matching html tag content in regex

Question

I know "Dont use regex for html", but seriously, loading an entire html parser isn't always an option.

So, here is the scenario

<script...>
    some stuff
</script>

<script...>
    var stuff = '<';
    anchortext
</script>

If you do this:

<script[^>]*?>.*?anchor.*?</script>

You will capture from the first script tag to the /script in the second block. Is there a way to do a .*? but by replacing the . with a match block, something like:

<script[^>]*?>(^</script>)*?anchor.*?</script>

I looked at negative lookaheads etc, but I can't get something to work properly. Usually I just use [^>]*? to avoid running past the closing block, but in this particular example, the script content has a "<" in it, and it stops matching on that before reaching the anchortext.

To simplify, I need something like [^z]*? but instead of a single character or character range, I need a capture group to fit a string.

.*?(?!z) doesn't have the same effect as [^z]*? as I assumed it would.

Here is where I am stuck at: http://regexr.com?34llp

Duplicate of around a million other roughly identical questions on StackOverflow. — Adrian, Apr 24 '13 at 19:24
"loading an entire html parser isn't always an option" What is the specific reason with the current issue you are trying to solve for which DOMDocument is not an option? — PeeHaa, Apr 24 '13 at 19:29
@PeeHaa埽 Speed issues, loading the php_simple_dom_parser object concurrently for thousands of simultaneous documents takes quite an overhead hit. — tweak2, Apr 24 '13 at 19:41

score 3 · Accepted Answer · answered Apr 24 '13 at 19:30

Match-anything-but is indeed commonly implemented with a negative lookahead:

 ((?!exclude).)*?

The trick is to not have the . dot repeated. But make it successively match any character while ensuring that character is not the beginning of the excluded word.

In your case you would want to have this instead of the initial .*?

 <script[^>]*?>((?!</script>).)*?anchor.*?</script>

score 0 · Answer 2 · answered Apr 24 '13 at 19:46

0

like that:

$pattern = '~<script[^>]*+>((?:[^<]+?|<++(?!/script>))*?\banchor(?:[^<]+?|<++(?!/script>))*+)</script>~';

But DOM is the better way as far to do that.

answered Apr 24 '13 at 19:46

Casimir et Hippolyte

88,009
5
94
125

matching html tag content in regex

2 Answers2