Because the 2 code snippets arent quite the same. A better comparison would be

open(IN, "/some/file") || die "Cant open /some/file: $!\n";
while (<IN>) {
   push(@foo, $_) if ( m/^$match/ );
}
close(IN);

# as opposed to

open(IN, "/some/file") || die "Cant access /some/file: $!\n";
@foo = grep(m/^$match/, <IN>);
close(IN);
[download]

The differences in the first post between the loops are as follows:

In the while loop, it continues to iterate over the input list, each time storing the value if it matches the regex. So if you have 3 values which match, then the last value matched will be placed in the variable.

In the grep loop, due to my excessive use of parentheses I forced grep to return a list, but only captured the first element. Along the same lines as say

$f = 'foo:bar:baz';
($blah) = split(/:/, $f);
[download]

$blah now contains 'foo', bar and baz are silently discarded. So grep finds all the matches, and returns them as a list, but I only grab the first value and toss the rest away. Use the first snippet in this post as a better example of the differences. Also remember that I stated on files themselves, not on directories, as grep really begins to out perform the loop as the data set gets larger. The loop gives you a far greater granularity of control in what to do with the contents of what you are iterating over. It will be more efficient to loop over a data set once in a loop if you are planning on doing different things with different pieces of the data. If you already know you want a specific piece of data, and only need to process the data set once, then grep is your friend.

MMMMM... Chocolaty Perl Goodness.....