-# Copyright (C) 2014, Eric Wong <normalperson@yhbt.net> and all contributors
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2014-2019 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
use strict;
use warnings;
use Test::More;
use Email::MIME;
use File::Temp qw/tempdir/;
-use Cwd;
-use IPC::Run qw(run);
+use Cwd qw(getcwd);
+use PublicInbox::MID qw(mid2path);
+eval { require IPC::Run };
+plan skip_all => "missing IPC::Run for t/mda.t" if $@;
my $mda = "blib/script/public-inbox-mda";
-my $tmpdir = tempdir(CLEANUP => 1);
+my $learn = "blib/script/public-inbox-learn";
+my $tmpdir = tempdir('pi-mda-XXXXXX', TMPDIR => 1, CLEANUP => 1);
my $home = "$tmpdir/pi-home";
my $pi_home = "$home/.public-inbox";
my $pi_config = "$pi_home/config";
my $fail_path = "$fail_bin:$ENV{PATH}"; # for spamc spam mock
my $addr = 'test-public@example.com';
my $cfgpfx = "publicinbox.test";
-my $failbox = "$home/fail.mbox";
+my $faildir = "$home/faildir/";
+my $mime;
{
ok(-x "$main_bin/spamc",
}
}
+local $ENV{GIT_COMMITTER_NAME} = eval {
+ use PublicInbox::MDA;
+ use PublicInbox::Address;
+ use Encode qw/encode/;
+ my $mbox = 't/utf8.mbox';
+ open(my $fh, '<', $mbox) or die "failed to open mbox: $mbox\n";
+ my $str = eval { local $/; <$fh> };
+ close $fh;
+ my $msg = Email::MIME->new($str);
+
+ my $from = $msg->header('From');
+ my ($author) = PublicInbox::Address::names($from);
+ my ($email) = PublicInbox::Address::emails($from);
+ my $date = $msg->header('Date');
+
+ is('Eléanor',
+ encode('us-ascii', my $tmp = $author, Encode::HTMLCREF),
+ 'HTML conversion is correct');
+ is($email, 'e@example.com', 'email parsed correctly');
+ is($date, 'Thu, 01 Jan 1970 00:00:00 +0000',
+ 'message date parsed correctly');
+ $author;
+};
+die $@ if $@;
+
{
my $good_rev;
- local $ENV{PI_FAILBOX} = $failbox;
+ local $ENV{PI_EMERGENCY} = $faildir;
local $ENV{HOME} = $home;
- local $ENV{RECIPIENT} = $addr;
+ local $ENV{ORIGINAL_RECIPIENT} = $addr;
my $simple = Email::Simple->new(<<EOF);
From: Me <me\@example.com>
To: You <you\@example.com>
# ensure successful message delivery
{
local $ENV{PATH} = $main_path;
- run([$mda], \$in);
- local $ENV{GIT_DIR} = $maindir;
- my $rev = `git rev-list HEAD`;
+ IPC::Run::run([$mda], \$in);
+ my $rev = `git --git-dir=$maindir rev-list HEAD`;
like($rev, qr/\A[a-f0-9]{40}/, "good revision committed");
chomp $rev;
- my $cmt = `git cat-file commit $rev`;
+ my $cmt = `git --git-dir=$maindir cat-file commit $rev`;
like($cmt, qr/^author Me <me\@example\.com> 0 \+0000\n/m,
"author info set correctly");
like($cmt, qr/^committer test <test-public\@example\.com>/m,
# ensure failures work, fail with bad spamc
{
- ok(!-e $failbox, "nothing in PI_FAILBOX before");
+ my @prev = <$faildir/new/*>;
+ is(scalar @prev, 0 , "nothing in PI_EMERGENCY before");
local $ENV{PATH} = $fail_path;
- run([$mda], \$in);
- local $ENV{GIT_DIR} = $maindir;
- my @revs = `git rev-list HEAD`;
+ IPC::Run::run([$mda], \$in);
+ my @revs = `git --git-dir=$maindir rev-list HEAD`;
is(scalar @revs, 1, "bad revision not committed");
- ok(-s $failbox > 0, "PI_FAILBOX is written to");
+ my @new = <$faildir/new/*>;
+ is(scalar @new, 1, "PI_EMERGENCY is written to");
}
fail_bad_header($good_rev, "bad recipient", <<"");
}
+# spam training
+{
+ local $ENV{PI_EMERGENCY} = $faildir;
+ local $ENV{HOME} = $home;
+ local $ENV{ORIGINAL_RECIPIENT} = $addr;
+ local $ENV{PATH} = $main_path;
+ my $mid = 'spam-train@example.com';
+ my $simple = Email::Simple->new(<<EOF);
+From: Spammer <spammer\@example.com>
+To: You <you\@example.com>
+Cc: $addr
+Message-ID: <$mid>
+Subject: this message will be trained as spam
+Date: Thu, 01 Jan 1970 00:00:00 +0000
+
+EOF
+ my $in = $simple->as_string;
+
+ {
+ # deliver the spam message, first
+ IPC::Run::run([$mda], \$in);
+ my $path = mid2path($mid);
+ my $msg = `git --git-dir=$maindir cat-file blob HEAD:$path`;
+ like($msg, qr/\Q$mid\E/, "message delivered");
+
+ # now train it
+ local $ENV{GIT_AUTHOR_EMAIL} = 'trainer@example.com';
+ local $ENV{GIT_COMMITTER_EMAIL} = 'trainer@example.com';
+ local $ENV{GIT_COMMITTER_NAME} = undef;
+ IPC::Run::run([$learn, "spam"], \$msg);
+ is($?, 0, "no failure from learning spam");
