How to match emoticons with regular expressions?

I need to capture smiley faces like

:)
:P
:-P
=)
:D
;)

And so on, along with general text. This is my current regex:

\b[0-9A-Za-z'\&\-\./()=:;]+\b

However, it doesn't match ()=:; for some reason. Am I missing something?

Edi开发者_开发问答t: Based on Mark's feedback here is an example that I need to parse:

hi =as.) friend :) haha yay! ;) =) test test) R&R I.O.U. 24/7

This should extract:

hi
friend
:)
haha
yay
;)
=)
test
test
R&R
I.O.U.
24/7

I'm having trouble getting this to work using any of the solutions proposed.

This is an example that captures a word followed by the above examples. It captures a single word and a following emoticon in separate capture groups. The Rubular link.

\s(\w+)\s((?::|;|=)(?:-)?(?:\)|D|P))

Edit Based on the edits and the given example, this might be what is desired. It defines two capture groups, one for the general text and one for an emoticon. Here is the Rubular link.

([0-9A-Za-z'\&\-\.\/\(\)=:;]+)|((?::|;|=)(?:-)?(?:\)|D|P))

I tested it here with Rubular. If I escape the / then it works. (Update: and remove the word boundaries.)

[0-9A-Za-z'\&\-\.\/()=:;]+

Update: The forward slash escaping was the error message I got from rubular. The real problem here are the \b anchors. They are matching a word boundary, i.e. a boundary from [A-Za-z0-9_] to something else. that means it will not match a :-) because there is no word boundary.

If you remove \b in front of the character class and behind the character class or escape / it works.

I think the reason why it works when you remove the word boundary is because it will only match words with alphanumeric characters...if I don't remember it wrong.