Merge pull request #383 from TeamMsgExtractor/next-release

Version 0.44.0

Author: TheElementalOfDestruction

CHANGELOG.md

@@ -1,3 +1,12 @@
+**v0.44.0**
+* Fixed a bug that caused `MessageBase.headerInit` to always return `False` after the 0.42.0 update.
+* Changed `MessageBase.headerInit` to a property.
+* Fixed `extract_msg.utils.__all__`.
+* Minor regoanization within `extract_msg/utils.py`.
+* Minor changes to docstrings.
+* Minor README updates.
+* Fix issue with folded header fields decoding incorrectly when given to `extract_msg.utils.decodeRfc2047`.
+
 **v0.43.0**
 * [[TeamMsgExtractor #56](https://github.com/TeamMsgExtractor/msg-extractor/issues/56)] [[TeamMsgExtractor #248](https://github.com/TeamMsgExtractor/msg-extractor/issues/248)] Added new function `MessageBase.asEmailMessage` which will convert the `MessageBase` instance, if possible, to an `email.message.EmailMessage` object. If an embedded MSG file on a `MessageBase` object is of a class that does not have this function, it will simply be attached to the instance as bytes.
 * Changed imports in `message_base.py` to help with type checkers.

README.rst

@@ -1,4 +1,4 @@
-|License: GPL v3| |PyPI3| |PyPI2|
+|License: GPL v3| |PyPI3| |PyPI2| |Read the Docs|
 
 extract-msg
 =============
@@ -9,7 +9,7 @@ The python package extract_msg automates the extraction of key email
 data (from, to, cc, date, subject, body) and the email's attachments.
 
 Documentation can be found in the code, on the `wiki`_, and on the
-`read the docs`_ page.
+`Read the Docs`_ page.
 
 NOTICE
 ======
@@ -242,11 +242,16 @@ your access to the newest major version of extract-msg.
 .. |License: GPL v3| image:: https://img.shields.io/badge/License-GPLv3-blue.svg
    :target: LICENSE.txt
 
-.. |PyPI3| image:: https://img.shields.io/badge/pypi-0.43.0-blue.svg
-   :target: https://pypi.org/project/extract-msg/0.43.0/
+.. |PyPI3| image:: https://img.shields.io/badge/pypi-0.44.0-blue.svg
+   :target: https://pypi.org/project/extract-msg/0.44.0/
 
 .. |PyPI2| image:: https://img.shields.io/badge/python-3.8+-brightgreen.svg
    :target: https://www.python.org/downloads/release/python-3816/
+
+.. |Read the Docs| image:: https://readthedocs.org/projects/msg-extractor/badge/?version=latest
+    :target: https://msg-extractor.readthedocs.io/en/stable/?badge=latest
+    :alt: Documentation Status
+
 .. _Matthew Walker: https://github.com/mattgwwalker
 .. _Destiny Peterson (The Elemental of Destruction): https://github.com/TheElementalOfDestruction
 .. _JP Bourget: https://github.com/punkrokk
@@ -261,5 +266,5 @@ your access to the newest major version of extract-msg.
 .. _Patreon: https://www.patreon.com/DestructionE
 .. _msg-explorer: https://pypi.org/project/msg-explorer/
 .. _wiki: https://github.com/TeamMsgExtractor/msg-extractor/wiki
-.. _read the docs: https://msg-extractor.rtfd.io/
+.. _Read the Docs: https://msg-extractor.rtfd.io/
 .. _Changelog: https://github.com/TeamMsgExtractor/msg-extractor/blob/master/CHANGELOG.md

docs/conf.py

@@ -12,7 +12,7 @@
 sys.path.insert(0, os.path.abspath(".."))
 
 __author__ = 'Destiny Peterson & Matthew Walker'
-__version__ = '0.43.0'
+__version__ = '0.44.0'
 __year__ = '2023'
 
 

extract_msg/__init__.py

@@ -27,8 +27,8 @@
 #    along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 __author__ = 'Destiny Peterson & Matthew Walker'
-__date__ = '2023-08-02'
-__version__ = '0.43.0'
+__date__ = '2023-08-03'
+__version__ = '0.44.0'
 
 __all__ = [
     # Modules:

extract_msg/msg_classes/calendar_base.py

@@ -35,7 +35,7 @@ def _genRecipient(self, recipientType, recipientInt : MeetingRecipientType) -> O
         recipientInt = MeetingRecipientType(recipientInt)
         value = None
         # Check header first.
-        if self.headerInit():
+        if self.headerInit:
             value = self.header[recipientType]
             if value:
                 value = value.replace(',', self.recipientSeparator)
@@ -44,7 +44,7 @@ def _genRecipient(self, recipientType, recipientInt : MeetingRecipientType) -> O
         # it manually.
         if not value:
             # Check if the header has initialized.
-            if self.headerInit():
+            if self.headerInit:
                 logger.info(f'Header found, but "{recipientType}" is not included. Will be generated from other streams.')
 
