How does one get all possible matches from regex?

supriyoch_2008 has asked for the wisdom of the Perl Monks concerning the following question:

Hi Perlmonks,

I am interested to get all the possible matches from a string using regex. But I have got fewer matches than expected with the perl script g1.pl. I have given the script, the incorrect results and the expected results below. I request perlmonks to provide suggestions how to write the regex to get the correct results.

Here goes the script:

#!/usr/bin/perl  
use warnings; 
   
  $seq="TT TATAAT CGCG ATG CAG GAG TGG TAA TGA TAG CC TGA TATAAT CCC A
+TG CTA 
CAT TGA TT"; 

 $seq=~ s/\s//gs;   

while ($seq=~ /([AG]TG).*?(TAA|TAG|TGA)+?/gs) {
           
    my $match=$&; 
     $match=~ s/\s//g; push  @matches,$match;}  

print"\n Matches are:\n\n"; 
print join ("\n",@matches);
print"\n\n";  
exit;
[download]

I have got the incorrect results like:

C:\Users\Dr Supriyo>cd desktop

C:\Users\Dr Supriyo\Desktop>g1.pl

 Matches are:

ATGCAGGAGTGGTAA
ATGCTACATTGA
[download]

The correct results should be:

Matches are:

ATGCAGGAGTGGTAA
ATGCAGGAGTGGTAATGA
ATGCAGGAGTGGTAATGATAG
ATGCAGGAGTGGTAATGATAGCCTGA
ATGCTACATTGA
[download]

Comment on How does one get all possible matches from regex? Select or Download Code

Replies are listed 'Best First'.
Re: How does one get all possible matches from regex? by educated_foo (Vicar) on Dec 10, 2013 at 03:31 UTC
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 $&})(?!)/;` [download] i.e. "match REGEX, print what it matched, then fail."	[reply] [d/l]
Re^2: How does one get all possible matches from regex? by LanX (Saint) on Dec 10, 2013 at 03:38 UTC
I didnt understand the OP, but is this not easier? `print $1 while /(REGEX)/g` Cheers Rolf ( addicted to the Perl Programming Language)	[reply] [d/l]
Re^3: How does one get all possible matches from regex? by educated_foo (Vicar) on Dec 10, 2013 at 04:17 UTC
They're different: `/X/g` does non-overlapping matches; `/X(?{print$&})(?!)/` does all matches. It depends what you want.	[reply] [d/l] [select]
Re^4: How does one get all possible matches from regex? by LanX (Saint) on Dec 10, 2013 at 16:27 UTC
Re: How does one get all possible matches from regex? (combinations permutations) by Anonymous Monk on Dec 10, 2013 at 03:58 UTC
Try Regexp::Exhaustive - Find all possible matches, including backtracked and overlapping, of a pattern against a string	[reply]
Re: How does one get all possible matches from regex? by 2teez (Vicar) on Dec 10, 2013 at 06:13 UTC
You don't want to be using '$&' in your regex because "..once Perl sees that you need one of $`, $&, or $' anywhere in the program, it provides them for every pattern match. This will slow down your program a bit..." -- `Programming Perl` See also:Why does using $&, $`, or $' slow my program down? Just my 2 kobo advice. If you tell me, I'll forget. If you show me, I'll remember. if you involve me, I'll understand. --- Author unknown to me	[reply]
Re^2: How does one get all possible matches from regex? by oiskuu (Hermit) on Dec 10, 2013 at 14:19 UTC
Is this still a sound advice to give these days? The doc that you link to says: As of the 5.005 release, the $& variable is no longer "expensive" the way the other two are. Why Is $& Bad? gives a better technical explanation why one might want to avoid `$&`. Even so, the issue with `$&` is perhaps more of a concern for core module writers. Can you point to some recent benchmarks? On a related note, does `study()` actually do anything in more recent perls?	[reply]


No such thing as a small change
	PerlMonks