extract-msg 0.55.0__tar.gz → 0.56.1__tar.gz

{extract_msg-0.55.0 → extract_msg-0.56.1}/CHANGELOG.md

@@ -1,3 +1,13 @@
+**v0.56.1**
+* [[TeamMsgExtractor #484](https://github.com/TeamMsgExtractor/msg-extractor/issues/484)] Corrected the way the HTML body is decoded in `MessageBase.asEmailMessage()` to prevent encoding errors.
+
+**v0.56.0**
+* [[TeamMsgExtractor #478](https://github.com/TeamMsgExtractor/msg-extractor/issues/478)] Expand allowable versions of `ebcdic` to include `2.x`. The 2.0 release is a pure technical release with no functional changes; it only drops support for Python 2 and 3.8. The lower bound is unchanged so installs on Python 3.8 will continue to resolve to `ebcdic 1.1.1`.
+* [[TeamMsgExtractor #469](https://github.com/TeamMsgExtractor/msg-extractor/issues/469)] Expanded `beautifulsoup` to allow for versions up to `5.*.*`. I was previously hesitant to do so because they sometimes do what seems like non backwards compatible changes in minor versions, but it looks like this should be fine.
+* [[TeamMsgExtractor #471](https://github.com/TeamMsgExtractor/msg-extractor/issues/471)] Added `.strip()` to file name determination to ensure that folders with trailing spaces would be corrected before being used.
+* [[TeamMsgExtractor #481](https://github.com/TeamMsgExtractor/msg-extractor/issues/481)] Fixed a typo that caused an HTML body created from plain text to end with `</head>` instead of `</html>`.
+* [[TeamMsgExtractor #476](https://github.com/TeamMsgExtractor/msg-extractor/issues/476)] Fixed `MessageBase.asEmailMessage()` raising `ValueError` when a message has multiple `To` recipients.
+
 **v0.55.0**
 * [[TeamMsgExtractor #465](https://github.com/TeamMsgExtractor/msg-extractor/issues/465)] Added missing `msg.close()` to `openMsg()`. If the MSG file was actually just a plain OLE file, it would be left open.
 * Adjusted the default value of `maxNameLength` for `MessageBase.save()` to 40 instead of 256.

{extract_msg-0.55.0 → extract_msg-0.56.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: extract_msg
-Version: 0.55.0
+Version: 0.56.1
 Summary: Extracts emails and attachments saved in Microsoft Outlook's .msg files
 Home-page: https://github.com/TeamMsgExtractor/msg-extractor
 Download-URL: https://github.com/TeamMsgExtractor/msg-extractor/archives/master
@@ -278,8 +278,8 @@ 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-0.55.0-blue.svg
-   :target: https://pypi.org/project/extract-msg/0.55.0/
+.. |PyPI3| image:: https://img.shields.io/badge/pypi-0.56.1-blue.svg
+   :target: https://pypi.org/project/extract-msg/0.56.1/
 
 .. |PyPI2| image:: https://img.shields.io/badge/python-3.8+-brightgreen.svg
    :target: https://www.python.org/downloads/release/python-3810/

{extract_msg-0.55.0 → extract_msg-0.56.1}/README.rst

@@ -260,8 +260,8 @@ 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-0.55.0-blue.svg
-   :target: https://pypi.org/project/extract-msg/0.55.0/
+.. |PyPI3| image:: https://img.shields.io/badge/pypi-0.56.1-blue.svg
+   :target: https://pypi.org/project/extract-msg/0.56.1/
 
 .. |PyPI2| image:: https://img.shields.io/badge/python-3.8+-brightgreen.svg
    :target: https://www.python.org/downloads/release/python-3810/

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/__init__.py

@@ -27,8 +27,8 @@ https://github.com/TeamMsgExtractor/msg-extractor
 #    along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 __author__ = 'Destiny Peterson & Matthew Walker'
-__date__ = '2025-08-12'
-__version__ = '0.55.0'
+__date__ = '2026-08-13'
+__version__ = '0.56.1'
 
 __all__ = [
     # Modules:

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/__main__.py

@@ -101,7 +101,7 @@ def main(argv: List[str] = sys.argv) -> None:
                     msg.saveAttachments(**kwargs)
                 else:
                     msg.save(**kwargs)
-        except Exception as e:
+        except Exception:
             try:
                 print(f'Error with file "{x}": {traceback.format_exc()}')
             except UnicodeEncodeError:
@@ -111,5 +111,6 @@ def main(argv: List[str] = sys.argv) -> None:
     if createdZip:
         _zip.close()
 
