From f4b0accddbfda5a21a68d46b1d95998266647a55 Mon Sep 17 00:00:00 2001
From: Folker Bernitt <fbernitt@thoughtworks.com>
Date: Wed, 3 Dec 2014 15:26:15 +0100
Subject: Fixed handling of mails with non us-ascii body.

---
 service/pixelated/adapter/mail.py      | 19 +++++++++++++++----
 service/test/unit/adapter/mail_test.py |  8 ++++++++
 2 files changed, 23 insertions(+), 4 deletions(-)

diff --git a/service/pixelated/adapter/mail.py b/service/pixelated/adapter/mail.py
index 472bdad4..949daa9e 100644
--- a/service/pixelated/adapter/mail.py
+++ b/service/pixelated/adapter/mail.py
@@ -22,8 +22,10 @@ import dateutil.parser as dateparser
 from pixelated.adapter.status import Status
 import pixelated.support.date
 from email.MIMEMultipart import MIMEMultipart
-from email.MIMEText import MIMEText
+from email.mime.text import MIMEText
 from pycryptopp.hash import sha256
+import re
+
 
 class Mail(object):
     @property
@@ -61,10 +63,16 @@ class Mail(object):
         mime = MIMEMultipart()
         for key, value in self.headers.items():
             mime[str(key)] = str(value)
-        mime.attach(MIMEText(self.body, 'plain'))
+        mime.attach(MIMEText(self.body, 'plain', self._charset()))
         self._mime = mime
         return mime
 
+    def _charset(self):
+        if 'content_type' in self.headers and 'charset' in self.headers['content_type']:
+            return re.compile('.*charset=(.*)').match(self.headers['content_type']).group(1)
+        else:
+            return 'us-ascii'
+
     @property
     def raw(self):
         return self._mime_multipart.as_string()
@@ -207,19 +215,22 @@ class PixelatedMail(Mail):
     @property
     def headers(self):
         _headers = {}
+        hdoc_headers = self.hdoc.content['headers']
 
         for header in ['To', 'Cc', 'Bcc']:
-            header_value = self.hdoc.content['headers'].get(header)
+            header_value = hdoc_headers.get(header)
             if not header_value:
                 continue
             _headers[header] = header_value if type(header_value) is list else header_value.split(',')
             _headers[header] = map(lambda x: x.strip(), _headers[header])
 
         for header in ['From', 'Subject']:
-            _headers[header] = self.hdoc.content['headers'].get(header)
+            _headers[header] = hdoc_headers.get(header)
         _headers['Date'] = self._get_date()
         if self.parts and len(self.parts['alternatives']) > 1:
             _headers['content_type'] = 'multipart/alternative; boundary="%s"' % self.boundary
+        elif self.hdoc.content['headers'].get('Content-Type'):
+            _headers['content_type'] = hdoc_headers.get('Content-Type')
         return _headers
 
     def _get_date(self):
diff --git a/service/test/unit/adapter/mail_test.py b/service/test/unit/adapter/mail_test.py
index 4ce398b8..e0879b44 100644
--- a/service/test/unit/adapter/mail_test.py
+++ b/service/test/unit/adapter/mail_test.py
@@ -126,6 +126,14 @@ class TestPixelatedMail(unittest.TestCase):
                 self.assertNotIn(',', address)
             self.assertEquals(4, len(mail.headers[header_label]))
 
+    def test_content_type_is_read_from_headers_for_plain_mail_when_converted_to_raw(self):
+        fdoc, hdoc, bdoc = test_helper.leap_mail(flags=['\\Recent'], body=u'some umlaut \xc3', extra_headers={'Content-Type': 'text/plain; charset=ISO-8859-1'})
+        hdoc.content['headers']['Subject'] = 'The subject'
+        hdoc.content['headers']['From'] = 'me@pixelated.org'
+        mail = PixelatedMail.from_soledad(fdoc, hdoc, bdoc, soledad_querier=self.querier)
+
+        mail.raw
+
 
 class InputMailTest(unittest.TestCase):
     mail_dict = lambda x: {
-- 
cgit v1.2.3