qwd

main.py

@@ -2,6 +2,7 @@ from fastapi import FastAPI, HTTPException
 from fastapi.middleware.cors import CORSMiddleware
 from curl_cffi import requests as cffi_requests
 import re
+import json
 from pydantic import BaseModel
 
 app = FastAPI()
@@ -42,45 +43,47 @@ async def get_transcript(req: TranscriptRequest):
         raise HTTPException(status_code=400, detail="Invalid YouTube URL")
 
     try:
-        print(f"Fetching InnerTube data for video_id: {video_id} using Chrome Impersonation...")
+        print(f"Scraping HTML page for video_id: {video_id} using Chrome Impersonation...")
         
-        # YouTube's internal API endpoint
-        innertube_url = "https://www.youtube.com/youtubei/v1/player?prettyPrint=false"
-        
-        payload = {
-            "context": {
-                "client": {
-                    "clientName": "WEB",
-                    "clientVersion": "2.20241210.00.00",
-                    "hl": "en",
-                    "gl": "US"
-                }
-            },
-            "videoId": video_id
-        }
+        # 1. Fetch the actual YouTube watch page HTML (Exactly like your local Node app does)
+        html_url = f"https://www.youtube.com/watch?v={video_id}"
         
         headers = {
-            "Content-Type": "application/json",
             "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36",
-            "Origin": "https://www.youtube.com",
+            "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8",
+            "Accept-Language": "en-US,en;q=0.9",
             "Referer": "https://www.youtube.com/",
             "Cookie": "CONSENT=YES+1; SOCS=CAESHAgBEhJnd3NfMjAyMzEwMjUtMF9SQzIaAmVuIAEaBgiA-LaoBg"
         }
         
-        # 1. Fetch the video player data (Impersonating Chrome to bypass SSLError)
-        response = cffi_requests.post(innertube_url, json=payload, headers=headers, impersonate="chrome131", timeout=10)
+        # Impersonate Chrome 131 to bypass SSL/TLS fingerprinting
+        response = cffi_requests.get(html_url, headers=headers, impersonate="chrome131", timeout=15)
         
         if response.status_code != 200:
-            raise HTTPException(status_code=500, detail=f"YouTube API returned status {response.status_code}")
+            raise HTTPException(status_code=500, detail=f"YouTube returned status {response.status_code}")
             
-        data = response.json()
+        html = response.text
+        
+        # 2. Find the ytInitialPlayerResponse JSON embedded in the HTML
+        match = re.search(r'var ytInitialPlayerResponse\s*=\s*(\{.*?\});', html)
+        if not match:
+            match = re.search(r'ytInitialPlayerResponse\s*=\s*(\{.*?\});', html)
+            
+        if not match:
+            raise HTTPException(status_code=404, detail="Could not find player data in HTML. YouTube might be blocking the request.")
+        
+        # 3. Parse the JSON data
+        try:
+            data = json.loads(match.group(1))
+        except json.JSONDecodeError:
+            raise HTTPException(status_code=500, detail="Failed to parse YouTube player data.")
         
         # Check if video is playable
         playability = data.get("playabilityStatus", {})
         if playability.get("status") == "ERROR":
             raise HTTPException(status_code=404, detail=playability.get("reason", "Video unavailable"))
             
-        # 2. Extract caption tracks
+        # 4. Extract caption tracks
         caption_tracks = data.get("captions", {}).get("playerCaptionsTracklistRenderer", {}).get("captionTracks", [])
         
         if not caption_tracks:
@@ -100,15 +103,15 @@ async def get_transcript(req: TranscriptRequest):
         if not caption_url:
             raise HTTPException(status_code=404, detail="Caption URL not found.")
             
-        print(f"Fetching captions from URL using Chrome Impersonation...")
+        print(f"Found captions, fetching XML from {caption_url}...")
         
-        # 3. Fetch the actual caption XML (Impersonating Chrome)
+        # 5. Fetch the actual caption XML
         caption_response = cffi_requests.get(caption_url, impersonate="chrome131", timeout=10)
         
         if caption_response.status_code != 200:
             raise HTTPException(status_code=500, detail=f"Caption fetch failed with status {caption_response.status_code}")
             
-        # 4. Parse the XML into plain text
+        # 6. Parse the XML into plain text
         full_text = parse_caption_xml(caption_response.text)
         
         if not full_text.strip():