Fall back to content-based MIME detection via puremagic

emails/store/file.py

@@ -2,6 +2,7 @@
 
 import uuid
 from mimetypes import guess_type
+import puremagic
 from email.mime.base import MIMEBase
 from email.encoders import encode_base64
 from os.path import basename
@@ -102,6 +103,23 @@ def get_mime_type(self) -> str:
             filename = self.filename
             if filename:
                 r = self._mime_type = guess_type(filename)[0]
+        if not r:
+            _data = self._data
+            if isinstance(_data, bytes):
+                header = _data
+            elif isinstance(_data, str):
+                header = _data.encode()
+            elif _data is not None:
+                pos = _data.tell()
+                header = _data.read(128)
+                _data.seek(pos)
+            else:
+                header = None
+            if header:
+                try:
+                    r = puremagic.from_string(header, mime=True)
+                except puremagic.PureError:
+                    pass
         if not r:
             r = MIMETYPE_UNKNOWN
         self._mime_type = r

emails/testsuite/store/test_store.py

@@ -68,6 +68,33 @@ def test_get_data_none():
     assert f.data is None
 
 
+def test_mime_type_from_content():
+    # PNG magic bytes, no file extension
+    png_header = (b'\x89PNG\r\n\x1a\n\x00\x00\x00\rIHDR'
+                  b'\x00\x00\x00\x01\x00\x00\x00\x01\x08\x02'
+                  b'\x00\x00\x00\x90wS\xde')
+    f = BaseFile(data=png_header, filename='image_no_ext')
+    assert f.mime_type == 'image/png'
+
+    # JPEG magic bytes, no file extension
+    jpeg_header = b'\xff\xd8\xff\xe0\x00\x10JFIF'
+    f = BaseFile(data=jpeg_header, filename='photo')
+    assert f.mime_type == 'image/jpeg'
+
+    # Unknown bytes, no extension — should fall back to unknown
+    f = BaseFile(data=b'\x00\x01\x02\x03', filename='mystery')
+    assert f.mime_type == 'application/unknown'
+
+    # Extension still takes priority
+    f = BaseFile(data=png_header, filename='image.gif')
+    assert f.mime_type == 'image/gif'
+
+    # File-like data: mime detected without exhausting stream
+    f = BaseFile(data=BytesIO(png_header), filename='no_ext')
+    assert f.mime_type == 'image/png'
+    assert f.data == png_header  # stream not consumed by mime detection
+
+
 def test_store_commons2():
     store = emails.store.MemoryFileStore()
     f1 = store.add({'uri': '/a/c.gif'})

requirements/base.txt

@@ -4,3 +4,4 @@ chardet
 python-dateutil
 requests
 premailer>=2.8.3
+puremagic

setup.py

@@ -127,7 +127,7 @@ def find_version(*file_paths):
     package_data={'emails': ['py.typed']},
     scripts=['scripts/make_rfc822.py'],
     python_requires='>=3.10',
-    install_requires=['python-dateutil'],
+    install_requires=['python-dateutil', 'puremagic'],
     extras_require={
         'html': ['cssutils', 'lxml', 'chardet', 'requests', 'premailer'],
     },