but it does not work

One of the obvious problems is that here:

    if ($h_a < $h_b){
        -1;
    }
    if ($h_a > $h_b){
        1;
    }
[download]

There should really be return statements. But all of this could be written rather compactly as $h_a <=> $h_b || $m_a <=> $m_b - but as I said before, for simple formats there's no need for that at all.

Thanks a lot moritz, I updated the messages several times, sorry for the trouble

However, I can't understand why $n is 0

my %in;

while (<$INFILE>){
m[(\d\d:\d\d) (.+)$] and push @{ $in{ $1 } }, $2 ;
print STDOUT "1 is $1\n";
print STDOUT "2 is $2\n";
}

my @keys = keys %in;
my @times = sort @keys;

foreach my $hr (@times) {
    print STDOUT "hr is $hr\n";
        my $n = $#{ $in{ $hr } };
}

__DATA__
11:10 A1
11:30 E4
11:30 Z4
11:50 H5
12:02 H6
12:25 B2
12:25 A8
12:30 F3
12:30 E7
12:50 E15
12:55 E16
[download]

Update: OP changed, so this might not be of relevance any more.

I'm sorry, but I can't read your mind. I don't see a problem in the code you show, and stating that there is one doesn't help me to identify it.

Is a cloaked ninja attacking you every time you run this program? If not, of what nature is that problem?

If your output is not what you expect, you have to tell us what your input and your expectations are.

Please read I know what I mean. Why don't you?.

Where does $INFILE come from?

---

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.

"Science is about questioning the status quo. Questioning authority".

In the absence of evidence, opinion is indistinguishable from prejudice.

"Too many [] have been sedated by an oppressive environment of political correctness and risk aversion."

However, I can't understand why $n is 0

Well, it's not always, at least not if I run your code and actually read from <DATA>. If there's one item in the array, it contains 0, which is the last index in the array. What do you want it to contain?

if ($a =~ /(\d\d):(\d\d)/){
        $h_a = $1;
        $mn_a = 2;
    }
[download]

In addition to what the others have said, shouldn't it be

$mn_a = $2
[download]

For example, my understanding is that $a and $b are magic variables used by sort. Did you ever print out what the routine received in those variables?

A better coding style might be to pass those two values as parameters to the subroutine, and to store the result of the tests in a variable that would be passed in a return statement. Then you could display the returned value as well as the parameters, $h_a, $mn_a, etc.

Thanks a lot apl

Here is the program to take into account your remarks. However, there is still a problem.

#! perl -slw
use strict;
use diagnostics;
use Data::Dump qw[ pp ];

sub by_hour{
    my ($a,$b) = @_;
    my $h_a;
    my $mn_a;
    my $h_b;
    my $mn_b;
    
    my $res;
    
    if ($a =~ /(\d\d):(\d\d)/){
            $h_a = $1;
            $mn_a = $2;
        }
    if ($b =~ /(\d\d):(\d\d)/){
            $h_a = $1;
            $mn_a = $2;
        }
    if ($h_a < $h_b){
        $res =    -1;
        }
    if ($h_a > $h_b){
        $res = 1;
        }
    if ($h_a == $h_b){
        
        if ($mn_a < $mn_b){
        $res = -1;
        }
        if ($mn_a > $mn_b){
        $res =1;
        }
        if ($mn_a == $mn_b){
        $res =    0;
        }
    }
    return $res;    
}

my %in;

while (<DATA>){
m[(\d\d:\d\d) (.+)$] and push @{ $in{ $1 } }, $2;
}

my @keys = keys %in;
my @times = sort @keys;

for my $hr ( sort by_hour keys %in ) {
    print STDOUT "$hr\n";
}


__DATA__
11:10 A1
12:30 E4
11:30 Z4
15:50 H5
12:02 H6
12:25 B2
11:25 A8
12:30 F3
14:30 E7
12:50 E15
12:55 E16
[download]

You still don't have the print statements showing $a, $b, $h_a, $h_b, $mn_a, $mn_b, $res.

Notice also that you're processing two strings even if they're not formatted correctly (that is, match neither of the =~ statements, the second of which should be an else.