+
 if __name__ == '__main__':
     main(sys.argv)

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/attachments/custom_att_handler/lnk_obj_att.py

@@ -7,7 +7,7 @@ __all__ = [
 
 
 from functools import cached_property
-from typing import List, Optional, TYPE_CHECKING
+from typing import Optional, TYPE_CHECKING
 
 from . import registerHandler
 from .custom_handler import CustomAttachmentHandler
@@ -106,4 +106,4 @@ class LinkedObjectAttachment(CustomAttachmentHandler):
 
 
 
-registerHandler(LinkedObjectAttachment)
+registerHandler(LinkedObjectAttachment)

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/attachments/web_att.py

@@ -71,4 +71,4 @@ class WebAttachment(AttachmentBase):
         The url for the web attachment. If this is not set, that is probably an
         error.
         """
-        return self.longPathname
+        return self.longPathname

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/constants/ps.py

@@ -43,4 +43,4 @@ PSETID_AIRSYNC: Final[str] = '{71035549-0739-4DCB-9163-00F0580DBBDF}'
 PSETID_SHARING: Final[str] = '{00062040-0000-0000-C000-000000000046}'
 PSETID_XMLEXTRACTEDENTITIES: Final[str] = '{23239608-685D-4732-9C55-4C95CB4E8E33}'
 PSETID_ATTACHMENT: Final[str] = '{96357F7F-59E1-47D0-99A7-46515C183B54}'
-PSETID_CALENDAR_ASSISTANT: Final[str] = '{11000E07-B51B-40D6-AF21-CAA85EDAB1D0}'
+PSETID_CALENDAR_ASSISTANT: Final[str] = '{11000E07-B51B-40D6-AF21-CAA85EDAB1D0}'

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/constants/st.py

@@ -93,7 +93,7 @@ ST_BE_UI8: Final[struct.Struct] = struct.Struct('>B')
 ST_BE_UI16: Final[struct.Struct] = struct.Struct('>H')
 ST_BE_UI32: Final[struct.Struct] = struct.Struct('>I')
 ST_BE_UI64: Final[struct.Struct] = struct.Struct('>Q')
-ST_BE_F32: Final[struct.Struct]  = struct.Struct('>f')
+ST_BE_F32: Final[struct.Struct] = struct.Struct('>f')
 ST_BE_F64: Final[struct.Struct] = struct.Struct('>d')
 # Structs that use the system byte order, where consistency on a single system
 # is all that matters. Mainly used for quick casts between signed and unsigned.
@@ -107,4 +107,4 @@ ST_SBO_UI32: Final[struct.Struct] = struct.Struct('@I')
 ST_SBO_UI64: Final[struct.Struct] = struct.Struct('@Q')
 # Struct for an RGB value that, in little endian, would be an int written in hex
 # as 0x00BBGGRR.
-ST_RGB: Final[struct.Struct] = struct.Struct('<BBBx')
+ST_RGB: Final[struct.Struct] = struct.Struct('<BBBx')

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/msg_classes/message_base.py

@@ -25,6 +25,7 @@ import RTFDE.exceptions
 
 from email import policy
 from email.charset import Charset, QP
+from email.header import decode_header as _decode_header, Header as _Header
 from email.message import EmailMessage
 from email.mime.multipart import MIMEMultipart
 from email.mime.text import MIMEText
@@ -56,6 +57,77 @@ from ..utils import (
 logger = logging.getLogger(__name__)
 logger.addHandler(logging.NullHandler())
 
