Remove firefox puppeteer fallback

2025-11-22 20:34:29 -05:00
parent e5d769d59b
commit 9df8c33e5e
5 changed files with 329 additions and 128 deletions
--- a/go.mod
+++ b/go.mod
@@ -6,8 +6,6 @@ require (
        github.com/charmbracelet/bubbles v0.18.0
        github.com/charmbracelet/bubbletea v0.26.6
        github.com/charmbracelet/lipgloss v0.13.0
-	github.com/chromedp/cdproto v0.0.0-20250803210736-d308e07a266d
-	github.com/chromedp/chromedp v0.14.2
 )

 require (
@@ -16,7 +14,6 @@ require (
 	github.com/charmbracelet/x/input v0.1.0 // indirect
 	github.com/charmbracelet/x/term v0.1.1 // indirect
 	github.com/charmbracelet/x/windows v0.1.0 // indirect
-	github.com/chromedp/sysutil v1.1.0 // indirect
        github.com/erikgeiser/coninput v0.0.0-20211004153227-1c3628e74d0f // indirect
        github.com/go-json-experiment/json v0.0.0-20250725192818-e39067aee2d2 // indirect
        github.com/gobwas/httphead v0.1.0 // indirect
--- a/internal/app.go
+++ b/internal/app.go
@@ -107,13 +107,13 @@ type Model struct {
 // ENTRY POINT
 // ────────────────────────────────

-func Run() error {
-	p := tea.NewProgram(New(), tea.WithAltScreen())
+func Run(debug bool) error {
+	p := tea.NewProgram(New(debug), tea.WithAltScreen())
 	_, err := p.Run()
 	return err
 }

-func New() Model {
+func New(debug bool) Model {
 	base := BaseURLFromEnv()
 	client := NewClient(base, 15*time.Second)
 	styles := NewStyles()
@@ -128,6 +128,10 @@ func New() Model {
 		debugLines:  []string{},
 	}

+	if debug {
+		m.currentView = viewDebug
+	}
+
 	m.sports = NewListColumn[Sport]("Sports", func(s Sport) string { return s.Name })
 	m.matches = NewListColumn[Match]("Popular Matches", func(mt Match) string {
 		when := time.UnixMilli(mt.Date).Local().Format("Jan 2 15:04")
@@ -452,7 +456,7 @@ func (m Model) runExtractor(st Stream) tea.Cmd {
 			}
 		}

-		logcb(fmt.Sprintf("[extractor] Starting Chrome-based extractor for %s", st.EmbedURL))
+		logcb(fmt.Sprintf("[extractor] Starting puppeteer extractor for %s", st.EmbedURL))

 		m3u8, hdrs, err := extractM3U8Lite(st.EmbedURL, func(line string) {
 			m.debugLines = append(m.debugLines, line)
--- a/internal/columns.go
+++ b/internal/columns.go
@@ -71,7 +71,11 @@ func (c *ListColumn[T]) SetWidth(w int) {
 	c.width = w - 4
 }

-func (c *ListColumn[T]) SetHeight(h int)       { if h > 5 { c.height = h - 5 } }
+func (c *ListColumn[T]) SetHeight(h int) {
+	if h > 5 {
+		c.height = h - 5
+	}
+}

 func (c *ListColumn[T]) CursorUp() {
 	if c.selected > 0 {
--- a/internal/extractor.go
+++ b/internal/extractor.go
@@ -1,118 +1,260 @@
 package internal

 import (
-	"context"
+	"bytes"
+	"encoding/json"
 	"errors"
 	"fmt"
 	"os"
 	"os/exec"
+	"path/filepath"
 	"strings"
 	"time"
-
-	"github.com/chromedp/cdproto/network"
-	"github.com/chromedp/chromedp"
 )

-// extractM3U8Lite loads an embed page in headless Chrome via chromedp,
-// runs any JavaScript, and extracts the final .m3u8 URL and its HTTP headers.
-// It streams live log lines via the provided log callback.
+type puppeteerResult struct {
+	URL     string            `json:"url"`
+	Headers map[string]string `json:"headers"`
+	Browser string            `json:"browser"`
+}
+
+func ensurePuppeteerAvailable() error {
+	if _, err := exec.LookPath("node"); err != nil {
+		return fmt.Errorf("node executable not found: %w", err)
+	}
+
+	// Verify both puppeteer-extra and the stealth plugin are available from the
+	// project directory so the temporary runner can load them relative to cwd.
+	requireScript := strings.Join([]string{
+		"const { createRequire } = require('module');",
+		"const req = createRequire(process.cwd() + '/');",
+		"req.resolve('puppeteer-extra/package.json');",
+		"req.resolve('puppeteer-extra-plugin-stealth/package.json');",
+	}, "")
+
+	check := exec.Command("node", "-e", requireScript)
+	if err := check.Run(); err != nil {
+		return fmt.Errorf("puppeteer-extra or stealth plugin missing. Run `npm install puppeteer-extra puppeteer-extra-plugin-stealth puppeteer` in the project directory: %w", err)
+	}
+
+	return nil
+}
+
+// extractM3U8Lite invokes a small Puppeteer runner that loads the embed page,
+// watches for .m3u8 requests, and returns the first match plus its request
+// headers.
 func extractM3U8Lite(embedURL string, log func(string)) (string, map[string]string, error) {
 	if log == nil {
 		log = func(string) {}
 	}

