[RESOLVED] PCRE matching numeric entities in strings

bunner_bob

I need a little help with a PCRE pattern I'm trying to write to use in preg_replace.

I'm trying to match any empty paragraphs. The input may contain tags, and any combination of space characters including the numeric nonbreaking space *

Here's what I have so far. It does match what I'm after, but also matches any of the individual characters within * - for instance, a 1, 6, ¡ etc., which I don't want. I think it's probably pretty clunky and welcome any advice.

$pattern = '/<p>(<strong>)*[\s\W(*)]*(<\/strong>)*<\/p>/';

What I think I'm asking for is:
- a tag
- possible opening tag
- 0 or more spaces, non-word characters or * characters
- possible closing tag
- closing tag

Thanks in advance for any advice!

dalecosp

What do you mean by "numeric nonbreaking space" ... this?

echo chr(160);

The regex you've posted appears to work just fine ... I might use a different delimiter (I prefer "#" or "%")...

bunner_bob

dalecosp;11035309 wrote:
What do you mean by "numeric nonbreaking space" ... this?
echo chr(160);

Whoops - looks like the forum edited my post and stripped out the numeric character entity from my pattern. I'm not sure how to get it to appear, but in the parentheses following the \W you should see the characters & # 1 6 0 ; (but without spaces), which is the numeric character entity for nonbreaking space. This string correctly matches on the input, but it also matches & # 1 6 1 ; - I need it to match the full pattern.

Here's another try using PHP tags - I see it converted it to a space in this case.

$pattern = '/<p>(<strong>)*[\s\W(&#160;)]*(<\/strong>)*<\/p>/';

Weedpacket

bunner bob wrote:
in the parentheses following the \W you should see the characters

If that's where you've got it then it won't work; [noparse][\s\W(⅋#160😉][/noparse] would match any whitespace character, any non-wordlike character, and any [font=monospace]([/font], [font=monospace]([/font], [font=monospace])[/font], [font=monospace]&[/font], [font=monospace]#[/font], [font=monospace]1[/font], [font=monospace]6[/font], [font=monospace]0[/font], or [font=monospace];[/font] character.

noparse*[/noparse]

Note that the same entity can be written as ⅋#xA0;

bunner_bob

Ahh perfect - thank you!