From 9316364e9deb206cff63afc0b8ca6f919b4f63e0 Mon Sep 17 00:00:00 2001
From: Paul Buetow <paul@buetow.org>
Date: Mon, 2 Jan 2023 14:20:48 +0200
Subject: initial script, but WIP

---
 README.md   |  2 +-
 foostats.pl | 66 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 67 insertions(+), 1 deletion(-)
 create mode 100644 foostats.pl

diff --git a/README.md b/README.md
index 13f65a6..b9e747d 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,3 @@
 # foostats
 
-Small Perl script reporting anonymous site stats for my foo.zone web and gemini capsule. 
\ No newline at end of file
+Small Perl script reporting anonymous site stats for my foo.zone web and gemini capsule running on OpenBSD using `httpd` webserver and `relayd`, `inetd` + `vger` for Gemini. 
diff --git a/foostats.pl b/foostats.pl
new file mode 100644
index 0000000..2d82916
--- /dev/null
+++ b/foostats.pl
@@ -0,0 +1,66 @@
+#!/usr/bin/perl
+
+use v5.32;
+use strict;
+use warnings;
+use feature 'signatures';
+no warnings 'experimental::signatures';
+
+use constant {
+  WWW_LOGS_GLOB => '/var/www/logs/access.log*',
+};
+
+use Data::Dumper;
+use Time::Piece;
+use Digest::SHA3 'sha3_512_base64';
+use PerlIO::gzip;
+
+sub anonymize_ip ($ip) {
+  my $ip_proto = (index $ip, ':') == -1 ? 'ipv4' : 'ipv6';
+  my $ip_hash = sha3_512_base64 $ip;
+  return ($ip_hash, $ip_proto);
+}
+
+sub process_lines ($glob, $cb) {
+  my sub open_file ($path) {
+    my $flag = $path =~ /\.gz$/ ? '<:gzip' : '<';
+    open my $file, $flag, $path or die $!;
+    return $file;
+  }
+
+  for my $path (glob $glob) {
+    my $file = open_file $path;
+    $cb->(split / +/) while <$file>;
+    close $file;
+  }
+}
+
+sub process_www_logs ($cb) {
+  my sub parse_date ($date) {
+    eval {
+      Time::Piece->strptime($date, '[%d/%b/%Y:%H:%M:%S')->strftime('%Y-%m-%d')
+    }
+  }
+
+  my sub parse_line (@line) {
+    my ($ip_hash, $ip_proto) = anonymize_ip $line[1];
+    {
+      host => $line[0],
+      ip_hash => $ip_hash,
+      ip_proto => $ip_proto,
+      date => parse_date($line[4]),
+      uripath => $line[7],
+      status => $line[9],
+      #line => \@line,
+    }
+  }
+
+  process_lines WWW_LOGS_GLOB, sub (@line) {
+    $cb->(parse_line @line);
+  };
+}
+
+process_www_logs sub {
+  say Dumper @_;
+};
+
-- 
cgit v1.2.3