up

server.js

@@ -1,14 +1,31 @@
 const express = require('express');
 const cors = require('cors');
-const { YoutubeTranscript } = require('youtube-transcript');
 
 const app = express();
 const PORT = 7860;
 
-// Enable CORS for your Vercel frontend
 app.use(cors());
 app.use(express.json());
 
+function extractVideoId(url) {
+  const match = url.match(/(?:youtube\.com\/(?:[^\/]+\/.+\/|(?:v|e(?:mbed)?)\/|.*[?&]v=)|youtu\.be\/)([^"&?\/\s]{11})/);
+  return match ? match[1] : null;
+}
+
+function parseCaptionXml(xml) {
+  const matches = xml.match(/<text[^>]*>([^<]*)<\/text>/g);
+  if (!matches) return "";
+  return matches
+    .map(m => m.replace(/<[^>]+>/g, ''))
+    .join(" ")
+    .replace(/&amp;#39;/g, "'")
+    .replace(/&amp;quot;/g, '"')
+    .replace(/&amp;amp;/g, "&")
+    .replace(/&#39;/g, "'")
+    .replace(/\s+/g, ' ')
+    .trim();
+}
+
 // ========================================================
 // KEEP-ALIVE ENDPOINT
 // ========================================================
@@ -22,47 +39,88 @@ app.get('/api/health', (req, res) => {
 app.post('/api/transcript', async (req, res) => {
   try {
     const { url } = req.body;
-
-    const videoIdMatch = url.match(
-      /(?:youtube\.com\/(?:[^\/]+\/.+\/|(?:v|e(?:mbed)?)\/|.*[?&]v=)|youtu\.be\/)([^"&?\/\s]{11})/
-    );
-    if (!videoIdMatch) {
+    const videoId = extractVideoId(url);
+    
+    if (!videoId) {
       return res.status(400).json({ error: "Invalid YouTube URL" });
     }
-    const videoId = videoIdMatch[1];
-
-    let transcriptData;
-    let retries = 3;
-    let lastError;
-
-    while (retries > 0) {
-      try {
-        transcriptData = await YoutubeTranscript.fetchTranscript(videoId);
-        break; // success
-      } catch (err) {
-        lastError = err;
-        retries--;
-        if (retries > 0) {
-          await new Promise((resolve) => setTimeout(resolve, 2000));
-        }
-      }
+
+    console.log(`Fetching InnerTube data for video_id: ${videoId}`);
+
+    // Anti-bot headers and cookies to bypass YouTube blocks on Cloud IPs
+    const 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",
+      "Referer": "https://www.youtube.com/",
+      "Cookie": "CONSENT=YES+1; SOCS=CAESHAgBEhJnd3NfMjAyMzEwMjUtMF9SQzIaAmVuIAEaBgiA-LaoBg"
+    };
+
+    // 1. Fetch video player data from InnerTube API
+    const playerRes = await fetch("https://www.youtube.com/youtubei/v1/player?prettyPrint=false", {
+      method: "POST",
+      headers: headers,
+      body: JSON.stringify({
+        context: {
+          client: {
+            clientName: "WEB",
+            clientVersion: "2.20241210.00.00",
+            hl: "en",
+            gl: "US"
+          }
+        },
+        videoId: videoId
+      })
+    });
+
+    if (!playerRes.ok) {
+      throw new Error(`YouTube API returned status ${playerRes.status}`);
+    }
+
+    const playerData = await playerRes.json();
+
+    // Check if video is playable
+    if (playerData.playabilityStatus?.status === "ERROR") {
+      return res.status(404).json({ error: playerData.playabilityStatus?.reason || "Video unavailable" });
     }
 
-    if (!transcriptData) {
-      throw lastError;
+    // 2. Extract caption tracks
+    const captionTracks = playerData?.captions?.playerCaptionsTracklistRenderer?.captionTracks;
+
+    if (!captionTracks || captionTracks.length === 0) {
+      return res.status(404).json({ error: "No captions found for this video." });
     }
 
-    const fullText = transcriptData
-      .map((t) => t.text)
-      .join(" ")
-      .replace(/&amp;#39;|&amp;quot;/g, "'");
+    // Prefer English, fallback to first available
+    let selectedTrack = captionTracks.find(t => t.languageCode === 'en') || captionTracks[0];
+    let captionUrl = selectedTrack.baseUrl;
 
+    console.log(`Fetching captions from URL...`);
+
+    // 3. Fetch the actual caption XML
+    const captionRes = await fetch(captionUrl, {
+      headers: { "User-Agent": headers["User-Agent"] }
+    });
+
+    if (!captionRes.ok) {
+      throw new Error(`Caption fetch failed with status ${captionRes.status}`);
+    }
+
+    const captionXml = await captionRes.text();
+
+    // 4. Parse the XML into plain text
+    const fullText = parseCaptionXml(captionXml);
+
+    if (!fullText) {
+      return res.status(404).json({ error: "Caption XML was empty." });
+    }
+
+    console.log(`Successfully fetched transcript. Length: ${fullText.length} chars`);
     res.json({ text: fullText });
+
   } catch (error) {
     console.error("Transcript Error:", error.message);
-    res.status(500).json({
-      error: "Could not fetch transcript. The video might not have captions enabled, or the request timed out."
-    });
+    res.status(500).json({ error: "Could not fetch transcript. " + error.message });
   }
 });