Update to the newest MFP version, including the Vixcloud PR

mediaflow_proxy/extractors/vixcloud.py

@@ -15,18 +15,18 @@ class VixCloudExtractor(BaseExtractor):
         super().__init__(*args, **kwargs)
         self.mediaflow_endpoint = "hls_manifest_proxy"
 
-    async def version(self, domain: str) -> str:
+    async def version(self, site_url: str) -> str:
         """Get version of VixCloud Parent Site."""
-        base_url = f"https://streamingcommunity.{domain}/richiedi-un-titolo"
+        base_url = f"{site_url}/richiedi-un-titolo"
         response = await self._make_request(
             base_url,
             headers={
-                "Referer": f"https://streamingcommunity.{domain}/",
-                "Origin": f"https://streamingcommunity.{domain}",
+                "Referer": f"{site_url}/",
+                "Origin": f"{site_url}",
             },
         )
         if response.status_code != 200:
-            raise ExtractorError("Outdated Domain")
+            raise ExtractorError("Outdated Url")
         # Soup the response
         soup = BeautifulSoup(response.text, "lxml", parse_only=SoupStrainer("div", {"id": "app"}))
         if soup:
@@ -39,8 +39,8 @@ class VixCloudExtractor(BaseExtractor):
 
     async def extract(self, url: str, **kwargs) -> Dict[str, Any]:
         """Extract Vixcloud URL."""
-        domain = url.split("://")[1].split("/")[0].split(".")[1]
-        version = await self.version(domain)
+        site_url = url.split("/iframe")[0]
+        version = await self.version(site_url)
         response = await self._make_request(url, headers={"x-inertia": "true", "x-inertia-version": version})
         soup = BeautifulSoup(response.text, "lxml", parse_only=SoupStrainer("iframe"))
         iframe = soup.find("iframe").get("src")

mediaflow_proxy/utils/crypto_utils.py

@@ -67,30 +67,24 @@ class EncryptionMiddleware(BaseHTTPMiddleware):
         encrypted_token = None
 
         # Check for token in path
-        if token_marker in path and self.encryption_handler:
+        if path.startswith(token_marker) and self.encryption_handler:
             try:
-                # Extract token from path
-                token_start = path.find(token_marker) + len(token_marker)
+                # Extract token from the beginning of the path
+                token_start = len(token_marker)
                 token_end = path.find("/", token_start)
 
-                if token_end == -1:  # No trailing slash (no filename after token)
-                    token_end = len(path)
-                    filename_part = ""
+                if token_end == -1:  # No trailing slash
+                    encrypted_token = path[token_start:]
+                    remaining_path = ""
                 else:
-                    # There's something after the token (likely a filename)
-                    filename_part = path[token_end:]
+                    encrypted_token = path[token_start:token_end]
+                    remaining_path = path[token_end:]
 
-                # Get the encrypted token
-                encrypted_token = path[token_start:token_end]
-
-                # Modify the path to remove the token part but preserve the filename
-                original_path = path[: path.find(token_marker)]
-                original_path += filename_part  # Add back the filename part
-
-                request.scope["path"] = original_path
+                # Modify the path to remove the token part
+                request.scope["path"] = remaining_path
 
                 # Update the raw path as well
-                request.scope["raw_path"] = original_path.encode()
+                request.scope["raw_path"] = remaining_path.encode()
 
             except Exception as e:
                 logging.error(f"Error processing token in path: {str(e)}")

mediaflow_proxy/utils/http_utils.py

@@ -330,15 +330,25 @@ def encode_mediaflow_proxy_url(
     # Handle encryption if needed
     if encryption_handler:
         encrypted_token = encryption_handler.encrypt_data(query_params, expiration, ip)
-        # Build the URL with token in path
-        path_parts = [base_url, f"_token_{encrypted_token}"]
+
+        # Parse the base URL to get its components
+        parsed_url = parse.urlparse(base_url)
+
+        # Insert the token at the beginning of the path
+        new_path = f"/_token_{encrypted_token}{parsed_url.path}"
+
+        # Reconstruct the URL with the token at the beginning of the path
+        url_parts = list(parsed_url)
+        url_parts[2] = new_path  # Update the path component
+
+        # Build the URL
+        url = parse.urlunparse(url_parts)
 
