Hi guys,
how can I match a block of
<li>item 1</li>
<li>item 2</li>
no matter there is a blank line before or after the block
and enclose it in <ul></ul> tags using PHP's preg_*
functions.
Thank you for answer
views:
80answers:
3
A:
Use an XML parser to do this, not a regex. PHP has one built in.
Rich Bradshaw
2010-06-24 09:53:46
As much as I agree to this, I think it should be a comment unless you add an example that shows how to do it
Gordon
2010-06-24 09:56:21
The code sample is not valid XML - it doesn't have a single top-level element. Use a HTML parser to parse HTML, and I think PHP has one of these built in too. (If not, I linked one in a comment to the question.)
Peter Boughton
2010-06-24 10:14:01
+1
A:
David already stated, that php got xml and html-parsers. However if you really want to use a regex, it probably would be something like:
preg_match('#<li>(.*?)</li>#', $string);
// Same thing
preg_match('#<li>(.*)</li>#U', $string);
faileN
2010-06-24 09:56:19
Good point. Just tested it and it also works without escaping. I removed the slashes. Thanks :)
faileN
2010-06-24 10:03:19
This works only if each `<li>..</li>` has a line to its self. If there are multiple list elements in on line it will match the whole line instead of each element for it's self.
jigfox
2010-06-24 10:13:22
No it works, if there are more <li>-tags on one line. That's why the `?` follows after the `*` or in the second version I used the `U`-modifier to reverse the greediness. If you would leave that out, things would go like you're saying and `item1</li><li>item2` would be matches. That's not the case here.
faileN
2010-06-24 10:19:20
ivan73 does this answer solve your problem? If so, mark is as the solution, if not explain how it fails.
Peter Boughton
2010-06-24 11:17:27
Ok, I tried that and it doesn't so I would try to explain it again, say you have text like this:`----``some pure text line``<li>item1</li>``<li>item2</li>``some pure text line``---`I need to match those items as a whole block and get something like this:`--``some pure text line``<ul>``<li>item1</li>``<li>item2</li>``</ul>``some pure text line``---`using preg_replace
ivan73
2010-06-24 11:37:46
Look at the answer I've added then - that should do what you're after.
Peter Boughton
2010-06-24 11:41:16
No, my *answer*, not that comment (which I'll go remove since I've linked to the core implementation in my answer). Here: http://stackoverflow.com/questions/3108812/how-to-match-a-block-of-li-li-using-regexp/3109512#3109512
Peter Boughton
2010-06-24 11:56:41
+2
A:
If this is safe, controlled input, and you just got LIs with missing parent ULs, you can do:
preg_replace ( '#\s*(?:<li>.*</li>\s*)+#' , '<ul>$0</ul>', $input )
(You may want to add some \n to the replacement string before or after the UL.)
NOTE: This will fail if:
- There are any existing UL/OL lists in the content.
- There is anything other than whitespace between consecutive list items.
- Any of the LIs span multiple lines (the
.excludes newline by default). - There are any attributes on the LIs.
- Possibly some things I haven't considered.
Some of these can relatively easily be catered for, but I'm not going to - if you haven't got known specific content, you should be using a real HTML parser instead.
The 'Regular' in Regular Expressions has a specific meaning, and full HTML is not a Regular language, so trying to handle all the intricacies of HTML with simple regex is liable to fail.
If you use a bad regex on user-supplied HTML, you may be introducing HTML injection vulnerabilities into your code.
Peter Boughton
2010-06-24 11:32:18