Merge pull request #96 from GetHappie/unicode-decode-fix

Fix UnicecodeDecodeError parsing email

imbox/parser.py

@@ -121,7 +121,7 @@ def decode_content(message):
 
 def parse_email(raw_email, policy=None):
     if isinstance(raw_email, bytes):
-        raw_email = str_encode(raw_email, 'utf-8')
+        raw_email = str_encode(raw_email, 'utf-8', errors='ignore')
     if policy is not None:
         email_parse_kwargs = dict(policy=policy)
     else:

tests/8422.msg

@@ -0,0 +1,22 @@
+Delivered-To: receiver@example.com
+Return-Path: <sender@example.com>
+Date: Thu, 20 Jul 2017 07:34:22 -0500
+Message-ID: <59705CFE.A95F.0016.0@journeys.com>
+Subject: Following up Re: Looking to connect, let's schedule a call!
+From: sender@example.com
+To: "Receiver" <receiver@example.com>
+Mime-Version: 1.0
+Content-Type: multipart/mixed; boundary="=__PartBD85995F.0__="
+
+This is a MIME message. If you are reading this text, you may want to 
+consider changing to a mail reader or gateway that understands how to 
+properly handle MIME multipart messages.
+
+--=__PartBD85995F.0__=
+Content-Type: multipart/alternative; boundary="=__PartBD85995F.1__="
+
+--=__PartBD85995F.1__=
+Content-Type: text/plain; charset=Windows-1252
+Content-Transfer-Encoding: 8bit
+
+Following up on my previous message. I�d love to connect you with

tests/parser_tests.py

@@ -1,13 +1,17 @@
 import unittest
 from imbox.parser import *
 
+import os
 import sys
 if sys.version_info.minor < 3:
     SMTP = False
 else:
     from email.policy import SMTP
 
 
+TEST_DIR = os.path.dirname(os.path.abspath(__file__))
+
+
 raw_email = """Delivered-To: johndoe@gmail.com
 X-Originating-Email: [martin@amon.cx]
 Message-ID: <test0@example.com>
@@ -192,6 +196,10 @@ def test_parse_email_encoded(self):
         self.assertEqual('Выписка по карте', parsed_email.subject)
         self.assertEqual('Выписка по карте 1234', parsed_email.body['html'][0])
 
+    def test_parse_email_invalid_unicode(self):
+        parsed_email = parse_email(open(os.path.join(TEST_DIR, '8422.msg'), 'rb').read())
+        self.assertEqual("Following up Re: Looking to connect, let's schedule a call!", parsed_email.subject)
+
     def test_parse_email_inline_body(self):
         parsed_email = parse_email(raw_email_encoded_another_bad_multipart)
         self.assertEqual("Re: Reaching Out About Peoples Home Equity", parsed_email.subject)