Rails Regex Match Group Overwriting itself

Question

I am trying to match this string:

NFPA 101 19.7.2.2

and am using this regex:

(NFPA) (\w+)(?: ?(?:([^.]+)\.?)+)?

This seems to match the string, but the captured groups are not what I'm looking for. I expect:

1. NFPA
2. 101
3. 19
4. 7
5. 2
6. 2

What I get is this:

1. NFPA
2. 101
3. 2

See this rubular example: http://rubular.com/r/43VY0yyNa7

It's as if that last recurring capture group is being overwritten by the final match. Is there a way to have all of these come back as capture groups as I need?

Added another regex that gives me the similar problem described above: (NFPA) (.+) (.+?.)+(.+)

Answer 1

The issue is you're using non-capturing group symbol : which isn't gonna work to select the string as separate capture group. To overcome the issue you need to use Positive / Negative Lookahead. So, the following regex should work in this case :

(\w+|\d+[-]\d+)(?=\s?)(?![-])

see demo