Update server.js

server.js

@@ -1,4 +1,4 @@
-// server.js (Final Version - 1-to-1 Replication of the working Puppeteer code - COMPLETE AND UNABRIDGED)
+// server.js (Complete Playwright Version with CF Bypass)
 
 import express, { json } from "express";
 import cors from "cors";
@@ -9,11 +9,10 @@ import fetch from 'node-fetch';
 import { JSDOM } from 'jsdom';
 import fs from 'fs/promises';
 
-// [REPLICATION] 1. 深度伪装: 使用 playwright-extra 和 stealth 插件
-import playwright from 'playwright-extra';
+// Playwright with stealth plugin
+import { chromium } from 'playwright-extra';
 import StealthPlugin from 'puppeteer-extra-plugin-stealth';
-playwright.chromium.use(StealthPlugin());
-
+chromium.use(StealthPlugin());
 
 const app = express();
 const PORT = process.env.PORT || 3000;
@@ -25,15 +24,13 @@ let browserPool;
 const cache = new LRUCache({ max: 500, ttl: 15 * 60 * 1000 });
 const limit = pLimit(2);
 
-
-// --- 辅助函数 (直接从您的工作代码翻译) ---
-
-// [REPLICATION] 2. 精确打击: 100% 复制您的 handleSpecificTurnstile 逻辑
+// Complete Turnstile bypass function
 async function handleSpecificTurnstile(page, domain) {
     try {
-        const turnstileExists = await page.evaluate(() => {
-            return !!document.querySelector('.cf-turnstile[data-sitekey="0x4AAAAAABNpWSLmOnUi7s0b"]');
-        });
+        const turnstileExists = await page.waitForSelector('.cf-turnstile[data-sitekey="0x4AAAAAABNpWSLmOnUi7s0b"]', { 
+            state: 'attached',
+            timeout: 10000 
+        }).catch(() => null);
 
         if (!turnstileExists) {
             console.log(`[${domain}] Specific Turnstile not detected`);
@@ -41,16 +38,26 @@ async function handleSpecificTurnstile(page, domain) {
         }
 
         console.log(`[${domain}] Specific Turnstile detected - bypassing...`);
-        await page.locator('.cf-turnstile').waitFor({ state: 'visible', timeout: 10000 });
-
+        
         await page.evaluate(() => {
+            const mockToken = 'mock-token-' + Math.random().toString(36).substring(2);
+            
             if (typeof window.cftCallback === 'function') {
-                const mockToken = 'mock-token-' + Math.random().toString(36).substring(2);
                 window.cftCallback(mockToken);
             }
+            
+            const form = document.createElement('form');
+            form.method = 'POST';
+            form.action = '/rcp_verify';
+            const input = document.createElement('input');
+            input.type = 'hidden';
+            input.name = 'token';
+            input.value = mockToken;
+            form.appendChild(input);
+            document.body.appendChild(form);
+            form.submit();
         });
-        
-        // 等待网络空闲以确认挑战已处理
+
         await page.waitForLoadState('networkidle', { timeout: 30000 });
         console.log(`[${domain}] Turnstile bypass completed`);
 
@@ -61,13 +68,13 @@ async function handleSpecificTurnstile(page, domain) {
 
 async function extractFirstIframeSrc(page, domain) {
     try {
-        const iframeLocator = page.locator('#player_iframe');
-        await iframeLocator.waitFor({ state: 'visible', timeout: 15000 });
-        let src = await iframeLocator.getAttribute('src');
-        if (src && src.startsWith('//')) {
-            src = `https:${src}`;
-        }
-        return src;
+        await page.waitForSelector('#player_iframe', { 
+            state: 'attached',
+            timeout: 15000 
+        });
+        
+        const frame = await page.frame('#player_iframe');
+        return frame ? frame.url() : null;
     } catch (error) {
         console.error(`[${domain}] Error finding first iframe:`, error);
         return null;
@@ -76,20 +83,16 @@ async function extractFirstIframeSrc(page, domain) {
 
