Use extension property for Outlook msg rather than guessing from mimetype

Author: martinburchell

cardinal_pythonlib/extract_text.py

@@ -1362,8 +1362,7 @@ def _gen_msg_content(
     for attachment in message.attachments:
         # null termination seen in the real world
         # https://github.com/TeamMsgExtractor/msg-extractor/issues/464
-        content_type = attachment.mimetype.replace("\x00", "")
-        ext = guess_extension(content_type)
+        ext = attachment.extension.replace("\x00", "")
         if ext is not None and ext in ext_map:
             yield document_to_text(
                 blob=attachment.data, extension=ext, config=config

cardinal_pythonlib/tests/extract_text_tests.py

@@ -677,7 +677,7 @@ def test_attachment_converted(self) -> None:
         mock_attachment = mock.Mock(
             # null termination seen in the real world
             # https://github.com/TeamMsgExtractor/msg-extractor/issues/464
-            mimetype="application/vnd.openxmlformats-officedocument.wordprocessingml.document\x00",  # noqa: E501
+            extension=".docx",
             data=BytesIO(docx).read(),
         )
         mock_msgfile = mock.Mock(