#!/usr/bin/perl
use warnings;
use strict;

my %violations_by_category;
my $category;
while (<DATA>) {
    if (/^([A-Z])_[0-9]+:/) {
        $category = $1;

    } elsif (/([0-9]+) violations?/) {
        $violations_by_category{$category} += $1;
    }
}

for my $category (keys %violations_by_category) {
    print $category, " ", $violations_by_category{$category}, "\n";
}

__DATA__
A_01: xxxxxxx
xxxxxxxxxx
xxx......... 1 violation

A_02: xxxxxxx
xxxxxxxxxx
xxx......... 4 violations
B_02: xxxxxxxx
xxxxxxxxxx
xxx......... 3 violations
[download]

($q=q:Sq=~/;[c](.)(.)/;chr(-||-|5+lengthSq)`"S|oS2"`map{chr |+ord
}map{substrSq`S_+|`|}3E|-|`7**2-3:)=~y+S|`+$1,++print+eval$q,q,a,
[download]

#!/usr/bin/perl

# http://perlmonks.org/?node_id=1194202

use strict;
use warnings;

my %totals;
local $/ = 'violation';

/^([A-Z])(?:_\d+).* (\d+) violation/sm and $totals{$1} += $2 while <DA
+TA>;

print "$_ number of violations = $totals{$_};\n" for sort keys %totals
+;

__DATA__
A_01: xxxxxxx
xxxxxxxxxx
xxx......... 1 violation

A_02: xxxxxxx
xxxxxxxxxx
xxx......... 4 violations
B_02: xxxxxxxx
xxxxxxxxxx
xxx......... 3 violations
[download]

I must admit that I do not well understand the question; is the following?

> But, my code is only able to read the violations all though, and sum it out.

So specify a bit what you want to implement in your program.

Anyway I have some general hints: always, that means ALWAYS, use strict; use warnings; this force you to write more robust program. strict force you to declare variables with my and this force you to limit the scope of such variables to the minimum required. This is good.

Second: use a more modern and safer way to open your filhadles: open(DATA, "<abc.txt, $!"; is a non sense..

Always use lexical filehandles (not bareword ones) like:

my $file_path = '/some/path/file.txt';

open my $fh, '<', $file_path or die "Impossible read from [$file_path]
+! $!";

# read the file and close it as soon as you do not need it anymore

close $fh or die "error closing filehandle! $!";
[download]

More; your bareword filehandle DATA is very risky: infact DATA is a special filehandle referring to the same program you are running and left open for read pointing just after the special token __DATA__ so do not use it anyway (use lexical filehandle and stop).

The DATA special filehandle is used this way:

use strict;
use warnings;

while (<DATA>){
  
     chomp;
     print "$_ ";
}
print "\n";


__DATA__
welcome
to 
Perl
[download]

L*

Hello DespacitoPerl,

Welcome to the monastery. Well your question can be resolved in many ways. Since I am not good with perlre I came up with another solution.

What I did, is first split each line (based on white space) of given input data (e.g. in.txt). As a second step I keep only the first character of the string (e.g. A_01), alternative ways of doing exactly the same thing What's the best way to get first character of the string?. Then I simply put together all data in a string with the help of join and finally push all strings into an array to have them all together.

I am using Data::Dumper to view the array since it helps me a lot to debug arrays and hashes.

Sample of solution with output and input: See update bellow

Update: I read your question again and I understood that you want to sum all violations together and also based on key (e.g. A_01, A_02 etc.). I have posted an updated code bellow using hashes.

Update 2: I was thinking that maybe on my previous Update I did not used the correct input data. So I created another script that does what you asked for. Solution provided bellow:

Hope this helps, BR.