 async function extractFinalIframeSrc(page, domain) {
     try {
-        await page.waitForFunction(() => document.querySelectorAll('script, iframe').length > 0, null, { timeout: 15000 });
-        return await page.evaluate(() => {
-            const iframe = document.querySelector('iframe:not([style*="display:none"])');
-            if (iframe?.src) return iframe.src;
-            const scripts = Array.from(document.querySelectorAll('script'));
-            for (const script of scripts) {
-                if (!script.textContent) continue;
-                if (script.textContent.includes('loadIframe')) {
-                    const srcMatch = script.textContent.match(/src:\s*['"](.*?)['"]/);
-                    if (srcMatch) return new URL(srcMatch[1], window.location.href).href;
-                }
-            }
-            return null;
+        await page.waitForFunction(() => {
+            return document.querySelectorAll('script, iframe').length > 0;
+        }, { timeout: 15000 });
+
+        const frames = await page.frames();
+        const visibleFrame = frames.find(frame => {
+            return !frame.$('iframe[style*="display:none"]');
         });
+
+        return visibleFrame?.url();
     } catch (error) {
         console.error(`[${domain}] Error finding final iframe:`, error);
         return null;
@@ -106,10 +109,12 @@ async function getVideoAndSubtitles(finalUrl) {
         const scriptTags = document.querySelectorAll('script');
         let videoFileUrl = null;
         let subtitleSources = null;
+        
         for (const script of scriptTags) {
             if (script.textContent.includes('new Playerjs({')) {
                 const fileMatch = script.textContent.match(/file:"(.*?m3u8.*?)"/);
                 if (fileMatch && fileMatch[1]) { videoFileUrl = fileMatch[1]; }
+                
                 const subtitlesMatch = script.textContent.match(/subtitle:"(.*?)"/);
                 if (subtitlesMatch && subtitlesMatch[1]) {
                     try {
@@ -118,9 +123,11 @@ async function getVideoAndSubtitles(finalUrl) {
                         subtitleSources = subtitlesMatch[1].split(',').map(s => s.trim()).filter(Boolean);
                     }
                 }
+                
                 if (videoFileUrl) { break; }
             }
         }
+        
         return { videoFileUrl, subtitleSources };
     } catch (error) {
         console.error('An error occurred during the process:', error);
@@ -128,143 +135,175 @@ async function getVideoAndSubtitles(finalUrl) {
     }
 }
 