-	ctx, cancel := chromedp.NewContext(context.Background())
-	defer cancel()
+	if strings.TrimSpace(embedURL) == "" {
+		return "", nil, errors.New("empty embed URL")
+	}

-	// Capture the first .m3u8 request Chrome makes
-	type capture struct {
-		URL     string
-		Headers map[string]string
-	}
-	found := make(chan capture, 1)
-
-	chromedp.ListenTarget(ctx, func(ev interface{}) {
-		if e, ok := ev.(*network.EventRequestWillBeSent); ok {
-			u := e.Request.URL
-			if strings.Contains(u, ".m3u8") {
-				h := make(map[string]string)
-				for k, v := range e.Request.Headers {
-					if s, ok := v.(string); ok {
-						h[k] = s
-					}
-				}
-				select {
-				case found <- capture{URL: u, Headers: h}:
-				default:
-				}
-			}
-		}
-	})
-
-	log(fmt.Sprintf("[chromedp] launching Chrome for %s", embedURL))
-
-	// Set reasonable headers for navigation
-	headers := network.Headers{
-		"User-Agent": "Mozilla/5.0 (X11; Linux x86_64) Gecko/20100101 Firefox/144.0",
-	}
-	ctxTimeout, cancelNav := context.WithTimeout(ctx, 30*time.Second)
-	defer cancelNav()
-
-	if err := chromedp.Run(ctxTimeout,
-		network.Enable(),
-		network.SetExtraHTTPHeaders(headers),
-		chromedp.Navigate(embedURL),
-		chromedp.WaitReady("body", chromedp.ByQuery),
-	); err != nil {
-		log(fmt.Sprintf("[chromedp] navigation error: %v", err))
+	if err := ensurePuppeteerAvailable(); err != nil {
 		return "", nil, err
 	}

-	log("[chromedp] page loaded, waiting for .m3u8 network requests...")
+	runnerPath, err := writePuppeteerRunner()
+	if err != nil {
+		return "", nil, err
+	}
+	defer os.Remove(runnerPath)

-	select {
-	case cap := <-found:
-		log(fmt.Sprintf("[chromedp] ✅ found .m3u8 via network: %s", cap.URL))
-		log(fmt.Sprintf("[chromedp] captured %d headers", len(cap.Headers)))
-		return cap.URL, cap.Headers, nil
-	case <-time.After(12 * time.Second):
-		log("[chromedp] timeout waiting for .m3u8 request, attempting DOM fallback...")
+	log(fmt.Sprintf("[puppeteer] launching chromium stealth runner for %s", embedURL))
+
+	cmd := exec.Command("node", runnerPath, embedURL)
+	var stdout bytes.Buffer
+	var stderr bytes.Buffer
+	cmd.Stdout = &stdout
+	cmd.Stderr = &stderr
+
+	if err := cmd.Run(); err != nil {
+		log(fmt.Sprintf("[puppeteer] runner error: %s", strings.TrimSpace(stderr.String())))
+		return "", nil, fmt.Errorf("puppeteer runner failed: %w", err)
 	}

