconds
Capturing subpattern count = 1
No options
First char = '/'
Need char = '.'
Subject length lower bound = 3
No set of starting bytes
data> http://php.net/manual/en/function.preg-match.php
Execute time 0.0005 milliseconds
 0: /php.net/manual/en/function.preg-match.
 1: /manual/en/function.preg-match

|/[^/]+\K.*(?=\.)|

find literal '/' followed by at least 1 or more non literal '/'
Reset select start \K
aggressive select everything before
look ahead last literal '.'

  re> |/[^/]+\K.*(?=\.)|
Compile time 0.0009 milliseconds
Memory allocation (code space): 22
  Study time 0.0002 milliseconds
Capturing subpattern count = 0
No options
First char = '/'
No need char
Subject length lower bound = 2
No set of starting bytes
data> http://php.net/manual/en/function.preg-match.php
Execute time 0.0005 milliseconds
 0: /manual/en/function.preg-match

|\w+\K/.*(?=\.)|

find one or more Word(\w) characters before a literal '/'
reset select start \K
select literal '/' followed by
anything before
look ahead last literal '.'

  re> |\w+\K/.*(?=\.)|
Compile time 0.0009 milliseconds
Memory allocation (code space): 22
  Study time 0.0003 milliseconds
Capturing subpattern count = 0
No options
No first char
Need char = '/'
Subject length lower bound = 2
Starting byte set: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P 
  Q R S T U V W X Y Z _ a b c d e f g h i j k l m n o p q r s t u v w x y z 
data> http://php.net/manual/en/function.preg-match.php
Execute time 0.0011 milliseconds
 0: /manual/en/function.preg-match

score 0 · Answer 9 · edited Jan 05 '12 at 18:07

This general URL match allows you to select parts of a URL:

if (preg_match('/\\b(?P<protocol>https?|ftp):\/\/(?P<domain>[-A-Z0-9.]+)(?P<file>\/[-A-Z0-9+&@#\/%=~_|!:,.;]*)?(?P<parameters>\\?[-A-Z0-9+&@#\/%=~_|!:,.;]*)?/i', $subject, $regs)) {
    $result = $regs['file'];
    //or you can append the $regs['parameters'] too
} else {
    $result = "";
}

score -1 · Answer 10 · answered Sep 01 '12 at 14:42

-1

http:[\/]{2}.+?[.][^\/]+(.+)[.].+

let's see, what it done:

http:[\/]{2}.+?[.][^\/] - non-capture group for http://php.net

(.+)[.] - capture part until last dot occur: /manual/en/function.preg-match

[.].+ - matching extension of file like this: .php

answered Sep 01 '12 at 14:42

gaussblurinc

3,642
9
35
64

score -1 · Answer 11 · edited Jan 24 '12 at 20:36

-1

Regular expression for matching everything after "net" and before ".php":

$pattern = "net([a-zA-Z0-9_]*)\.php";

In the above regular expression, you can find the matching group of characters enclosed by "()" to be what you are looking for.

Hope it's useful.

edited Jan 24 '12 at 20:36