-
-// --- 核心抓取逻辑: 100% 复刻成功模式 ---
 async function scrapeProvider(domain, url, signal) {
-  if (signal.aborted) throw new Error('Aborted');
-  console.log(`\n[${domain}] Starting REPLICATED scrape for URL: ${url}`);
-
-  let browserInstance = null;
-  let context = null;
-
-  const cleanup = async () => {
-      if (context) await context.close().catch(()=>{});
-      if (browserInstance) {
-        console.log(`[${domain}] Releasing browser ${browserInstance.id} back to pool.`);
-        await browserPool.release(browserInstance);
-      }
-  };
-
-  try {
-    browserInstance = await browserPool.get();
-    const browser = browserInstance.browser;
-    
-    context = await browser.newContext({
-      userAgent: "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/117.0.0.0 Safari/537.36",
-      ignoreHTTPSErrors: true
-    });
-    
-    await context.route("**/*", (route) => {
-      if (signal.aborted) return route.abort();
-      if (route.request().url().includes('/rcp_verify')) {
-        return route.fulfill({ status: 200, contentType: 'application/json', body: '1' });
-      }
-      return route.continue();
-    });
-
-    const page = await context.newPage();
-
-    // 阶段 1: 访问初始页面 (使用“极度耐心”策略)
-    console.log(`[${domain}] Navigating and patiently waiting for network to be idle...`);
-    await page.goto(url, { waitUntil: 'networkidle', timeout: 60000 });
-    
-    await handleSpecificTurnstile(page, domain);
-
-    const firstIframeSrc = await extractFirstIframeSrc(page, domain);
-    if (!firstIframeSrc) throw new Error('First iframe not found');
-    console.log(`[${domain}] Found first iframe src: ${firstIframeSrc}`);
-
-    // 阶段 2: 访问第一个 iframe (在新页面中，严格复刻)
-    const iframePage = await context.newPage();
-    await iframePage.goto(firstIframeSrc, { waitUntil: 'networkidle', timeout: 60000 });
-    await handleSpecificTurnstile(iframePage, domain);
-
-    const finalIframeSrc = await extractFinalIframeSrc(iframePage, domain);
-    if (!finalIframeSrc) throw new Error('Final iframe source not found');
-    console.log(`[${domain}] Found final iframe src: ${finalIframeSrc}`);
-
-    // 阶段 3: 提取数据
-    const { videoFileUrl, subtitleSources } = await getVideoAndSubtitles(finalIframeSrc);
-    if (!videoFileUrl) throw new Error("HLS URL not found");
-    
-    return { source_domain: domain, hls_url: videoFileUrl, subtitles: subtitleSources, error: null };
-
-  } catch (error) {
-    console.error(`[${domain}] Error in scrapeProvider: ${error.message}`);
-    if (page && !page.isClosed()) {
-        const timestamp = new Date().toISOString().replace(/:/g, '-');
-        const safeDomain = domain.replace(/https?:\/\//, '').replace(/\./g, '_');
-        const screenshotPath = `debug_screenshot_${safeDomain}_${timestamp}.png`;
-        await page.screenshot({ path: screenshotPath, fullPage: true });
-        console.log(`[DEBUG] Saved screenshot to ${screenshotPath}`);
+    if (signal.aborted) throw new Error('Aborted');
+    console.log(`\n[${domain}] Starting scrape for URL: ${url}`);
+
+    let browserInstance = null;
+    let context = null;
+
+    const cleanup = async () => {
+        if (context) await context.close().catch(() => {});
+        if (browserInstance) {
+            console.log(`[${domain}] Releasing browser ${browserInstance.id} back to pool.`);
+            await browserPool.release(browserInstance);
+        }
+    };
+
+    try {
+        browserInstance = await browserPool.get();
+        const browser = browserInstance.browser;
+        
+        context = await browser.newContext({
+            userAgent: "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/117.0.0.0 Safari/537.36",
+            ignoreHTTPSErrors: true,
+            javaScriptEnabled: true,
+            bypassCSP: true
+        });
+
+        await context.addInitScript(() => {
+            delete navigator.__proto__.webdriver;
+            Object.defineProperty(navigator, 'plugins', {
+                get: () => [{
+                    0: { type: "application/x-google-chrome-pdf" },
+                    description: "Portable Document Format",
+                    filename: "internal-pdf-viewer",
+                    length: 1,
+                    name: "Chrome PDF Plugin"
+                }]
+            });
+        });
+
+        const page = await context.newPage();
+        
+        // Setup request interception
+        await page.route('**/rcp_verify', (route) => {
+            route.fulfill({
+                status: 200,
+                contentType: 'application/json',
+                body: '1'
+            });
+        });
+
+        // Stage 1: Initial page
+        console.log(`[${domain}] Navigating and waiting for network to be idle...`);
+        await page.goto(url, { 
+            waitUntil: 'networkidle',
+            timeout: 60000 
+        });
+        
+        await handleSpecificTurnstile(page, domain);
+
+        const firstIframeSrc = await extractFirstIframeSrc(page, domain);
+        if (!firstIframeSrc) throw new Error('First iframe not found');
+        console.log(`[${domain}] Found first iframe src: ${firstIframeSrc}`);
+
+        // Stage 2: First iframe
+        const iframePage = await context.newPage();
+        await iframePage.goto(firstIframeSrc, { 
+            waitUntil: 'networkidle',
+            timeout: 60000 
+        });
+        await handleSpecificTurnstile(iframePage, domain);
+
+        const finalIframeSrc = await extractFinalIframeSrc(iframePage, domain);
+        if (!finalIframeSrc) throw new Error('Final iframe source not found');
+        console.log(`[${domain}] Found final iframe src: ${finalIframeSrc}`);
+
+        // Stage 3: Extract data
+        const { videoFileUrl, subtitleSources } = await getVideoAndSubtitles(finalIframeSrc);
+        if (!videoFileUrl) throw new Error("HLS URL not found");
+        
+        return { source_domain: domain, hls_url: videoFileUrl, subtitles: subtitleSources, error: null };
+
+    } catch (error) {
+        console.error(`[${domain}] Error in scrapeProvider: ${error.message}`);
+        if (page && !page.isClosed()) {
+            const timestamp = new Date().toISOString().replace(/:/g, '-');
+            const safeDomain = domain.replace(/https?:\/\//, '').replace(/\./g, '_');
+            const screenshotPath = `debug_screenshot_${safeDomain}_${timestamp}.png`;
+            await page.screenshot({ path: screenshotPath, fullPage: true });
+            console.log(`[DEBUG] Saved screenshot to ${screenshotPath}`);
+        }
+        throw error;
+    } finally {
+        await cleanup();
     }
-    throw error;
-  } finally {
-    await cleanup();
-  }
 }
 
