Bug #71898 for Mail-DeliveryStatus-BounceParser: new spam case

New spam case. Patch attached or in gitorious.

From d86d5686e1314a33798644f48971df428cc8c648 Mon Sep 17 00:00:00 2001 From: Michael Stevens <mstevens@etla.org> Date: Mon, 24 Oct 2011 17:05:42 +0100 Subject: [PATCH] Add more spam cases. --- MANIFEST | 2 + lib/Mail/DeliveryStatus/BounceParser.pm | 3 +- t/corpus/spam-rejection23.msg | 99 +++++++++++++++++++++++++++++++ t/spam-rejection23.t | 29 +++++++++ 4 files changed, 132 insertions(+), 1 deletions(-) create mode 100644 t/corpus/spam-rejection23.msg create mode 100644 t/spam-rejection23.t diff --git a/MANIFEST b/MANIFEST index 8f1575c..06f6339 100644 --- a/MANIFEST +++ b/MANIFEST @@ -91,6 +91,7 @@ t/corpus/spam-rejection19.msg t/corpus/spam-rejection20.msg t/corpus/spam-rejection21.msg t/corpus/spam-rejection22.msg +t/corpus/spam-rejection23.msg t/corpus/spam-rejection-uribl.msg t/corpus/spam-with-badly-parsed-email.msg t/corpus/spam-with-image.msg @@ -167,6 +168,7 @@ t/spam-rejection19.t t/spam-rejection20.t t/spam-rejection21.t t/spam-rejection22.t +t/spam-rejection23.t t/spam-rejection-uribl.t t/spam-with-image.t t/spamassassin.t diff --git a/lib/Mail/DeliveryStatus/BounceParser.pm b/lib/Mail/DeliveryStatus/BounceParser.pm index 79055c7..47ada91 100644 --- a/lib/Mail/DeliveryStatus/BounceParser.pm +++ b/lib/Mail/DeliveryStatus/BounceParser.pm @@ -1040,7 +1040,8 @@ sub _std_reason { /matches\s+a\s+profile\s+the\s+Internet\s+community\s+may\s+consider\s+spam/i or /accepted\s+due\s+to\s+spam\s+filter/i or /content\s+filter\s+rejection/i or - /using\s+a\s+mass\s+mailer/i + /using\s+a\s+mass\s+mailer/i or + /Spam\s+email/i ) { return "spam"; } diff --git a/t/corpus/spam-rejection23.msg b/t/corpus/spam-rejection23.msg new file mode 100644 index 0000000..e1276b2 --- /dev/null +++ b/t/corpus/spam-rejection23.msg @@ -0,0 +1,99 @@ +Delivered-To: automated-bounces@email.example.com +Received: by 10.142.82.8 with SMTP id f8cs13791wfb; + Sat, 22 Oct 2011 16:05:56 -0700 (PDT) +Received: by 10.227.19.203 with SMTP id c11mr7114016wbb.107.1319324754892; + Sat, 22 Oct 2011 16:05:54 -0700 (PDT) +Return-Path: <> +Received: from admin1.example.co.uk (admin1.example.co.uk. [94.236.45.212]) + by mx.google.com with ESMTP id 18si2548262wbx.45.2011.10.22.16.05.54; + Sat, 22 Oct 2011 16:05:54 -0700 (PDT) +Received-SPF: pass (google.com: domain of admin1.example.co.uk designates 94.236.45.212 as permitted sender) client-ip=94.236.45.212; +Authentication-Results: mx.google.com; spf=pass (google.com: domain of admin1.example.co.uk designates 94.236.45.212 as permitted sender) smtp.mail= +Received: by admin1.example.co.uk (Postfix) + id ED21C651165; Sun, 23 Oct 2011 00:05:53 +0100 (BST) +Date: Sun, 23 Oct 2011 00:05:53 +0100 (BST) +From: MAILER-DAEMON@admin1.example.co.uk (Mail Delivery System) +Subject: Undelivered Mail Returned to Sender +To: automated-bounces@email.example.com +Auto-Submitted: auto-replied +MIME-Version: 1.0 +Content-Type: multipart/report; report-type=delivery-status; + boundary="5822365117D.1319324753/admin1.example.co.uk" +Content-Transfer-Encoding: 8bit +Message-Id: <20111022230553.ED21C651165@admin1.example.co.uk> + +This is a MIME-encapsulated message. + +--5822365117D.1319324753/admin1.example.co.uk +Content-Description: Notification +Content-Type: text/plain; charset=us-ascii + +This is the mail system at host admin1.example.co.uk. + +I'm sorry to have to inform you that your message could not +be delivered to one or more recipients. It's attached below. + +For further assistance, please send mail to <postmaster> + +If you do so, please include this problem report. You can +delete your own text from the attached returned message. + + The mail system + +<Fred.Bloggs@example.net>: host mailgate.test.com[217.158.5.8] + said: 554 Spam email. Email Session ID: {4EA34C51-0-5059ED9-FFFF} (in reply + to end of DATA command) + +--5822365117D.1319324753/admin1.example.co.uk +Content-Description: Delivery report +Content-Type: message/delivery-status + +Reporting-MTA: dns; admin1.example.co.uk +X-Postfix-Queue-ID: 5822365117D +X-Postfix-Sender: rfc822; automated-bounces@email.example.com +Arrival-Date: Sun, 23 Oct 2011 00:05:53 +0100 (BST) + +Final-Recipient: rfc822; Fred.Bloggs@example.net +Action: failed +Status: 5.0.0 +Remote-MTA: dns; mailgate.test.com +Diagnostic-Code: smtp; 554 Spam email. Email Session ID: + {4EA34C51-0-5059ED9-FFFF} + +--5822365117D.1319324753/admin1.example.co.uk +Content-Description: Undelivered Message +Content-Type: message/rfc822 +Content-Transfer-Encoding: 8bit + +Received: by admin1.example.co.uk (Postfix, from userid 507) + id 5822365117D; Sun, 23 Oct 2011 00:05:53 +0100 (BST) +MIME-Version: 1.0 +Content-Transfer-Encoding: binary +Content-Type: multipart/alternative; boundary="_----------=_131932475319825307" +X-Mailer: MIME::Lite 3.027 (F2.78; T1.31; A2.07; B3.13; Q3.13) +Date: Sun, 23 Oct 2011 00:05:53 +0100 +From: John Smith <john.smith@example.com> +To: Fred.Bloggs@example.net, ms@example.com +Subject: foo +Message-Id: <20111022230553.5822365117D@admin1.example.co.uk> + +This is a multi-part message in MIME format. + +--_----------=_131932475319825307 +Content-Disposition: inline +Content-Transfer-Encoding: 8bit +Content-Type: text/plain + +Text + +--_----------=_131932475319825307 +Content-Disposition: inline +Content-Transfer-Encoding: 8bit +Content-Type: text/html + +HTML + +--_----------=_131932475319825307-- + + +--5822365117D.1319324753/admin1.example.co.uk-- diff --git a/t/spam-rejection23.t b/t/spam-rejection23.t new file mode 100644 index 0000000..f2c9c05 --- /dev/null +++ b/t/spam-rejection23.t @@ -0,0 +1,29 @@ +#!perl -wT +use strict; + +use Test::More tests => 3; + +use Mail::DeliveryStatus::BounceParser; + +# FH because we're being backcompat to pre-lexical +sub readfile { + my $fn = shift; + open FH, "$fn" or die $!; + local $/; + my $text = <FH>; + close FH; + return $text; +} + +my $message = readfile('t/corpus/spam-rejection23.msg'); + +my $bounce = Mail::DeliveryStatus::BounceParser->new($message); + +isa_ok($bounce, 'Mail::DeliveryStatus::BounceParser'); +ok($bounce->is_bounce, "This is a bounce"); + +my ($report) = $bounce->reports; + +my $std_reason = $report->get("std_reason"); + +is($std_reason, "spam", "std reason is spam"); -- 1.7.5.4