         # Add filename at the end if provided
         if filename:
-            path_parts.append(parse.quote(filename))
-
-        return "/".join(path_parts)
+            url = f"{url}/{parse.quote(filename)}"
 
+        return url
     else:
         # No encryption, use regular query parameters
         url = base_url

mediaflow_proxy/utils/mpd_utils.py

@@ -233,8 +233,21 @@ def parse_representation(
         profile["audioSamplingRate"] = representation.get("@audioSamplingRate") or adaptation.get("@audioSamplingRate")
         profile["channels"] = representation.get("AudioChannelConfiguration", {}).get("@value", "2")
     else:
-        profile["width"] = int(representation["@width"])
-        profile["height"] = int(representation["@height"])
+        # Handle video-specific attributes, making them optional with sensible defaults
+        if "@width" in representation:
+            profile["width"] = int(representation["@width"])
+        elif "@width" in adaptation:
+            profile["width"] = int(adaptation["@width"])
+        else:
+            profile["width"] = 0  # Default if width is missing
+
+        if "@height" in representation:
+            profile["height"] = int(representation["@height"])
+        elif "@height" in adaptation:
+            profile["height"] = int(adaptation["@height"])
+        else:
+            profile["height"] = 0  # Default if height is missing
+
         frame_rate = representation.get("@frameRate") or adaptation.get("@maxFrameRate") or "30000/1001"
         frame_rate = frame_rate if "/" in frame_rate else f"{frame_rate}/1"
         profile["frameRate"] = round(int(frame_rate.split("/")[0]) / int(frame_rate.split("/")[1]), 3)
@@ -317,7 +330,9 @@ def parse_segment_timeline(parsed_dict: dict, item: dict, profile: dict, source:
     """
     timelines = item["SegmentTimeline"]["S"]
     timelines = timelines if isinstance(timelines, list) else [timelines]
-    period_start = parsed_dict.get("availabilityStartTime", datetime.fromtimestamp(0, tz=timezone.utc)) + timedelta(seconds=parsed_dict.get("PeriodStart", 0))
+    period_start = parsed_dict.get("availabilityStartTime", datetime.fromtimestamp(0, tz=timezone.utc)) + timedelta(
+        seconds=parsed_dict.get("PeriodStart", 0)
+    )
     presentation_time_offset = int(item.get("@presentationTimeOffset", 0))
     start_number = int(item.get("@startNumber", 1))
 
@@ -354,13 +369,14 @@ def preprocess_timeline(
         for _ in range(repeat + 1):
             segment_start_time = period_start + timedelta(seconds=(start_time - presentation_time_offset) / timescale)
             segment_end_time = segment_start_time + timedelta(seconds=duration / timescale)
+            presentation_time = start_time - presentation_time_offset
             processed_data.append(
                 {
                     "number": start_number,
                     "start_time": segment_start_time,
                     "end_time": segment_end_time,
                     "duration": duration,
-                    "time": start_time,
+                    "time": presentation_time,
                 }
             )
             start_time += duration
@@ -475,7 +491,7 @@ def create_segment_data(segment: Dict, item: dict, profile: dict, source: str, t
     media = media.replace("$Bandwidth$", str(profile["bandwidth"]))
 
     if "time" in segment and timescale is not None:
-        media = media.replace("$Time$", str(int(segment["time"] * timescale)))
+        media = media.replace("$Time$", str(int(segment["time"])))
 
     if not media.startswith("http"):
         media = f"{source}/{media}"
@@ -496,16 +512,20 @@ def create_segment_data(segment: Dict, item: dict, profile: dict, source: str, t
             }
         )
     elif "start_time" in segment and "duration" in segment:
-        duration = segment["duration"]
+        duration_seconds = segment["duration"] / timescale
         segment_data.update(
             {
                 "start_time": segment["start_time"],
-                "end_time": segment["start_time"] + timedelta(seconds=duration),
-                "extinf": duration,
+                "end_time": segment["start_time"] + timedelta(seconds=duration_seconds),
+                "extinf": duration_seconds,
                 "program_date_time": segment["start_time"].isoformat() + "Z",
             }
         )
+    elif "duration" in segment and timescale is not None:
+        # Convert duration from timescale units to seconds
+        segment_data["extinf"] = segment["duration"] / timescale
     elif "duration" in segment:
+        # If no timescale is provided, assume duration is already in seconds
         segment_data["extinf"] = segment["duration"]
 
     return segment_data