Re: How does one get all possible matches from regex?

[educated_foo]

Your "correct results" don't correspond to what I get; if you want "all possible matches for REGEX," you should use this:

1 while /REGEX(?{print $&})(?!)/;

i.e. "match REGEX, print what it matched, then fail."

[LanX]

I didnt understand the OP, but is this not easier?

 print $1 while /(REGEX)/g

[educated_foo]

They're different: /X/g does non-overlapping matches; /X(?{print$&})(?!)/ does all matches. It depends what you want.

[LanX]

Well it took me a while to understand whats going on in your code...

Please note that the while-loop is redundant (and confusing =)

  DB<138> $_="  x1x2x3x  x4x5x"
 => "  x1x2x3x  x4x5x"

  DB<139> ; /x(\d)x(?{print "<$1>\t"})(?!)/
<1>    <2>    <3>    <4>    <5>

by placing an _empty_ negative lookahead you are forcing your regex to reject all matches and to backtrack through the whole string.

But the embedded perlcode (which is an experimental feature IIRC) prints the temporary match before the next attempt is started.

edit

as a side note, the same effect can be achieved with "conventional" means (i.e. manipulation of pos)

  DB<154> while ( /x(\d)x/g ) { print $1; pos($_) -= length($ &)-1 }
12345