Parse url with pattern in PHP?

Question

How to determine, using regexp or something else in PHP, that following urls match some patterns with tokens (url => pattern):

node/11221 => node/%node
node/38429/news => node/%node/news
album/34234/shadowbox/321023 => album/%album/shadowbox/%photo

Thanks in advance!

Update 1

Wrote the following script:

<?php

$patterns = [
    "node/%node",
    "node/%node/news",
    "album/%album/shadowbox/%photo",
    "media/photo",
    "blogs",
    "news",
    "node/%node/players",
];

$url = "node/11111/news";

foreach ($patterns as $pattern) {
    
    $result_pattern = preg_replace("/\/%[^\/]+/x", '/*', $pattern);
    
    $to_replace = ['/\\\\\*/']; // asterisks
    $replacements = ['[^\/]+'];
    
    $result_pattern = preg_quote($result_pattern, '/');
    $result_pattern = '/^(' . preg_replace($to_replace, $replacements, $result_pattern) . ')$/';
    
    if (preg_match($result_pattern, $url)) {
        echo "<pre>" . $pattern . "</pre>"; 
    }
    
}


?>

Could anyone analyze whether this code is good enough? And also explain why there is so many slashes in this part $to_replace = ['/\\\\\*/']; (regarding the replacement, found exactly such solution on the Internet).

The backslash '\' is an escape character. So, in order to have a literal '\', you need to make it into '\\'. Having '\\\\*' will give you the literal '\*'. Also note that, to avoid having to use '\/' to denote literal '/', you can use | or any other special characters as the markers on the preg_replace() (as I did on my answer). It will make the regex cleaner. — Sutandiono
– Sutandiono, Commented Nov 21, 2016 at 2:44
@Sutandiono Ok, thanks. But there is 5 slashes. And with 4 ones the script doesn't work properly. What does the fifth slash do? — jekahm
– jekahm, Commented Nov 21, 2016 at 10:28
The first \ escapes the second \. The third escapes the fourth. The fifth escapes the * sign. Without the fifth \, the star will be treated as "zero or more". With the fifth \, the star is treated literally as the * sign. Also in case you missed it, the "*" comes from your preg_replace() where you replaced slash-%-followed-by-non-slash (\/%[^\/]+) with a /*. — Sutandiono
– Sutandiono, Commented Nov 21, 2016 at 17:14

Aziz Saleh · Accepted Answer · 2016-11-20 15:15:18Z

3

If you know the format beforehand you can use preg_match. For example in the first example, you know %node can only be numbers. Matching multiples should be as as easy as we did it earlier, just store the regex in the array:

$patterns = array(
    'node/%node' => '|node/[0-9]+$|',
    'node/%node/news' => '|node/[0-9]+/news|',
    'album/%album/shadowbox/%photo' => '|album/[0-9]+/shadowbox/[0-9]+|',
    'media/photo' => '|media/photo|',
    'blogs' => '|blogs|',
    'news' => '|news|',
    'node/%node/players' => '|node/[0-9]+/players|',

);

$url = "node/11111/players";

foreach ($patterns as $pattern => $regex) {
    preg_match($regex, $url, $results);
    if (!empty($results)) {
        echo "<pre>" . $pattern . "</pre>"; 
    }
}

Notice how I added the question mark $ to end of the first rule, this will insure that it doesn't break into the second rule.

edited Nov 20, 2016 at 15:15

answered Nov 20, 2016 at 1:55

Aziz Saleh

2,7071 gold badge19 silver badges27 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

georoot · Accepted Answer · 2016-11-20 02:40:21Z

Here is the generic solution to the solution above

<?php
    // The url part
    $url     = "/node/123/hello/strText";
    // The pattern part
    $pattern = "/node/:id/hello/:test";

    // Replace all variables with * using regex
    $buffer = preg_replace("(:[a-z]+)", "*", $pattern);
    // Explode to get strings at *
    // In this case ['/node/','/hello/']
    $buffer = explode("*", $buffer);
    // Control variables for loop execution
    $IS_MATCH = True;
    $CAPTURE  = [];
    for ($i=0; $i < sizeof($buffer); $i++) { 
        $slug = $buffer[$i];
        $real_slug = substr($url, 0 , strlen($slug));
        if (!strcmp($slug, $real_slug)) {
            $url = substr($url, strlen($slug));
            $temp = explode("/", $url)[0];
            $CAPTURE[sizeof($CAPTURE)+1] = $temp;
            $url = substr($url,strlen($temp));
        }else {
            $IS_MATCH = False;
        }

    }
    unset($CAPTURE[sizeof($CAPTURE)]);
    if($IS_MATCH)
        print_r($CAPTURE);
    else
        print "Not a match";
?>

You can pretty much convert the code above into a function and pass parameters to check against the array case. The first step is regex to convert all variables into * and the explode by *. Finally loop over this array and keep comparing to the url to see if the pattern matches using simple string comparison.

Sutandiono · Accepted Answer · 2016-11-20 02:59:38Z

2

As long as the pattern is fixed, you can use preg_match() function:

$urls = array (
          "node/11221",
          "node/38429/news",
          "album/34234/shadowbox/321023",
        );

foreach ($urls as $url)
{
  if (preg_match ("|node/([\d]+$)|", $url, $matches))
  {
    print "Node is {$matches[1]}\n";
  }
  elseif (preg_match ("|node/([\d]+)/news|", $url, $matches))
  {
    print "Node is {$matches[1]}\n";
  }
  elseif (preg_match ("|album/([\d]+)/shadowbox/([\d]+)$|", $url, $matches))
  {
    print "Album is {$matches[1]} and photo is {$matches[2]}\n";
  }
}

For other patterns to match, adjust as necessary.

answered Nov 20, 2016 at 2:59

Sutandiono

1,7501 gold badge12 silver badges22 bronze badges

Collectives™ on Stack Overflow

Parse url with pattern in PHP?

Update 1

3 Answers 3

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

Update 1

3 Answers 3

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related