Regular expression for anchor tag with all attributes

Question

I'm trying to get a regular expression to replace all the links out of a text string for the value of the link.

A link may look like these:

<a href="http://whatever" id="an_id" rel="a_rel">the link</a>
<a href="/absolute_url/whatever" id="an_id" rel="a_rel">the link</a>

I want a regular expression that I get: the link

Solution 1

/<a[^>]*>([^<]+)<\/a>/g

It's far from being perfect, but you need to provide more examples of what is a correct match and what isn't (e.g. what about whitespaces?)

Solution 2

/<a[\s]+([^>]+)>((?:.(?!\<\/a\>))*.)<\/a>/g

This one will match any <a ...>...</a> tag including correctly matching ones that contain a < or any full tags such as:

blah blah <a href="test.html">This line contains an HTML opening < bracket.</a> blah blah
blah blah <a href="test.html">This line contains <strong>bold</strong> text.</a> blah blah

Would capture:

<a href="test.html">This line contains an HTML opening < bracket.</a>

• with capture groups:
  • href="test.html"
  • This line contains an HTML opening < bracket.

and

<a href="test.html">This line contains <strong>bold</strong> text.</a>

• with capture groups:
  • href="test.html"
  • This line contains <strong>bold</strong> text.

It also includes capturing groups for the tag attributes (like class="", href="", etc) and contain (what is between the tag) that can be removed if you do not need them.

If you want to capture across multiple lines add an "s" before or after the "g" flag at the end. Note that the "s" flag may not work in all flavors of regular expression.

Capture example (not using the "s" flag - not supported by regexr yet): http://regexr.com/39rsv

Solution 3

Just a little correction from the accepted answer. This is the correct regex: /<a[^>]*>([^<]+)<\/a>/g. The forward slash (/) for closing the anchor tag </a> was not escaped so no match will be made.

Comments

[Lobo]

Hi Florian, others example: <a href="/absolute_url/whatever" id="an_id" rel="a_rel"></a> <a href="/absolute_url/whatever">a link</a> <a href="domain.com">a link</a>

[Brian Leishman]

You have an unescaped forward slash near the end

[Jerry]

how would you modify this to cover bla bla <a href="test.html" data-annoying=">" >yikes</a>? That's the one killing me right now.

[Kshitij]

Note: This would not work for nested elements. Regex should be case insensitive as <a> and <A> both are valid.

[idungotnosn]

Good question, @Jerry. I don't really know how to answer your question (and this post is over a year too late), but I would think that any HTML attributes that contain XML special characters like that should have those characters encoded somehow.

[robrecord]

Escaped < and > where it shouldn't be... correct version is <a[\s]+([^>]+)>((?:.(?!<\/a>))*.)<\/a>