Regular Expression for Detecting Emoticons

[Chia-you Chai]

Hello, I have a question on detecting the emoticons in a string.

If I have a sentence

This is a nice book :) I really love it:D

"

How can I use regular expression to extract the emoticons

:)

and

:D

in this sentence ?

Thank you helping !!

[Nicola Garofalo]

Unfortunately i am not so good with regular expressions since i don't use them so often and i can't practice them as i would, but i would like to try to give you an answer because i think regular exepressions are a really expressive language.
Then

i arrived to this one

.*([:)])|([:D]).*

If i am not wrong this expression recognises any character (zero or more times) followed by or sequence followed by any character (zero or more times)

But, as i told you, i am not so good with it and you surely find better answers

[Rob Spoor]

That's indeed not such a good regex ;)

It says "any character any number of times" followed by a : or ) (because the [] denote a character class, or a choice of characters) or a : or D, followed again by any character any number of times.

The regular expression needed is simple in this case:
- a :. Use the literal.
- a ) or D. Use either a character class or a | for this.

So you'd get ":[)D]" or ":(\\)|D)". Note that in the second example you need to escape the ). The first form is easier but allows extending only with two-character like :(. The regex would become ":[)D(]". The second form allows you to add any number of characters after the :. With additions :(, :'( and :wink: the regex would become ":(\\(|D|\\(|'\\(|wink:)". With a given set of smilies that can be automatically generated, using Pattern.quote (the : is taken inside the parentheses so you can also add smilies like ;) that don't start with :):

Set<String> smilies = ...
StringBuilder regex = new StringBuilder();
regex.append("(");
Iterator<String> iterator = smilies.iterator();
regex.append(Pattern.quote(iterator.next());
while (iterator.hasNext())
{
    regex.append('|');
    regex.append(Pattern.quote(iterator.next());
}
regex.append(")");

This will create a regex like this:

(\Q:)\E|\Q:D\E|\Q:(\E|\Q:wink:\E)

Don't be put off by those \Q and \E; \Q means to treat everything until the next \E as literals instead of meta characters. See also the Javadoc of java.util.regex.Pattern.

[Nicola Garofalo]

I was sure of it
Thank you for the explanation

[Rob Spoor]

You're welcome.

[Chia-you Chai]

Thanks for reply ! That's really a nice post to improve my knowledge about RE.
However, when I try to apply this regular expression to detect emoticon, I only can get the first emoticon in the sentence.

  
String s = "This is a great book ! :D I really love it :)";
Pattern pattern = Pattern.compile("(\\Q:)\\E|\\Q:D\\E|\\Q:(\\E|\\Q:wink:\\E)");
Matcher matcher = pattern.matcher(s);
if(matcher.find()) {
   System.out.println(matcher.group());  
}			

Is there any approach that I get get all the emoticons in the sentence ?

Thanks again

[Rob Spoor]

Change the if into a while would be a start

[Chia-you Chai]

Yes ! I change

if(metcher.find())

to

while(matcher.find())

and it works ! Thanks for helpin !