From eab11ed53e4094ba97e4365de518b3e5af918bb5 Mon Sep 17 00:00:00 2001
From: medmunds <medmunds@gmail.com>
Date: Sat, 24 Mar 2018 17:46:18 -0700
Subject: [PATCH] Inbound: test parsing RFC2231 MIME header parameters

And decide not to work around a Python 3.3 bug accessing MIME headers
that have non-ASCII characters in params. The bug is fixed in the
Python 3.4 email package (and didn't exist in Python 2.7). Python 3.3
was only supported with Django 1.8.
---
 tests/test_inbound.py | 33 ++++++++++++++++++++++++++++++++-
 tests/utils.py        | 17 +++++++++++++++++
 2 files changed, 49 insertions(+), 1 deletion(-)

diff --git a/tests/test_inbound.py b/tests/test_inbound.py
index 353673c..aff8215 100644
--- a/tests/test_inbound.py
+++ b/tests/test_inbound.py
@@ -2,12 +2,15 @@
 from __future__ import unicode_literals
 
 from base64 import b64encode
+from email.utils import collapse_rfc2231_value
 from textwrap import dedent
+from unittest import skipIf
 
 from django.test import SimpleTestCase
 
 from anymail.inbound import AnymailInboundMessage
-from .utils import SAMPLE_IMAGE_FILENAME, sample_image_content
+
+from .utils import SAMPLE_IMAGE_FILENAME, python_has_broken_mime_param_handling, sample_image_content
 
 SAMPLE_IMAGE_CONTENT = sample_image_content()
 
@@ -458,3 +461,31 @@ class EmailParserWorkaroundTests(SimpleTestCase):
 
         # Replace illegal encodings (rather than causing error):
         self.assertEqual(msg["X-Broken"], "Not a char: \N{REPLACEMENT CHARACTER}.")
+
+    @skipIf(python_has_broken_mime_param_handling(),
+            "This Python has a buggy email package that crashes on non-ASCII "
+            "characters in RFC2231-encoded MIME header parameters")
+    def test_parse_encoded_params(self):
+        raw = dedent("""\
+            MIME-Version: 1.0
+            Content-Type: multipart/mixed; boundary="this_is_a_boundary"
+
+            --this_is_a_boundary
+            Content-Type: text/plain; charset="UTF-8"
+
+            This is the body
+
+            --this_is_a_boundary
+            Content-Type: text/plain; name*=us-ascii''TPS%20Report
+            Content-Disposition: attachment;
+             filename*=iso-8859-1''Une%20pi%E8ce%20jointe%2Etxt
+
+            This is an attachment
+            --this_is_a_boundary--
+            """)
+        msg = AnymailInboundMessage.parse_raw_mime(raw)
+        att = msg.attachments[0]
+        self.assertTrue(att.is_attachment())
+        self.assertEqual(att.get_content_disposition(), "attachment")
+        self.assertEqual(collapse_rfc2231_value(att.get_param("Name", header="Content-Type")), "TPS Report")
+        self.assertEqual(att.get_filename(), "Une pièce jointe.txt")
diff --git a/tests/utils.py b/tests/utils.py
index 479a4e0..b833a18 100644
--- a/tests/utils.py
+++ b/tests/utils.py
@@ -222,3 +222,20 @@ class ClientWithCsrfChecks(Client):
     def __init__(self, **defaults):
         super(ClientWithCsrfChecks, self).__init__(
             enforce_csrf_checks=True, **defaults)
+
+
+def python_has_broken_mime_param_handling():
+    # In Python 3.3 (only), trying to access any parsed MIME header will crash if the header
+    # has parameters with non-ASCII characters. (Common in, e.g., attachment filenames.)
+    # The bug is somewhere within email._header_value_parser.parse_mime_parameters, and is too
+    # complicated to work around for an uncommon version combination (Django 1.8 on Python 3.3).
+    # If you run into it, please upgrade to (at least) Python 3.4.
+    try:
+        from email.policy import default
+        default.header_fetch_parse("Content-Type", "plain; name*=iso-8859-1''Une%20pi%E8ce")
+    except ImportError:
+        return False  # Python 2 (or pre Python 3.3) -- bug doesn't apply
+    except UnicodeEncodeError:
+        return True  # this is the bug
+    else:
+        return False  # worked fine