Is there a token for capture line breaks in multiline regex?

Question

I've run into this problems several times before when trying to do some html scraping with php and the preg* functions.

Most of the time I've to capture structures like that:

<!-- comment -->
<tag1>lorem ipsum</tag>

<p>just more text with several html tags in it, sometimes CDATA encapsulated…</p>
<!-- /comment -->

In particular I want something like this:

/<tag1>(.*?)<\/tag1>\n\n<p>(.*?)<\/p>/mi

but the \n\n doesn't look like it would work.

Is there a general line-break switch?

Answer 1

I think you could replace the \n\n with (\r?\n){2} this way you capture the CRLF pair instead of just the LF char.

Answer 2

Are you sure you want to parse HTML using regexps ? HTML isn't regular and there are too many corner cases.

I would investigate some form of HTML parser (perhaps this one ?), and then identify the pattern you're interested in via the returned HTML data structure.

Answer 3

Or you could look at the Dom Extension to php. It has a function to load html from a string or a file. You can then use the php dom methods to traverse the dom and find the data you are interested in.