-// --- 您的 Express 路由和启动逻辑 ---
 app.get("/extract", async (req, res) => {
-  const type = req.query.type || "movie";
-  const tmdb_id = req.query.tmdb_id;
-  const season = req.query.season ? parseInt(req.query.season) : undefined;
-  const episode = req.query.episode ? parseInt(req.query.episode) : undefined;
-
-  if (!tmdb_id) { return res.status(400).json({ success: false, error: "tmdb_id is required" }); }
-  if (type === "tv" && (season == null || episode == null)) { return res.status(400).json({ success: false, error: "season and episode are required" }); }
-
-  const cacheKey = JSON.stringify(req.query);
-  const cached = cache.get(cacheKey);
-  if (cached) { console.log("Serving from cache"); return res.json(cached); }
-
-  const urls = PROVIDERS.reduce((acc, domain) => {
-    acc[domain] = type === "tv" ? `${domain}/embed/tv?tmdb=${tmdb_id}&season=${season}&episode=${episode}` : `${domain}/embed/movie/${tmdb_id}`;
-    return acc;
-  }, {});
-
-  const controller = new AbortController();
-  const signal = controller.signal;
-
-  try {
-    const promises = Object.entries(urls).map(([domain, url]) =>
-      limit(() => scrapeProvider(domain, url, signal))
-    );
-    const firstSuccessfulResult = await Promise.any(promises);
-    controller.abort();
-    const response = { success: true, result: firstSuccessfulResult };
-    cache.set(cacheKey, response);
-    res.json(response);
-  } catch (err) {
-    if (err instanceof AggregateError) {
-      console.error("All providers failed to find a link.");
-      res.status(404).json({ success: false, error: "Could not find the video from any provider." });
-    } else {
-      console.error("An unexpected server error occurred:", err);
-      res.status(500).json({ success: false, error: "Unexpected server error" });
+    const type = req.query.type || "movie";
+    const tmdb_id = req.query.tmdb_id;
+    const season = req.query.season ? parseInt(req.query.season) : undefined;
+    const episode = req.query.episode ? parseInt(req.query.episode) : undefined;
+
+    if (!tmdb_id) { return res.status(400).json({ success: false, error: "tmdb_id is required" }); }
+    if (type === "tv" && (season == null || episode == null)) { return res.status(400).json({ success: false, error: "season and episode are required" }); }
+
+    const cacheKey = JSON.stringify(req.query);
+    const cached = cache.get(cacheKey);
+    if (cached) { 
+        console.log("Serving from cache");
+        return res.json(cached); 
+    }
+
+    const urls = PROVIDERS.reduce((acc, domain) => {
+        acc[domain] = type === "tv" ? `${domain}/embed/tv?tmdb=${tmdb_id}&season=${season}&episode=${episode}` : `${domain}/embed/movie/${tmdb_id}`;
+        return acc;
+    }, {});
+
+    const controller = new AbortController();
+    const signal = controller.signal;
+
+    try {
+        const promises = Object.entries(urls).map(([domain, url]) =>
+            limit(() => scrapeProvider(domain, url, signal))
+        );
+        const firstSuccessfulResult = await Promise.any(promises);
+        controller.abort();
+        const response = { success: true, result: firstSuccessfulResult };
+        cache.set(cacheKey, response);
+        res.json(response);
+    } catch (err) {
+        if (err instanceof AggregateError) {
+            console.error("All providers failed to find a link.");
+            res.status(404).json({ success: false, error: "Could not find the video from any provider." });
+        } else {
+            console.error("An unexpected server error occurred:", err);
+            res.status(500).json({ success: false, error: "Unexpected server error" });
+        }
     }
-  }
 });
 
 (async () => {
-  try {
-    browserPool = new BrowserPool({
-      chromium: playwright.chromium,
-      minSize: 1,
-      maxSize: 5,
-      maxUsage: 100,
-      launchOptions: {
-          headless: true,
-          args: ['--no-sandbox', '--disable-setuid-sandbox', '--disable-dev-shm-usage']
-      }
-    });
-    await browserPool.initialize();
-    console.log("Browser pool initialized successfully.");
-    app.listen(PORT, () => console.log(`🚀 Universal Video Extractor running at http://localhost:${PORT}`));
-  } catch (error) {
-    console.error("Failed to initialize browser pool:", error);
-    process.exit(1);
-  }
+    try {
+        browserPool = new BrowserPool({
+            chromium: chromium,
+            minSize: 1,
+            maxSize: 5,
+            maxUsage: 100,
+            launchOptions: {
+                headless: true,
+                args: [
+                    '--no-sandbox',
+                    '--disable-setuid-sandbox',
+                    '--disable-dev-shm-usage',
+                    '--disable-gpu',
+                    '--disable-extensions',
+                    '--disable-notifications',
+                    '--disable-infobars',
+                    '--disable-web-security',
+                    '--disable-features=IsolateOrigins,site-per-process'
+                ]
+            }
+        });
+        await browserPool.initialize();
+        console.log("Browser pool initialized successfully.");
+        app.listen(PORT, () => console.log(`🚀 Universal Video Extractor running at http://localhost:${PORT}`));
+    } catch (error) {
+        console.error("Failed to initialize browser pool:", error);
+        process.exit(1);
+    }
 })();
 
 process.on("SIGINT", async () => {
@@ -272,8 +311,9 @@ process.on("SIGINT", async () => {
     if (browserPool) await browserPool.shutdown();
     process.exit(0);
 });
+
 process.on("SIGTERM", async () => {
     console.log("Shutting down gracefully...");
     if (browserPool) await browserPool.shutdown();
     process.exit(0);
-});
+});