log_parser.pl in Perl

Buckets log lines by severity with one named-capture regex.

use strict;
use warnings;

# log_parser: bucket log lines by level with a named-capture regex

my $log = <<'LOG';
2026-08-02 09:14:02 INFO  server started on port 8080
2026-08-02 09:14:05 INFO  connected to database
2026-08-02 09:15:11 WARN  slow query took 1200 ms
2026-08-02 09:16:40 ERROR connection reset by peer
2026-08-02 09:16:41 INFO  retrying in 2 s
2026-08-02 09:17:03 WARN  cache miss rate above 20 pct
2026-08-02 09:18:22 ERROR disk usage at 95 pct
LOG

my $pat = qr/^\S+ (?<time>\S+) (?<level>INFO|WARN|ERROR)\s+(?<msg>.+)$/;

my %by_level;
my @errors;

for my $line (split /\n/, $log) {
    next unless $line =~ $pat;
    my ($time, $level, $msg) = @+{qw(time level msg)};
    push @{ $by_level{$level} }, $msg;
    push @errors, "$time  $msg" if $level eq "ERROR";
}

print "lines by level:\n";
for my $level (sort keys %by_level) {
    printf "  %-5s %d\n", $level, scalar @{ $by_level{$level} };
}

print "errors:\n";
print "  $_\n" for @errors;

How it works

1. qr// compiles the pattern once; (?<level>...) names each capture.
2. Matches land in %+, and a hash slice unpacks time, level, and msg.
3. push @{ $by_level{$level} } groups messages under their severity.
4. The report counts each bucket, then replays the ERROR lines with times.

Keywords and builtins used here

• for
• if
• keys
• my
• next
• print
• printf
• push
• scalar
• sort
• split
• unless
• use

The run, in numbers

Lines

34

Characters to type

942

Tokens

144

Three-star pace

85 tpm

At the three-star pace of 85 tokens a minute, this run takes about 102 seconds.

Type this snippet

Step 3 of 3 in Encore, step 27 of 27 in Language basics.

← Previous