-	// DOM fallback: look for <video> src or inline JS with a URL
-	var candidate string
-	if err := chromedp.Run(ctx,
-		chromedp.EvaluateAsDevTools(`(function(){
+	var res puppeteerResult
+	if err := json.Unmarshal(stdout.Bytes(), &res); err != nil {
+		log(fmt.Sprintf("[puppeteer] decode error: %v", err))
+		return "", nil, err
+	}
+
+	if res.URL == "" {
+		if stderr.Len() > 0 {
+			log(strings.TrimSpace(stderr.String()))
+		}
+		return "", nil, errors.New("m3u8 not found")
+	}
+
+	log(fmt.Sprintf("[puppeteer] ✅ found .m3u8 via %s: %s", res.Browser, res.URL))
+	return res.URL, res.Headers, nil
+}
+
+// writePuppeteerRunner materializes a temporary Node.js script that performs
+// the actual page load and .m3u8 discovery with puppeteer-extra stealth
+// protections.
+func writePuppeteerRunner() (string, error) {
+	script := `const { createRequire } = require('module');
+const requireFromCwd = createRequire(process.cwd() + '/');
+
+let puppeteer;
+let StealthPlugin;
+try {
+  puppeteer = requireFromCwd('puppeteer-extra');
+  StealthPlugin = requireFromCwd('puppeteer-extra-plugin-stealth');
+  puppeteer.use(StealthPlugin());
+} catch (err) {
+  console.error('[puppeteer] required packages missing. install with "npm install puppeteer-extra puppeteer-extra-plugin-stealth puppeteer" in the project directory.');
+  process.exit(1);
+}
+
+const embedURL = process.argv[2];
+const timeoutMs = 45000;
+
+if (!embedURL) {
+  console.error('missing embed URL');
+  process.exit(1);
+}
+
+const viewport = { width: 1280, height: 720 };
+const launchArgs = ['--disable-blink-features=AutomationControlled', '--no-sandbox', '--disable-web-security', '--window-size=1920,1080'];
+const userAgent = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36';
+
+async function launchBrowser() {
+  const chromiumOptions = {
+    headless: 'new',
+    args: launchArgs,
+    defaultViewport: viewport,
+  };
+  const browser = await puppeteer.launch(chromiumOptions);
+  return { browser, flavor: 'chromium' };
+}
+
+function installTouchAndWindowSpoofing(page) {
+  return page.evaluateOnNewDocument(() => {
+    const { width, height } = window.screen || { width: 1920, height: 1080 };
+    Object.defineProperty(navigator, 'maxTouchPoints', { get: () => 1 });
+    Object.defineProperty(navigator, 'platform', { get: () => 'Linux x86_64' });
+    Object.defineProperty(navigator, 'hardwareConcurrency', { get: () => 8 });
+    Object.defineProperty(window, 'outerWidth', { get: () => width });
+    Object.defineProperty(window, 'outerHeight', { get: () => height });
+  });
+}
+
+(async () => {
+  const { browser, flavor } = await launchBrowser();
+  const page = await browser.newPage();
+  await installTouchAndWindowSpoofing(page);
+
+  await page.setUserAgent(userAgent);
+  await page.setViewport(viewport);
+  await page.setExtraHTTPHeaders({
+    'accept-language': 'en-US,en;q=0.9',
+    'sec-fetch-site': 'same-origin',
+    'sec-fetch-mode': 'navigate',
+    'sec-fetch-user': '?1',
+    'sec-fetch-dest': 'document',
+    'sec-ch-ua': '"Chromium";v="124", "Not=A?Brand";v="99", "Google Chrome";v="124"',
+    'sec-ch-ua-platform': 'Linux',
+    'sec-ch-ua-mobile': '?0',
+  });
+
+  let captured = null;
+  const capturePromise = new Promise(resolve => {
+    page.on('request', req => {
+      const url = req.url();
+      if (!captured && url.includes('.m3u8')) {
+        const headers = req.headers();
+        captured = { url, headers };
+        resolve();
+      }
+    });
+  });
+
  try {
-				const v = document.querySelector('video');
-				if(v){
-					if(v.currentSrc) return v.currentSrc;
-					if(v.src) return v.src;
-					const s = v.querySelector('source');
-					if(s && s.src) return s.src;
+    await page.goto(embedURL, { waitUntil: 'networkidle2', timeout: timeoutMs });
+    await page.waitForNavigation({ waitUntil: 'networkidle2', timeout: timeoutMs }).catch(() => {});
+  } catch (err) {
+    console.error('[puppeteer] navigation warning: ' + err.message);
+  }
+
+  await Promise.race([
+    capturePromise,
+    new Promise(resolve => setTimeout(resolve, 20000)),
+  ]);
+
+  if (!captured) {
+    const candidate = await page.evaluate(() => {
+      try {
+        const video = document.querySelector('video');
+        if (video) {
+          if (video.currentSrc) return video.currentSrc;
+          if (video.src) return video.src;
+          const source = video.querySelector('source');
+          if (source && source.src) return source.src;
        }
        const html = document.documentElement.innerHTML;
        const match = html.match(/https?:\/\/[^'"\s]+\.m3u8[^'"\s]*/i);
-				if(match) return match[0];
-			}catch(e){}
+        if (match) return match[0];
+      } catch (e) {}
      return '';
-		})()`, &candidate),
-	); err != nil {
-		log(fmt.Sprintf("[chromedp] DOM evaluation error: %v", err))
+    });
+    if (candidate && candidate.includes('.m3u8')) {
+      captured = { url: candidate, headers: {} };
+    }
  }

