Re: Regexes: finding ALL matches (including overlap)

I gave a talk about this amongst other things at YAPC::Europe::2004. This question started at slide 20.

I would want "abcdef" =~ m/..*..*./g to return 20 = 6 choose 3 matches.

Hmmm... that's not quite the same thing I was talking about. How is Perl to know that .* is different from . ?

As far as my solution (actually largely due to abigail) is concerned /..*..*../ is simply /.{4,}/ and all matches thereof in "abcdef" would be 6..

substr("abcdef",0,4)
substr("abcdef",0,5)
substr("abcdef",0,6)
substr("abcdef",1,4)
substr("abcdef",1,5)
substr("abcdef",2,4)

Update: changed /.*/ to /.{4,}/ and made resulting changes.

Comment on Re: Regexes: finding ALL matches (including overlap) Select or Download Code

Replies are listed 'Best First'.
Re^2: Regexes: finding ALL matches (including overlap) by demerphq (Chancellor) on Jun 04, 2005 at 17:15 UTC
/....../ is simply /./* That strikes me as somewhat odd. The pattern on the right can match a string of less than 4 characters, the pattern of the left can not. --- $world=~s/war/peace/g	[reply]
Re^3: Regexes: finding ALL matches (including overlap) by nobull (Friar) on Jun 05, 2005 at 19:29 UTC
Yes thanks. Updating the previous node.	[reply]
Re^2: Regexes: finding ALL matches (including overlap) by kaif (Friar) on Jun 04, 2005 at 17:03 UTC
I would want "abcdef" =~ m/...../g to return 20 = 6 choose 3 matches. Hmmm... that's not quite the same thing I was talking about. How is Perl to know that `.` is different from `.` ? Easy: imagine I was matching `m/\w.\w.\w/g` instead. There really is no other possibility than have this return 20 matches (each `\w` has to match one of the 6 letters). Here are some more examples of what I would want (assuming I made no mathematical mistakes): `"abcdef" =~ m/...../g` returns 20 = 6 choose 3 `"abcdef" =~ m/./g` returns 28 = (6+2) choose 2 = number of substrings of length 6 string `"abcdef" =~ m/..../g` returns 10 = number of length 3 or greater substrings of length 6 string `"abcdef" =~ m/^.$/g` returns 1 `"abcdef" =~ m/^..$/g` returns 7 = number of ways of splitting a length 6 string into two parts	[reply] [d/l] [select]


Perl-Sensitive Sunglasses
	PerlMonks