             # Get a list of the recipients of the specified type.

extract_msg/msg_classes/message_base.py

@@ -89,6 +89,7 @@ def __init__(self, path, **kwargs):
         # The rest needs to be in a try-except block to ensure the file closes
         # if an error occurs.
         try:
+            self.__headerInit = False
             self.__recipientSeparator = kwargs.get('recipientSeparator', ';')
             self.__deencap = kwargs.get('deencapsulationFunc')
             # Initialize properties in the order that is least likely to cause bugs.
@@ -125,7 +126,7 @@ def _genRecipient(self, recipientType, recipientInt : RecipientType) -> Optional
         recipientInt = RecipientType(recipientInt)
         value = None
         # Check header first.
-        if self.headerInit():
+        if self.headerInit:
             value = self.header[recipientType]
             if value:
                 value = decodeRfc2047(value)
@@ -135,7 +136,7 @@ def _genRecipient(self, recipientType, recipientInt : RecipientType) -> Optional
         # it manually.
         if not value:
             # Check if the header has initialized.
-            if self.headerInit():
+            if self.headerInit:
                 logger.info(f'Header found, but "{recipientType}" is not included. Will be generated from other streams.')
 
             # Get a list of the recipients of the specified type.
@@ -472,16 +473,6 @@ def getSaveRtfBody(self, **_) -> bytes:
         # Inject the header into the data.
         return self.injectRtfHeader()
 
-    def headerInit(self) -> bool:
-        """
-        Checks whether the header has been initialized.
-        """
-        try:
-            self._header
-            return True
-        except AttributeError:
-            return False
-
     def injectHtmlHeader(self, prepared : bool = False) -> bytes:
         """
         Returns the HTML body from the MSG file (will check that it has one) with
@@ -1080,6 +1071,7 @@ def header(self) -> email.message.Message:
             # TODO find authentication results outside of header
             header.add_header('Authentication-Results', None)
 
+        self.__headerInit = True
         return header
 
     @property
@@ -1139,6 +1131,13 @@ def headerFormatProperties(self) -> constants.HEADER_FORMAT_TYPE:
             },
         }
 
+    @property
+    def headerInit(self) -> bool:
+        """
+        Checks whether the header has been initialized.
+        """
+        return self.__headerInit
+
     @functools.cached_property
     def headerText(self) -> Optional[str]:
         """
@@ -1240,12 +1239,12 @@ def isSent(self) -> bool:
     @functools.cached_property
     def messageId(self) -> Optional[str]:
         headerResult = None
-        if self.headerInit():
+        if self.headerInit:
             headerResult = self.header['message-id']
         if headerResult is not None:
             return headerResult
 
-        if self.headerInit():
+        if self.headerInit:
             logger.info('Header found, but "Message-Id" is not included. Will be generated from other streams.')
         return self._getStringStream('__substg1.0_1035')
 
@@ -1323,7 +1322,7 @@ def sender(self) -> Optional[str]:
         Returns the message sender, if it exists.
         """
         # Check header first
-        if self.headerInit():
+        if self.headerInit:
             headerResult = self.header['from']
             if headerResult is not None:
                 return decodeRfc2047(headerResult)

extract_msg/utils.py

@@ -15,6 +15,7 @@
     'ceilDiv',
     'cloneOleFile',
     'createZipOpen',
+    'decodeRfc2047',
     'dictGetCasedKey',
     'divide',
     'filetimeToDatetime',
@@ -201,6 +202,9 @@ def decodeRfc2047(encoded : str) -> str:
     """
     Decodes text encoded using the method specified in RFC 2047.
     """
+    # Fix an issue with folded header fields.
+    encoded = encoded.replace('\r\n', '')
+
     # This returns a list of tuples containing the bytes and the encoding they
     # are using, so we decode each one and join them together.
     #
@@ -277,6 +281,13 @@ def filetimeToDatetime(rawTime : int) -> datetime.datetime:
         raise ValueError(f'Timestamp value of {filetimeToUtc(rawTime)} caused an exception. This was probably caused by the time stamp being too far in the future.')
 
 
+def filetimeToUtc(inp : int) -> float:
+    """
+    Converts a FILETIME into a unix timestamp.
+    """
+    return (inp - 116444736000000000) / 10000000.0
+
+
 def findWk(path = None):
     """
     Attempt to find the path of the wkhtmltopdf executable. If :param path: is
@@ -581,27 +592,20 @@ def inputToString(bytesInputVar, encoding) -> str:
         raise ConversionError('Cannot convert to str type.')
 
 
-def isEncapsulatedRtf(inp : bytes) -> bool:
-    """
-    Currently the detection is made to be *extremly* basic, but this will work
-    for now. In the future this will be fixed so that literal text in the body
-    of a message won't cause false detection.
-    """
-    return b'\\fromhtml' in inp
-
-
 def isEmptyString(inp : str) -> bool:
     """
     Returns true if the input is None or is an Empty string.
     """
     return (inp == '' or inp is None)
 
 
-def filetimeToUtc(inp : int) -> float:
+def isEncapsulatedRtf(inp : bytes) -> bool:
     """
-    Converts a FILETIME into a unix timestamp.
+    Currently the detection is made to be *extremly* basic, but this will work
+    for now. In the future this will be fixed so that literal text in the body
+    of a message won't cause false detection.
     """
-    return (inp - 116444736000000000) / 10000000.0
+    return b'\\fromhtml' in inp
 
 
 def makeWeakRef(obj : Optional[_T]) -> Optional[weakref.ReferenceType[_T]]: