Regex to match whole word [duplicate]

Answer 1

\b will match word boundaries, so any transition from a word character ([a-zA-Z0-9_]) to a non-word character.

Since it sounds like you do not want to match if / follows your whole word, you are going to need a different boundary check. Something like the following should work:

string TextToFind = @"(?<!\S)www.abc.com(?!\S)";

This will cause the match to fail if the character before your word is not whitespace, or if the character after your word is not whitespace. Note that I used negative lookbehind/lookahead here instead of (?<=\s) and (?=\s) so that you will still match if your word is at the beginning or end of the string.

Answer 2

\b is a word boundary and will match where there's on one side a 'word' character (a word character here means a character that matches \w) and on another side a non-word character (that which matches \W).

If your definition of a whole word is 'something which has a space before and after it, unless it's at the beginning or at the end of the string', then you can use the regex:

(?<!\S)www.abc.com(?!\S)

To check exactly for this.

However, would you treat www.abc.com in those strings?

I'm going to visit www.abc.com; there's lots of things there.
What's this 'www.abc.com' you're speaking about?

If you still consider those as 'whole words', then maybe it'd better you only check for the main domain (i.e. there's no forward slash after www.abc.com):

\bwww.abc.com\b(?!/)

Then I guess your definition of 'whole word' is matching links with only the main domain name.

Answer 3

You seem to want to have only white spaces or begin/end of line around your replaced string.

Use somehting like (^|\s)string to replace(\s|$). Not sure what language you are using you may need to tune this string a little bit.