improve preg / pcre / regex to find PHP variable

Question

String to parse:

$str = "
public   $xxxx123;
private  $_priv   ;
         $xxx     = 'test';
private  $arr_123 = array();
"; //    |       |
   //     ^^^^^^^---- get the variable name

What I got so far:

    $str = preg_match_all('/\$\S+(;|[[:space:]])/', $str, $matches);
    foreach ($matches[0] as $match) {
        $match = str_replace('$', '', $match);
        $match = str_replace(';', '', $match);
     }

It works but I want to know if I can improve the preg, e.g. get rid of the two str_replace and maybe include \t in (;|[[:space:]])

Elias Van Ootegem · Accepted Answer · 2013-07-26 07:49:28Z

4

Using a positive lookbehind, you can get only that what you need, to be sure you'll only match valid variable names, I've used this:

preg_match_all('/(?<=\$)[a-zA-Z_\x7f-\xff][a-zA-Z0-9_\x7f-\xff]*/',$str,$matches);
var_dump($matches);

which correctly shows:

array (
  0 => 
  array (
    0 => 'xxxx123',
    1 => '_priv',
    2 => 'xxx',
    3 => 'arr_123'
  )
)

Which is all you need, no memory waisted on an array containing all variables with their leading and/or trailing chars.

The expression:

(?<=\$) is a positive lookbehind
[a-zA-Z_\x7f-\xff][a-zA-Z0-9_\x7f-\xff]*: is the regex PHP's site suggests themselves on their document pages

edited Jul 26, 2013 at 7:49

answered Jul 26, 2013 at 7:40

Elias Van Ootegem

76.7k10 gold badges123 silver badges160 bronze badges

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

6 Comments

Andreas Linden Over a year ago

this would also match other random strings like $("=§/&="§/$. also it would not properly match something like $x=123;

Andreas Linden Over a year ago

still would match for example in here "!"§$%&/()=?`" -> $%&/()

Elias Van Ootegem Over a year ago

@andreasLinden: I've edited my answer, to now use the regex PHP recommend themselves to match only valid variable names

Andreas Linden Over a year ago

okay fine, what is missing now to have a perfect solution: exclude variables inside of single quoted strings :D

Elias Van Ootegem Over a year ago

@AndreasLinden: That'll be a bit tricky, since lookarounds have to be of fixed lengths, but then again, you could just preg_replace('/(?<=\')(.*)\$/', '$1', $str); to remove all $ from single quoted strings

|

Andreas Linden · Accepted Answer · 2013-07-26 07:40:48Z

1

simply use backreferences

preg_match_all('/\$(\S+?)[;\s=]/', $str, $matches);
foreach ($matches[1] as $match) {

     // $match is now only the name of the variable without $ and ;
}

edited Jul 26, 2013 at 7:40

answered Jul 26, 2013 at 7:35

Andreas Linden

12.8k7 gold badges54 silver badges70 bronze badges

Comments

Benz · Accepted Answer · 2013-07-26 08:00:31Z

1

I changed the regex a little bit, take a look:

$str = '
public   $xxxx123;
private  $_priv   ;
         $xxx     = "test";
private  $arr_123 = array();
';

$matches = array();

//$str = preg_match_all('/\$(\S+)[; ]/', $str, $matches);
$str = preg_match_all('/\$(\S+?)(?:[=;]|\s+)/', $str, $matches); //credits for mr. @booobs for this regex

print_r($matches);

The output:

Array
(
    [0] => Array
        (
            [0] => $xxxx123;
            [1] => $_priv 
            [2] => $xxx 
            [3] => $arr_123 
        )

    [1] => Array
        (
            [0] => xxxx123
            [1] => _priv
            [2] => xxx
            [3] => arr_123
        )

)

Now you can use the $matches[1] in the foreach loop.

::Update::

After using regex "/\$([a-zA-Z_\x7f-\xff][a-zA-Z0-9_\x7f-\xff]*)/" the output looks correct.

String:

$str = '
public   $xxxx123; $input1;$input3
private  $_priv   ;
         $xxx     = "test";
private  $arr_123 = array();

';

And the output:

Array
(
    [0] => Array
        (
            [0] => $xxxx123
            [1] => $input1
            [2] => $input3
            [3] => $_priv
            [4] => $xxx
            [5] => $arr_123
        )

    [1] => Array
        (
            [0] => xxxx123
            [1] => input1
            [2] => input3
            [3] => _priv
            [4] => xxx
            [5] => arr_123
        )

)

edited Jul 26, 2013 at 8:00

answered Jul 26, 2013 at 7:35

Benz

2,33513 silver badges19 bronze badges

5 Comments

nstCactus Over a year ago

I'd change the regex even more so that it matches even if there's no space before assignation operator: \$(\S+?)(?:[=;]|\s+)

Daniel W. Over a year ago

problem here: $input; $input2; it does not match $input2;

Benz Over a year ago

@sbooob Nice one, I'll change the regex.

Benz Over a year ago

@DanFromGermany I don't have that problem? What string are you using?

nstCactus Over a year ago

This simple one seems to work: /\$([a-zA-Z_\x7f-\xff][a-zA-Z0-9_\x7f-\xff]*)/. "Simple" as it is just the regex for a valid PHP var name according to PHP manual

Collectives™ on Stack Overflow

improve preg / pcre / regex to find PHP variable

3 Answers 3

6 Comments

Comments

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

6 Comments

Comments

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related