reg expression to strip links out of HTML?

steveryan

Hey all,

I'd really appreciate some help on this.

I need a regular expression which will strip all the links from a HTML document. Basically I am reading a websites HTML into a variable and then I need to strip the links from there.

Any idea?

Thanks!!

Weedpacket

I presume you want to retain the bit between the opening and closing tags?

$text=preg_replace("/<\sa\s[^>]>(.*?)<\/a>/i","\1",$text);

Um, that looks about right for a first approximation.

the_Igel

I think one can't put a space between < and tagname, like
<[space]A href="#">

So there's no need to write in regexp "<\s*a......", "<a..........." must be enough.

the_Igel

And, to be more accurate, i must mention, that simply looking for closing ">" may be incorrect. This character may be found in "title" attribute for link or even (i wonder) in "href".

But if you don't need to be superaccurate and just want to get the result, in most cases mentioned regexp would be nice enough.

Weedpacket

Originally posted by the_Igel
And, to be more accurate, i must mention, that simply looking for closing ">" may be incorrect. This character may be found in "title" attribute for link or even (i wonder) in "href".

Gee, that doesn't sound like very safe HTML... browser could get awfully confused about where tags and/or attribute values end. In attribute values you're supposed to use >.

If people are going to write > in attribute values their HTML deserves to break.

the_Igel

Yes, they're supposed to, but i think no browser will have problems if i write <A href="#" title="go here>>>, dude">. And there are so many people around the world who do things they aren't supposed to... 😉

Weedpacket

Originally posted by the_Igel
Yes, they're supposed to, but i think no browser will have problems if i write <A href="#" title="go here>>>, dude">. And there are so many people around the world who do things they aren't supposed to... 😉

Well I'm not going to pander to such people and write an entire non-compliant HTML parser for their sakes :mad:

the_Igel

You're completely right when speaking about writing a parser.
But "stripping links from a page" is completely different. If you need it, you'd better do your best for script to work.

Not that i'm defending those guys writing non-compliant code 🙂 ...