Regular Expression Matching More Than Expected

Question

I have a large chunk of HTML.

With this:

~<div>(?:.*?)<a[\s]+[^>]*?href[\s]?=[\s"\']+(#_ftnref([0-9]+))["\']+.*?>(?:[^<]+|.*?)?</a>(.*?)</div>~si

I am capturing this:

<div> </div><hr align="left" size="1" width="33%" /><div><p><a title="" href="#_ftnref1">[1]</a> This is not to suggest that there are only two possible arguments to be made in support of  blah blah <em>blah</em>.</p></div>

But! I want this:

<div><p><a title="" href="#_ftnref1">[1]</a> This is not to suggest that there are only two possible arguments to be made in support of  blah blah <em>blah</em>.</p></div>

Can you help?

PS: (?: ), in contrast to ( ), is used to avoid capturing text. I'm doing that on purpose because I want the returned $matches array to be consistent for several different regex not mentioned in this post.

http://stackoverflow.com/questions/3577641/how-to-parse-and-process-html-xml-with-php — nhahtdh, Feb 21 '13 at 03:42
Yes, I would mind. DOM is not appropriate because sometimes the markup is garbage. — conner_bw, Feb 21 '13 at 03:42
Mario, that worked! Please answer the question and I will give you the points. — conner_bw, Feb 21 '13 at 03:48

score 1 · Accepted Answer · answered Feb 21 '13 at 03:49

If lazy matching with .*? doesn't work, you need to come up with some exclusion pattern.

(?:(?!</div>).)*

Would for instance only match one div and stop/skip after any contained </div>

Alternatively a length constraint could be a workaround:

(?:.{0,20})

Regular Expression Matching More Than Expected

1 Answers1