m// vs grep for counting occurances of a string

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

Hello Monks, Could someone explain to me what's the difference between using m// vs grep for counting the number of times a pattern appears in text. Here's a same of code to help explain my question:

#!/usr/bin/perl -w
use strict;

my $count1;
my $count2;
my $pat = "apple";
my @apples;
while (<DATA>) {
        my $line = $_;
        @apples = grep /$pat/gi, $line;
        $count1 += @apples;
        @apples = m/$pat/gi;
        $count2 += @apples;
}
print "Count one is $count1\n";
print "Count two is $count2\n";
__DATA__
awna apple apple glais
16 Uriore Rd apple
Highapplegate corner, VT. 4xyz59
[download]

... the output is ...

Count one is 3
Count two is 4
[download]

Thanks!

Comment on m// vs grep for counting occurances of a string Select or Download Code

Replies are listed 'Best First'.
Re: m// vs grep for counting occurances of a string by wind (Priest) on Aug 11, 2007 at 01:22 UTC
grep returns the elements of a list for which expression is true. In this case, your list is the array ($line), and your expression is the /$pat/gi. Therefore, @apples will simply be set to ($line) if the pattern /$pat/gi matches, otherwise will be set to nothing. Because each of your lines of data has at least one 'apple' in them, the grep count will simply equal the number of lines. - Miller	[reply]
Re: m// vs grep for counting occurances of a string by GrandFather (Saint) on Aug 11, 2007 at 01:25 UTC
Read the grep documentation. grep returns a single item ($line) regardless of the number of matches found in the regex so @apples will contain either no elements, or a single element containing $line. The regex on the other hand returns an element for each match so you get two elements in @apples for the first line where there are two matches. DWIM is Perl's answer to Gödel	[reply]
Re^2: m// vs grep for counting occurances of a string by Plankton (Vicar) on Aug 11, 2007 at 02:05 UTC
Thank you monks! You just saved me from making a embarrassing bug.	[reply]
Re: m// vs grep for counting occurances of a string by dogz007 (Scribe) on Aug 11, 2007 at 01:44 UTC
If you would like to use `grep` to find all of the occurences of `$pat` just like the `m//`, then `split` the line before passing to `grep`, as shown below: `#!/usr/bin/perl -w use strict; my $count1; my $count2; my $pat = "apple"; my @apples; while (<DATA>) { @apples = grep /$pat/, split ' ', $_; $count1 += @apples; @apples = m/$pat/gi; $count2 += @apples; } print "Count one is $count1\n"; print "Count two is $count2\n"; __DATA__ awna apple apple glais 16 Uriore Rd apple Highapplegate corner, VT. 4xyz59` [download] Prints: `Count one is 4 Count two is 4` [download]	[reply] [d/l] [select]
Re: m// vs grep for counting occurances of a string by mreece (Friar) on Aug 11, 2007 at 05:28 UTC
incidentally, if you replace `grep` with `map`, you get the behavior you were expecting (a count of 4).	[reply] [d/l] [select]


Clear questions and runnable code get the best and fastest answer
	PerlMonks