From: Christian Herdtweck <christian.herdtweck@intra2net.com>
Date: Mon, 15 Jul 2019 11:04:03 +0000 (+0200)
Subject: Parse email from bytes instead of text
X-Git-Tag: v1.6~1^2~2
X-Git-Url: http://developer.intra2net.com/git/?a=commitdiff_plain;h=1d21262cc9765b2cf8192753fb148fb692a7de88;p=pyi2ncommon

Parse email from bytes instead of text

This allows usage of other encodings than the system encoding (e.g. latin1)
---

diff --git a/src/mail_utils.py b/src/mail_utils.py
index 9efc632..19adb76 100644
--- a/src/mail_utils.py
+++ b/src/mail_utils.py
@@ -42,7 +42,7 @@ from base64 import b64decode
 import re
 import logging
 from email.utils import parsedate_to_datetime
-from email.parser import Parser
+from email.parser import BytesParser
 
 from . import arnied_wrapper
 
@@ -127,7 +127,7 @@ def parse_mail_file(file_name, headers_only=True, attachment_filenames=False):
     """
     Parse given email file (e.g. a banned message).
 
-    This is basically a `email.parser.Parser().parse(...)` with given
+    This is basically a `email.parser.BytesParser().parse(...)` with given
     `headers_only`, that can handle BSMTP. As an extra bonus, you can just
     request headers plus the names of attached files.
 
@@ -149,27 +149,27 @@ def parse_mail_file(file_name, headers_only=True, attachment_filenames=False):
     """
     msg = None
     start_pos = 0
-    with open(file_name, 'r') as read_handle:
+    with open(file_name, 'rb') as read_handle:
         line = read_handle.readline()
-        if line.startswith('EHLO'):
+        if line.startswith(b'EHLO'):
             # there is a smtp header. skip to its end
-            while line.strip() != 'DATA':
+            while line.strip() != b'DATA':
                 line = read_handle.readline()
             # the rest is the email plus a trailing '.' (ignored by parser if
             # multipart)
         else:
             read_handle.seek(0)  # forget we read the first line already
         start_pos = read_handle.tell()
-        msg = Parser().parse(read_handle, headersonly=headers_only)
+        msg = BytesParser().parse(read_handle, headersonly=headers_only)
 
     if not attachment_filenames:
         return msg
 
     # otherwise need to parse complete message to get attachment file names
     if headers_only:
-        with open(file_name, 'r') as read_handle:
+        with open(file_name, 'rb') as read_handle:
             read_handle.seek(start_pos)
-            full_msg = Parser().parse(read_handle, headersonly=False)
+            full_msg = BytesParser().parse(read_handle, headersonly=False)
     else:
         full_msg = msg
     filenames = [get_filename(part) for part in full_msg.walk()]