+_RFC2047_WORD = re.compile(r'=\?[^?]+\?[bBqQ]\?[^?]*\?=')
+
+# Encodings to try when a declared charset fails, keyed by normalised charset name.
+# GBK is a strict superset of GB2312 and accepts the ASCII-range second bytes that
+# GB2312 rejects, so real-world GB2312-labelled headers often decode correctly as GBK.
+_CHARSET_FALLBACKS: Dict[str, Tuple[str, ...]] = {
+    'gb2312': ('gbk', 'cp936'),
+    'gb_2312': ('gbk', 'cp936'),
+    'gb_2312-80': ('gbk', 'cp936'),
+}
+
+
+def _fix_encoded_word(m: re.Match) -> str:
+    """
+    Regex substitution callback: decode one RFC 2047 encoded word and
+    re-emit it as a valid UTF-8 encoded word.
+
+    If the declared charset fails (e.g. GB2312-labelled GBK bytes), charset
+    fallbacks from _CHARSET_FALLBACKS are tried before falling back to
+    latin-1 with replacement characters.
+    """
+    word = m.group(0)
+    try:
+        parts = _decode_header(word)
+        if len(parts) != 1 or not isinstance(parts[0][0], bytes):
+            return word
+        btext, cs = parts[0]
+        try:
+            text = btext.decode(cs or 'ascii')
+        except (UnicodeDecodeError, LookupError):
+            for fallback in _CHARSET_FALLBACKS.get((cs or '').lower(), ()):
+                try:
+                    text = btext.decode(fallback)
+                    break
+                except (UnicodeDecodeError, LookupError):
+                    continue
+            else:
+                text = btext.decode('latin-1', 'replace')
+        return str(_Header(text, charset='utf-8'))
+    except Exception:
+        return word
+
+
+def _preprocess_encoded_words(text: str) -> str:
+    """
+    Replace every RFC 2047 encoded word in *text* with a clean UTF-8
+    encoded word.  Safe to apply to an entire raw header block before
+    feeding it to the modern email policy parser.
+    """
+    return _RFC2047_WORD.sub(_fix_encoded_word, text)
+
+
+def _sanitize_header(value: str) -> str:
+    """
+    Prepare a header value from a compat32-parsed Message for safe assignment
+    to an EmailMessage (fallback path when no raw headerText is available).
+
+    RFC 2047 encoded words are re-encoded as UTF-8; raw non-ASCII characters
+    are also RFC 2047-encoded so that EmailMessage.as_bytes() never raises
+    UnicodeEncodeError.
+    """
+    if '=?' in value:
+        value = _RFC2047_WORD.sub(_fix_encoded_word, value)
+
+    try:
+        value.encode('ascii')
+    except (UnicodeEncodeError, UnicodeDecodeError):
+        value = str(_Header(value, charset='utf-8'))
+
+    return value
+
 
 class MessageBase(MSGFile):
     """
@@ -165,10 +237,48 @@ class MessageBase(MSGFile):
         """
         ret = EmailMessage()
 
-        # Copy the headers.
-        for key, value in self.header.items():
-            if key.lower() != 'content-type':
-                ret[key] = value.replace('\r\n', '').replace('\n', '')
+        # Prefer the raw transport-header block: pre-fix any malformed RFC 2047
+        # encoded words (e.g. GB2312-labelled GBK bytes), then parse with the
+        # modern policy so that RFC 5322 unfolding and RFC 2047 decoding are
+        # handled natively and the values arrive as clean Unicode strings.
+        # Fall back to the compat32-derived self.header when no raw text is
+        # stored (synthesised headers from MAPI properties).
+        _ADDRESS_HEADERS = frozenset({'to', 'cc', 'bcc', 'reply-to'})
+        if self.headerText:
+            raw = self.headerText
+            if raw.startswith('Microsoft Mail Internet Headers Version 2.0'):
+                raw = raw[43:].lstrip()
+            raw = _preprocess_encoded_words(raw)
+            source_items = list(HeaderParser(policy = policy.default).parsestr(raw).items())
+            sanitize = False
+        else:
+            source_items = list(self.header.items())
+            sanitize = True
+
+        # Address headers (To/CC/BCC/Reply-To) may appear once per recipient in
+        # the stored header block; merge them into a single comma-separated value.
+        # All other headers — including multi-valued trace headers like Received
+        # and Authentication-Results — are forwarded as-is; EmailMessage appends
+        # rather than replaces, so natural repetition is preserved correctly.
+        address_merged: Dict[str, Tuple[str, str]] = {}
+        for key, value in source_items:
+            if key.lower() == 'content-type':
+                continue
+            if sanitize:
+                # compat32 values are folded and may contain raw encoded words.
+                value = re.sub(r'\r?\n[ \t]', ' ', value)
+                value = value.replace('\r\n', '').replace('\n', '')
+                value = _sanitize_header(value)
+            lower = key.lower()
+            if lower in _ADDRESS_HEADERS:
+                if lower in address_merged:
+                    address_merged[lower] = (address_merged[lower][0], address_merged[lower][1] + ', ' + value)
+                else:
+                    address_merged[lower] = (key, value)
+            else:
+                ret[key] = value
+        for _, (key, value) in address_merged.items():
+            ret[key] = value
 
         ret['Content-Type'] = 'multipart/mixed'
 
