Question

I have three possible forms that will go through a regular expression, and I’d like to match them all.

text1
text1 text2
text1 text2 ;text3

I’ve got so far (.*?)(?:\s)(.*) working for 'text1 text2', but I able to handle all three cases if the semicolon is present. Any ideas?

Answer 1

The following should work, it would also put ‘text1’, ‘text2’, and ‘text3’ into the correct groups:

^(.+?)(?:\s(.*?)(?:\s;(.*))?)?$

See it working: http://www.rubular.com/r/IyPyF3wXLx

Here is an explanation:

^                # start of string
(.+?)            # put text1 in group 1
(?:              # start an optional non-capturing group
  \s(.*?)        # space followed by text2, put text2 in group 2
  (?:            # start an optional non-capturing group
    \s;(.*)      # space and semicolon, followed by text3, put text3 in group 3
  )?             # end of optional non-capturing group
)?               # end of optional non-capturing group
$                # end of string

The optional groups in the middle allow your regex to match text2 and text3 if they are present, but still match if they are not.