huge refactor of encoding handling

[public-inbox.git] / t / mda.t

diff --git a/t/mda.t b/t/mda.t
index d37c03cd32305c5def7478b03dddb90ed8c59813..fad96e5bf44eb733600f5369c4dfed7130559a3b 100644 (file)
--- a/t/mda.t
+++ b/t/mda.t
@@ -4,11 +4,13 @@ use strict;
 use warnings;
 use Test::More;
 use Email::MIME;
+use Email::Filter;
 use File::Temp qw/tempdir/;
 use Cwd;
 use IPC::Run qw(run);
 
 my $mda = "blib/script/public-inbox-mda";
+my $learn = "blib/script/public-inbox-learn";
 my $tmpdir = tempdir(CLEANUP => 1);
 my $home = "$tmpdir/pi-home";
 my $pi_home = "$home/.public-inbox";
@@ -42,11 +44,31 @@ my $failbox = "$home/fail.mbox";
        }
 }
 
+local $ENV{GIT_COMMITTER_NAME} = eval {
+       use PublicInbox::MDA;
+       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::Filter->new(data => $str);
+       $msg = Email::MIME->new($msg->simple->as_string);
+       my ($author, $email, $date) = PublicInbox::MDA->author_info($msg);
+       is('El&#233;anor',
+               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} = $failbox;
        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>
@@ -76,13 +98,13 @@ EOF
 
        # ensure failures work, fail with bad spamc
        {
-               ok(!-e $failbox, "nothing in PI_FAILBOX before");
+               ok(!-e $failbox, "nothing in PI_EMERGENCY before");
                local $ENV{PATH} = $fail_path;
                run([$mda], \$in);
                local $ENV{GIT_DIR} = $maindir;
                my @revs = `git rev-list HEAD`;
                is(scalar @revs, 1, "bad revision not committed");
-               ok(-s $failbox > 0, "PI_FAILBOX is written to");
+               ok(-s $failbox > 0, "PI_EMERGENCY is written to");
        }
 
        fail_bad_header($good_rev, "bad recipient", <<"");
@@ -130,6 +152,93 @@ Date: deadbeef
 
 }
 
+# spam training
+{
+       local $ENV{PI_EMERGENCY} = $failbox;
+       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 = $simple->as_string;
+
+       {
+               # deliver the spam message, first
+               run([$mda], \$in);
+               my $msg = `ssoma cat $mid $maindir`;
+               like($msg, qr/\Q$mid\E/, "message delivered");
+
+               # now train it
+               local $ENV{GIT_AUTHOR_EMAIL} = 'trainer@example.com';
+               local $ENV{GIT_COMMITTER_EMAIL} = 'trainer@example.com';
+               run([$learn, "spam"], \$msg);
+               is($?, 0, "no failure from learning spam");
+               run([$learn, "spam"], \$msg);
+               is($?, 0, "no failure from learning spam idempotently");
+       }
+}
+
+# train ham message
+{
+       local $ENV{PI_EMERGENCY} = $failbox;
+       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
+       local $ENV{GIT_AUTHOR_EMAIL} = 'trainer@example.com';
+       local $ENV{GIT_COMMITTER_EMAIL} = 'trainer@example.com';
+       run([$learn, "ham"], \$in);
+       is($?, 0, "learned ham without failure");
+       my $msg = `ssoma cat $mid $maindir`;
+       like($msg, qr/\Q$mid\E/, "ham message delivered");
+       run([$learn, "ham"], \$in);
+       is($?, 0, "learned ham idempotently ");
+}
+
+# faildir - emergency destination is maildir
+{
+       my $faildir= "$home/faildir/";
+       local $ENV{PI_EMERGENCY} = $faildir;
+       local $ENV{HOME} = $home;
+       local $ENV{ORIGINAL_RECIPIENT} = $addr;
+       local $ENV{PATH} = $fail_path;
+       my $in = <<EOF;
+From: Faildir <faildir\@example.com>
+To: You <you\@example.com>
+Cc: $addr
+Message-ID: <faildir\@example.com>
+Subject: faildir subject
+Date: Thu, 01 Jan 1970 00:00:00 +0000
+
+EOF
+       run([$mda], \$in);
+       ok(-d $faildir, "emergency exists");
+       my @new = glob("$faildir/new/*");
+       is(scalar(@new), 1, "message delivered");
+}
+
+done_testing();
+
 sub fail_bad_header {
        my ($good_rev, $msg, $in) = @_;
        open my $fh, '>', $failbox or die "failed to open $failbox: $!\n";
@@ -141,8 +250,6 @@ sub fail_bad_header {
        my $rev = `git rev-list HEAD`;
        chomp $rev;
        is($rev, $good_rev, "bad revision not commited ($msg)");
-       ok(-s $failbox > 0, "PI_FAILBOX is written to ($msg)");
+       ok(-s $failbox > 0, "PI_EMERGENCY is written to ($msg)");
        [ $in, $out, $err ];
 }
-
-done_testing();