df

Provider/matrix.py

@@ -15,6 +15,115 @@ from ProviderCore.base import Provider
 _MATRIX_INIT_CHECK_CACHE: Dict[str, Tuple[bool, Optional[str]]] = {}
 
 
+def _sniff_mime_from_header(path: Path) -> Optional[str]:
+	"""Best-effort MIME sniffing from file headers.
+
+	Used when the file has no/unknown extension (common for exported/temp files).
+	Keeps dependencies to stdlib only.
+	"""
+	try:
+		if not path.exists() or not path.is_file():
+			return None
+		with open(path, "rb") as handle:
+			header = handle.read(512)
+		if not header:
+			return None
+
+		# Images
+		if header.startswith(b"\xFF\xD8\xFF"):
+			return "image/jpeg"
+		if header.startswith(b"\x89PNG\r\n\x1a\n"):
+			return "image/png"
+		if header.startswith(b"GIF87a") or header.startswith(b"GIF89a"):
+			return "image/gif"
+		if header.startswith(b"BM"):
+			return "image/bmp"
+		if header.startswith(b"RIFF") and len(header) >= 12 and header[8:12] == b"WEBP":
+			return "image/webp"
+
+		# Audio
+		if header.startswith(b"fLaC"):
+			return "audio/flac"
+		if header.startswith(b"OggS"):
+			# Could be audio or video; treat as audio unless extension suggests video.
+			return "audio/ogg"
+		if header.startswith(b"ID3"):
+			return "audio/mpeg"
+		if len(header) >= 2 and header[0] == 0xFF and (header[1] & 0xE0) == 0xE0:
+			return "audio/mpeg"
+		if header.startswith(b"RIFF") and len(header) >= 12 and header[8:12] == b"WAVE":
+			return "audio/wav"
+
+		# Video
+		if header.startswith(b"RIFF") and len(header) >= 12 and header[8:12] == b"AVI ":
+			return "video/x-msvideo"
+		if header.startswith(b"\x1A\x45\xDF\xA3"):
+			# EBML container: Matroska/WebM.
+			return "video/x-matroska"
+		if len(header) >= 12 and header[4:8] == b"ftyp":
+			# ISO BMFF: mp4/mov/m4a. Default to mp4; extension can refine.
+			return "video/mp4"
+		# MPEG-TS / M2TS (sync byte every 188 bytes)
+		try:
+			if path.stat().st_size >= 188 * 2 and header[0] == 0x47:
+				with open(path, "rb") as handle:
+					handle.seek(188)
+					b = handle.read(1)
+					if b == b"\x47":
+						return "video/mp2t"
+		except Exception:
+			pass
+
+		return None
+	except Exception:
+		return None
+
+
+def _classify_matrix_upload(path: Path, *, explicit_mime_type: Optional[str] = None) -> Tuple[str, str]:
+	"""Return (mime_type, msgtype) for Matrix uploads."""
+	mime_type = str(explicit_mime_type or "").strip() or None
+
+	if not mime_type:
+		# `mimetypes.guess_type` expects a string/URL; Path can return None on some platforms.
+		mime_type, _ = mimetypes.guess_type(str(path))
+
+	if not mime_type:
+		mime_type = _sniff_mime_from_header(path)
+
+	# Refinements based on extension for ambiguous containers.
+	ext = path.suffix.lower()
+	if ext in {".m4a", ".aac"}:
+		mime_type = mime_type or "audio/mp4"
+	if ext in {".mkv", ".webm"}:
+		mime_type = mime_type or "video/x-matroska"
+	if ext in {".ogv"}:
+		mime_type = mime_type or "video/ogg"
+
+	msgtype = "m.file"
+	if mime_type:
+		mt = mime_type.casefold()
+		if mt.startswith("image/"):
+			msgtype = "m.image"
+		elif mt.startswith("audio/"):
+			msgtype = "m.audio"
+		elif mt.startswith("video/"):
+			msgtype = "m.video"
+
+	# Final fallback for unknown MIME types.
+	if msgtype == "m.file":
+		audio_exts = {".mp3", ".flac", ".wav", ".m4a", ".aac", ".ogg", ".opus", ".wma", ".mka", ".alac"}
+		video_exts = {".mp4", ".mkv", ".webm", ".mov", ".avi", ".flv", ".mpg", ".mpeg", ".ts", ".m4v", ".wmv", ".m2ts", ".mts", ".3gp", ".ogv"}
+		image_exts = {".jpg", ".jpeg", ".png", ".gif", ".webp", ".bmp", ".tiff"}
+		if ext in audio_exts:
+			msgtype = "m.audio"
+		elif ext in video_exts:
+			msgtype = "m.video"
+		elif ext in image_exts:
+			msgtype = "m.image"
+
+	return (mime_type or "application/octet-stream"), msgtype
+
+
 def _normalize_homeserver(value: str) -> str:
 	text = str(value or "").strip()
 	if not text:
@@ -189,9 +298,8 @@ class Matrix(Provider):
 			"Content-Type": "application/octet-stream",
 		}
 
-		mime_type, _ = mimetypes.guess_type(path)
-		if mime_type:
-			headers["Content-Type"] = mime_type
+		mime_type, msgtype = _classify_matrix_upload(path, explicit_mime_type=kwargs.get("mime_type"))
+		headers["Content-Type"] = mime_type
 
 		filename = path.name
 
@@ -222,19 +330,6 @@ class Matrix(Provider):
 		except Exception:
 			download_url_for_store = ""
 
-		# Determine message type
-		msgtype = "m.file"
-		ext = path.suffix.lower()
-		audio_exts = {".mp3", ".flac", ".wav", ".m4a", ".aac", ".ogg", ".opus", ".wma", ".mka", ".alac"}
-		video_exts = {".mp4", ".mkv", ".webm", ".mov", ".avi", ".flv", ".mpg", ".mpeg", ".ts", ".m4v", ".wmv"}
-		image_exts = {".jpg", ".jpeg", ".png", ".gif", ".webp", ".bmp", ".tiff"}
-		if ext in audio_exts:
-			msgtype = "m.audio"
-		elif ext in video_exts:
-			msgtype = "m.video"
-		elif ext in image_exts:
-			msgtype = "m.image"
-
 		info = {"mimetype": mime_type, "size": path.stat().st_size}
 		payload = {"msgtype": msgtype, "body": filename, "url": content_uri, "info": info}