@@ -184,7 +294,7 @@ class MessageBase(MSGFile):
         if self.body:
             bodyParts.attach(MIMEText(self.body, 'plain', c))
         if self.htmlBody:
-            bodyParts.attach(MIMEText(self.htmlBody.decode('utf-8'), 'html', c))
+            bodyParts.attach(MIMEText(str(bs4.BeautifulSoup(self.htmlBody, features = 'html.parser')), 'html', c))
 
         # Process attachments.
         for att in self.attachments:
@@ -855,9 +965,9 @@ class MessageBase(MSGFile):
                     raise ValueError(f'Invalid filename found in self.filename: "{self.filename}"')
 
                 # Add the file name to the path.
-                path /= filename[:maxNameLength]
+                path /= filename[:maxNameLength].strip()
             else:
-                path /= self.defaultFolderName[:maxNameLength]
+                path /= self.defaultFolderName[:maxNameLength].strip()
 
             # Create the folders.
             if not _zip:
@@ -1178,7 +1288,7 @@ class MessageBase(MSGFile):
             # Convert the plain text body to html.
             logger.info('HTML body was not found, attempting to generate from plain text body.')
             correctedBody = html.escape(self.body).replace('\r', '').replace('\n', '<br />')
-            htmlBody = f'<html><body>{correctedBody}</body></head>'.encode('ascii', 'xmlcharrefreplace')
+            htmlBody = f'<html><body>{correctedBody}</body></html>'.encode('ascii', 'xmlcharrefreplace')
 
         if not htmlBody:
             logger.info('HTML body could not be found nor generated.')

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg/utils.py

@@ -458,7 +458,7 @@ def getCommandArgs(args: Sequence[str]) -> argparse.Namespace:
                         help='Extracts the embedded MSG files as MSG files instead of running their save functions.')
     # --overwrite-existing
     parser.add_argument('--overwrite-existing', dest='overwriteExisting', action='store_true',
-                        help='Disables filename conflict resolution code for attachments when saving a file, causing files to be overwriten if two attachments with the same filename are on an MSG file.')
+                        help='Disables filename conflict resolution code for attachments when saving a file, causing files to be overwritten if two attachments with the same filename are on an MSG file.')
     # --skip-not-implemented
     parser.add_argument('--skip-not-implemented', '--skip-ni', dest='skipNotImplemented', action='store_true',
                         help='Skips any attachments that are not implemented, allowing saving of the rest of the message.')

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: extract-msg
-Version: 0.55.0
+Version: 0.56.1
 Summary: Extracts emails and attachments saved in Microsoft Outlook's .msg files
 Home-page: https://github.com/TeamMsgExtractor/msg-extractor
 Download-URL: https://github.com/TeamMsgExtractor/msg-extractor/archives/master
@@ -278,8 +278,8 @@ 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-0.55.0-blue.svg
-   :target: https://pypi.org/project/extract-msg/0.55.0/
+.. |PyPI3| image:: https://img.shields.io/badge/pypi-0.56.1-blue.svg
+   :target: https://pypi.org/project/extract-msg/0.56.1/
 
 .. |PyPI2| image:: https://img.shields.io/badge/python-3.8+-brightgreen.svg
    :target: https://www.python.org/downloads/release/python-3810/

{extract_msg-0.55.0 → extract_msg-0.56.1}/extract_msg.egg-info/requires.txt

@@ -1,8 +1,8 @@
 olefile==0.47
 tzlocal<6,>=4.2
 compressed-rtf<2,>=1.0.6
-ebcdic<2,>=1.1.1
-beautifulsoup4<4.14,>=4.11.1
+ebcdic<3,>=1.1.1
+beautifulsoup4<5,>=4.11.1
 RTFDE<0.2,>=0.1.1
 red-black-tree-mod<=1.23,>=1.20
 

{extract_msg-0.55.0 → extract_msg-0.56.1}/requirements.txt

@@ -5,7 +5,9 @@
 olefile==0.47
 tzlocal>=4.2,<6
 compressed-rtf>=1.0.6,<2
-ebcdic>=1.1.1,<2
-beautifulsoup4>=4.11.1,<4.14
+ebcdic>=1.1.1,<3
+# Beautiful soup can sometimes make weird versioning, so something *could*
+# break, but so long as they at least reasonably do sem-var this should be fine.
+beautifulsoup4>=4.11.1,<5
 RTFDE>=0.1.1,<0.2
 red-black-tree-mod>=1.20, <=1.23