Merge branch 'release/3.12.0'

Author: fedelemantuano

README.md

@@ -187,7 +187,13 @@ mail.text_not_managed: all not managed text (check the warning logs to find cont
 mail.to
 mail.to_domains
 mail.timezone: returns the timezone, offset from UTC
-mail_partial: returns only the mains parts of emails
+mail.mail_partial: returns only the mains parts of emails
+```
+
+It's possible to write the attachments on disk with the method:
+
+```
+mail.write_attachments(base_path)
 ```
 
 ## Usage from command-line

mailparser/mailparser.py

@@ -42,6 +42,7 @@
     ported_open,
     ported_string,
     receiveds_parsing,
+    write_attachments,
 )
 
 from .exceptions import MailParserEnvironmentError
@@ -348,10 +349,14 @@ def parse(self):
         # walk all mail parts
         for i, p in enumerate(parts):
             if not p.is_multipart():
-                filename = decode_header_part(p.get_filename())
                 charset = p.get_content_charset('utf-8')
                 charset_raw = p.get_content_charset()
                 log.debug("Charset {!r} part {!r}".format(charset, i))
+                content_id = ported_string(p.get('content-id'))
+                log.debug("content-id {!r} part {!r}".format(
+                    content_id, i))
+                filename = decode_header_part(
+                    p.get_filename("{}".format(content_id)))
 
                 # this is an attachment
                 if filename:
@@ -365,9 +370,6 @@ def parse(self):
                         p.get('content-transfer-encoding', '')).lower()
                     log.debug("Transfer encoding {!r} part {!r}".format(
                         transfer_encoding, i))
-                    content_id = ported_string(p.get('content-id'))
-                    log.debug("content-id {!r} part {!r}".format(
-                        content_id, i))
                     content_disposition = ported_string(
                         p.get('content-disposition'))
                     log.debug("content-disposition {!r} part {!r}".format(
@@ -473,6 +475,16 @@ def get_server_ipaddress(self, trust):
                             log.debug("IP {!r} not private".format(ip_str))
                             return ip_str
 
+    def write_attachments(self, base_path):
+        """ This method writes the attachments of mail on disk
+
+        Arguments:
+            base_path {str} -- Base path where write the attachments
+        """
+        write_attachments(
+            attachments=self.attachments,
+            base_path=base_path)
+
     def __getattr__(self, name):
         name = name.strip("_").lower()
         name_header = name.replace("_", "-")

mailparser/utils.py

@@ -31,8 +31,10 @@
 import hashlib
 import logging
 import os
+import random
 import re
 import simplejson as json
+import string
 import subprocess
 import sys
 import tempfile
@@ -456,12 +458,19 @@ def get_header(message, name):
         name (string): header to get
 
     Returns:
-        decoded header
+        str if there is an header
+        list if there are more than one
     """
-    header = message.get(name)
-    log.debug("Getting header {!r}: {!r}".format(name, header))
-    if header:
-        return decode_header_part(header)
+
+    headers = message.get_all(name)
+    log.debug("Getting header {!r}: {!r}".format(name, headers))
+    if headers:
+        headers = [decode_header_part(i) for i in headers]
+        if len(headers) == 1:
+            # in this case return a string
+            return headers[0]
+        # in this case return a list
+        return headers
     return six.text_type()
 
 
@@ -551,3 +560,16 @@ def write_sample(binary, payload, path, filename):  # pragma: no cover
     else:
         with open(sample, "w") as f:
             f.write(payload)
+
+
+def random_string(string_length=10):
+    """ Generate a random string of fixed length
+
+    Keyword Arguments:
+        string_length {int} -- String length (default: {10})
+
+    Returns:
+        str -- Random string
+    """
+    letters = string.ascii_lowercase
+    return ''.join(random.choice(letters) for i in range(string_length))

mailparser/version.py

@@ -17,7 +17,7 @@
 limitations under the License.
 """
 
-__version__ = "3.11.0"
+__version__ = "3.12.0"
 
 if __name__ == "__main__":
     print(__version__)

requirements-dev.txt

@@ -1,7 +1,7 @@
 # tool
 ipaddress==1.0.23
 simplejson==3.17.0
-six==1.13.0
+six==1.14.0
 
 # dev
 coverage==5.0.2

requirements.txt

@@ -1,3 +1,3 @@
 ipaddress==1.0.23
 simplejson==3.17.0
-six==1.13.0
+six==1.14.0

setup.py

@@ -41,6 +41,7 @@
     license="Apache License, Version 2.0",
     url="https://github.com/SpamScope/mail-parser",
     long_description=long_description,
+    long_description_content_type="text/markdown",
     version=__version__,
     author="Fedele Mantuano",
     author_email="[email protected]",

tests/mails/mail_test_14

@@ -4,6 +4,8 @@ Date: Wed, 24 Apr 2019 10:05:02 +0200 (CEST)
 Mime-Version: 1.0
 Content-Type: multipart/mixed; boundary="===============8544575414772382491=="
 To: [email protected]
+Received-SPF: custom_header1
+Received-SPF: custom_header2
 
 --===============8544575414772382491==
 Content-Type: text/html; charset=UTF-8

tests/test_mail_parser.py

@@ -20,6 +20,7 @@
 import datetime
 import logging
 import os
+import shutil
 import six
 import sys
 import unittest
@@ -42,6 +43,7 @@
     ported_string,
     receiveds_parsing,
     parse_received,
+    random_string,
 )
 
 from mailparser.exceptions import MailParserEnvironmentError
@@ -87,6 +89,27 @@ def setUp(self):
             mail_malformed_2,
             mail_malformed_3)
 
+    def test_write_attachments(self):
+        attachments = [
+            "<_1_0B4E44A80B15F6FC005C1243C12580DD>",
+            "<_1_0B4E420C0B4E3DD0005C1243C12580DD>",
+            "<_1_0B4E24640B4E1564005C1243C12580DD>",
+            "Move To Eight ZWEP6227F.pdf"]
+        random_path = os.path.join(root, "tests", random_string())
+        mail = mailparser.parse_from_file(mail_test_10)
+        os.makedirs(random_path)
+        mail.write_attachments(random_path)
+        for i in attachments:
+            self.assertTrue(os.path.exists(os.path.join(random_path, i)))
+        shutil.rmtree(random_path)
+
+    def test_issue62(self):
+        mail = mailparser.parse_from_file(mail_test_14)
+        received_spf = mail.Received_SPF
+        self.assertIsInstance(received_spf, list)
+        self.assertIn("custom_header1", received_spf)
+        self.assertIn("custom_header2", received_spf)
+
     def test_html_field(self):
         mail = mailparser.parse_from_file(mail_malformed_1)
         self.assertIsInstance(mail.text_html, list)
@@ -117,6 +140,8 @@ def test_mail_partial(self):
         self.assertNotIn("x-ibm-av-version", mail.mail_partial)
         result = mail.mail_partial_json
         self.assertIsInstance(result, six.text_type)
+        nr_attachments = len(mail._attachments)
+        self.assertEqual(nr_attachments, 4)
 
     def test_not_parsed_received(self):
         mail = mailparser.parse_from_file(mail_test_9)