+ IPC::Run::run([$learn, "spam"], \$msg);
+ is($?, 0, "no failure from learning spam idempotently");
+ }
+}
+
+# train ham message
+{
+ local $ENV{PI_EMERGENCY} = $faildir;
+ local $ENV{HOME} = $home;
+ local $ENV{ORIGINAL_RECIPIENT} = $addr;
+ local $ENV{PATH} = $main_path;
+ my $mid = 'ham-train@example.com';
+ my $simple = Email::Simple->new(<<EOF);
+From: False-positive <hammer\@example.com>
+To: You <you\@example.com>
+Cc: $addr
+Message-ID: <$mid>
+Subject: this message will be trained as spam
+Date: Thu, 01 Jan 1970 00:00:00 +0000
+
+EOF
+ my $in = $simple->as_string;
+
+ # now train it
+ # these should be overridden
+ local $ENV{GIT_AUTHOR_EMAIL} = 'trainer@example.com';
+ local $ENV{GIT_COMMITTER_EMAIL} = 'trainer@example.com';
+
+ IPC::Run::run([$learn, "ham"], \$in);
+ is($?, 0, "learned ham without failure");
+ my $path = mid2path($mid);
+ my $msg = `git --git-dir=$maindir cat-file blob HEAD:$path`;
+ like($msg, qr/\Q$mid\E/, "ham message delivered");
+ IPC::Run::run([$learn, "ham"], \$in);
+ is($?, 0, "learned ham idempotently ");
+
+ # ensure trained email is filtered, too
+ my $html_body = "<html><body>hi</body></html>";
+ my $parts = [
+ Email::MIME->create(
+ attributes => {
+ content_type => 'text/html; charset=UTF-8',
+ encoding => 'base64',
+ },
+ body => $html_body,
+ ),
+ Email::MIME->create(
+ attributes => {
+ content_type => 'text/plain',
+ encoding => 'quoted-printable',
+ },
+ body => 'hi = "bye"',
+ )
+ ];
+ $mid = 'multipart-html-sucks@11';
+ $mime = Email::MIME->create(
+ header_str => [
+ From => 'a@example.com',
+ Subject => 'blah',
+ Cc => $addr,
+ 'Message-ID' => "<$mid>",
+ 'Content-Type' => 'multipart/alternative',
+ ],
+ parts => $parts,
+ );
+
+ {
+ $in = $mime->as_string;
+ IPC::Run::run([$learn, "ham"], \$in);
+ is($?, 0, "learned ham without failure");
+ my $path = mid2path($mid);
+ $msg = `git --git-dir=$maindir cat-file blob HEAD:$path`;
+ like($msg, qr/<\Q$mid\E>/, "ham message delivered");
+ unlike($msg, qr/<html>/i, '<html> filtered');
+ }
+}
+
+# List-ID based delivery
+{
+ local $ENV{PI_EMERGENCY} = $faildir;
+ local $ENV{HOME} = $home;
+ local $ENV{ORIGINAL_RECIPIENT} = undef;
+ local $ENV{PATH} = $main_path;
+ my $list_id = 'foo.example.com';
+ my $mid = 'list-id-delivery@example.com';
+ my $simple = Email::Simple->new(<<EOF);
+From: user <user\@example.com>
+To: You <you\@example.com>
+Cc: $addr
+Message-ID: <$mid>
+List-Id: <$list_id>
+Subject: this message will be trained as spam
+Date: Thu, 01 Jan 1970 00:00:00 +0000
+
+EOF
+ system(qw(git config --file), $pi_config, "$cfgpfx.listid", $list_id);
+ $? == 0 or die "failed to set listid $?";
+ my $in = $simple->as_string;
+ IPC::Run::run([$mda], \$in);
+ is($?, 0, 'mda OK with List-Id match');
+ my $path = mid2path($mid);
+ my $msg = `git --git-dir=$maindir cat-file blob HEAD:$path`;
+ like($msg, qr/\Q$list_id\E/, 'delivered message w/ List-ID matches');
+
+ # try a message w/o precheck
+ $simple = Email::Simple->new(<<EOF);
+To: You <you\@example.com>
+List-Id: <$list_id>
+
+this message would not be accepted without --no-precheck
+EOF
+ $in = $simple->as_string;
+ my ($out, $err) = ('', '');
+ IPC::Run::run([$mda, '--no-precheck'], \$in, \$out, \$err);
+ is($?, 0, 'mda OK with List-Id match and --no-precheck');
+ my $cur = `git --git-dir=$maindir diff HEAD~1..HEAD`;
+ like($cur, qr/this message would not be accepted without --no-precheck/,
+ '--no-precheck delivered message anyways');
+}
+
+done_testing();
+
sub fail_bad_header {
my ($good_rev, $msg, $in) = @_;
- open my $fh, '>', $failbox or die "failed to open $failbox: $!\n";
- close $fh or die "failed to close $failbox: $!\n";
+ my @f = glob("$faildir/*/*");
+ unlink @f if @f;
my ($out, $err) = ("", "");
local $ENV{PATH} = $main_path;
- run([$mda], \$in, \$out, \$err);
- local $ENV{GIT_DIR} = $maindir;
- my $rev = `git rev-list HEAD`;
+ IPC::Run::run([$mda], \$in, \$out, \$err);
+ my $rev = `git --git-dir=$maindir rev-list HEAD`;
chomp $rev;
is($rev, $good_rev, "bad revision not commited ($msg)");
- ok(-s $failbox > 0, "PI_FAILBOX is written to ($msg)");
+ @f = glob("$faildir/*/*");
+ is(scalar @f, 1, "faildir written to");
[ $in, $out, $err ];
}
-
-done_testing();