Question

Can someone please tell me why my pattern: <p(\s+(.*)?)?>(.[^</p>]*)?</p> does not work correctly. Example matches:

1. <p>This is a test and anything can be here even other <tags>tags</tags></p>
2. <p style="test">This is a test</p>
3. <p></p>

And if the above were all on one line it should find 3 separate patterns. The link below demonstrates its true behaviour which is very odd…

http://regexr.com?33jrn

The matches it finds should always immediately start when it finds <p and immediately stop when it finds </p>

Answer 1

There are a couple of problems with your regex. Let’s see what they look like.

Here’s your regex: –

<p(\s+(.*)?)?>(.[^</p>]*)?</p>

• Problem 1: – Notice the pattern (.*)?. It is not doing what you think. It is not a enforcing a reluctant behaviour on * quantifier. Rather it’s a enforcing an optional quantifier (?) over a greedy * quantifier. It simply means match 0 or 1 repetition of (.*). For making it reluctant, you need to move ? inside the bracket. So, you need to use (.*?) instead of (.*)?.
• Problem 2: – [^</p>] does not negate </p> rather it negates – <, /, p, > as separate characters. Note that in a character class, each character is taken literally. There is not grouping in there. So, (.[^</p>]*) means match a character if is not followed by 0 or more repetition of either of [</p>]. That is not what you want. If you want to match a sequence that is not </p>, then you can use a negative look-ahead like this: – ((?!</p>).)*. Now this will check first whether the following sequence is not </p>, then it matches the next character.

So, your regex pattern should be: –

<p(\s+(.*?))?>((?!</p>).)*</p>

Or, you can even simplify your regex to: –

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