Compare strings and extract variables? - php

Question

Could someone tell me how I would do this. I have 3 strings.
$route = '/user/$1/profile/$2';
$path = '/user/profile/$1/$2';
$url = '/user/jason/profile/friends';
What I need to do is check to see if the url conforms to the route. I am trying to do this as follows.
$route_segments = explode('/', $route);
$url_segments = explode('/', $url);
$count = count($url_segments);
for($i=0; $i < $count; $i++) {
if ($route_segments[$i] != $url_segments[$i] && ! preg_match('/\$[0-9]/', $route_segments[$i])) {
return false;
}
}
I assume the regex works, it's the first I have ever written by myself. :D
This is where I am stuck. How do I compare the following strings:
$route = '/user/$1/profile/$2';
$url = '/user/jason/profile/friends';
So I end up with:
array (
'$1' => 'jason',
'$2' => 'friends'
);
I assume that with this array I could then str_replace these values into the $path variable?

Answer 1

$route_segments = explode('/',$route);
$url_segments = explode('/',$url);
$combined_segments = array_combine($route_segments,$url_segments);
Untested and not sure how it reacts with unequal array lengths, but that's probably what you're looking for regarding an element-to-element match. Then you can pretty much iterate the array and look for $ and use the other value to replace it.
EDIT
/^\x24[0-9]+$/
Close on the RegEx except you need to "Escape" the $ in a regex because this is a flag for end of string (thus the \x24). The [0-9]+ is a match for 1+ number(s). The ^ means match to the beginning of the string, and, as explained, the $ means match to the end. This will insure it's always a dollar sign then a number.
(actually, netcoder has a nice solution)

Answer 2

I did something similar in a small framework of my own.
My solution was to transform the template URL: /user/$1/profile/$2
into a regexp capable of parsing parameters: ^\/user\/([^/]*)/profile/([^/]*)\/$
I then check if the regexp matches or not.
You can have a look at my controller code if you need to.

Answer 3

You could do this:
$route = '/user/$1/profile/$2';
$path = '/user/profile/$1/$2';
$url = '/user/jason/profile/friends';
$regex_route = '#'.preg_replace('/\$[0-9]+/', '([^/]*)', $route).'#';
if (preg_match($regex_route, $url, $matches)) {
$real_path = $path;
for ($i=1; $i<count($matches); $i++) {
$real_path = str_replace('$'.$i, $matches[$i], $real_path);
}
echo $real_path; // outputs /user/profile/jason/friends
} else {
// route does not match
}

Answer 4

You could replace any occurrence of $n by a named group with the same number (?P<_n>[^/]+) and then use it as pattern for preg_match:
$route = '/user/$1/profile/$2';
$path = '/user/profile/$1/$2';
$url = '/user/jason/profile/friends';
$pattern = '~^' . preg_replace('/\\\\\$([1-9]\d*)/', '(?P<_$1>[^/]+)', preg_quote($route, '~')) . '$~';
if (preg_match($pattern, $url, $match)) {
var_dump($match);
}
This prints in this case:
array(5) {
[0]=>
string(28) "/user/jason/profile/friends"
["_1"]=>
string(5) "jason"
[1]=>
string(5) "jason"
["_2"]=>
string(7) "friends"
[2]=>
string(7) "friends"
}
Using a regular expression allows you to use the wildcards at any position in the path and not just as a separate path segment (e.g. /~$1/ for /~jason/ would work too). And named subpatterns allows you to use an arbitrary order (e.g. /$2/$1/ works as well).
And for a quick fail you can additionally use the atomic grouping syntax (?>…).