-	candidate = strings.TrimSpace(candidate)
-	if candidate != "" && strings.Contains(candidate, ".m3u8") {
-		log(fmt.Sprintf("[chromedp] ✅ found .m3u8 via DOM: %s", candidate))
-		return candidate, map[string]string{}, nil
+  if (captured) {
+    // Enrich headers with cookies and referer if missing.
+    const cookies = await page.cookies();
+    if (cookies && cookies.length > 0) {
+      const cookieHeader = cookies.map(c => c.name + '=' + c.value).join('; ');
+      if (!captured.headers) captured.headers = {};
+      captured.headers['cookie'] = captured.headers['cookie'] || cookieHeader;
+    }
+    captured.headers = captured.headers || {};
+    captured.headers['user-agent'] = userAgent;
+    captured.headers['referer'] = captured.headers['referer'] || embedURL;
+    try {
+      const origin = new URL(embedURL).origin;
+      captured.headers['origin'] = captured.headers['origin'] || origin;
+    } catch (e) {}
  }

-	log("[chromedp] ❌ failed to find .m3u8 via network or DOM")
-	return "", nil, errors.New("m3u8 not found")
+  await browser.close();
+
+  const output = captured || { url: '', headers: {} };
+  output.browser = flavor;
+  console.log(JSON.stringify(output));
+})().catch(err => {
+  console.error(err.stack || err.message);
+  process.exit(1);
+});
+`
+
+	dir := os.TempDir()
+	path := filepath.Join(dir, fmt.Sprintf("puppeteer-runner-%d.js", time.Now().UnixNano()))
+	if err := os.WriteFile(path, []byte(script), 0o600); err != nil {
+		return "", err
+	}
+	return path, nil
 }

-// LaunchMPVWithHeaders spawns mpv to play the given M3U8 URL,
-// reusing all captured HTTP headers to mimic browser playback.
+// lookupHeaderValue returns the first header value matching name, using a
+// case-insensitive comparison for keys sourced from the Puppeteer request map.
+func lookupHeaderValue(hdrs map[string]string, name string) string {
+	for k, v := range hdrs {
+		if strings.EqualFold(k, name) {
+			return v
+		}
+	}
+	return ""
+}
+
+// LaunchMPVWithHeaders spawns mpv to play the given M3U8 URL using the minimal
+// header set required for successful playback (User-Agent, Origin, Referer).
 // Logs are streamed via the provided callback.
 func LaunchMPVWithHeaders(m3u8 string, hdrs map[string]string, log func(string)) error {
 	if log == nil {
@@ -124,20 +266,27 @@ func LaunchMPVWithHeaders(m3u8 string, hdrs map[string]string, log func(string))

 	args := []string{"--no-terminal", "--really-quiet"}

-	for k, v := range hdrs {
-		if k == "" || v == "" {
-			continue
+	// Only forward the minimal headers mpv requires to mirror the working
+	// curl→mpv handoff: User-Agent, Origin, and Referer. Extra headers
+	// captured in the browser session can cause mpv to reject the request
+	// or send malformed values when duplicated, so we constrain the set
+	// explicitly and tolerate case-insensitive keys from Puppeteer.
+	headerKeys := []struct {
+		lookup  string
+		display string
+	}{
+		{lookup: "user-agent", display: "User-Agent"},
+		{lookup: "origin", display: "Origin"},
+		{lookup: "referer", display: "Referer"},
 	}
-		switch strings.ToLower(k) {
-		case "accept-encoding", "sec-fetch-site", "sec-fetch-mode", "sec-fetch-dest",
-			"sec-ch-ua", "sec-ch-ua-platform", "sec-ch-ua-mobile":
-			continue // ignore internal Chromium headers
+	for _, hk := range headerKeys {
+		if v := lookupHeaderValue(hdrs, hk.lookup); v != "" {
+			args = append(args, fmt.Sprintf("--http-header-fields=%s: %s", hk.display, v))
 		}
-		args = append(args, fmt.Sprintf("--http-header-fields=%s: %s", k, v))
 	}

 	args = append(args, m3u8)
-	log(fmt.Sprintf("[mpv] launching with %d headers: %s", len(hdrs), m3u8))
+	log(fmt.Sprintf("[mpv] launching with %d headers: %s", len(args)-3, m3u8))

 	cmd := exec.Command("mpv", args...)
 	cmd.Stdout = os.Stdout
@@ -151,3 +300,37 @@ func LaunchMPVWithHeaders(m3u8 string, hdrs map[string]string, log func(string))
 	log(fmt.Sprintf("[mpv] started (pid %d)", cmd.Process.Pid))
 	return nil
 }
+
+// RunExtractorCLI provides a non-TUI entry point to run the extractor directly
+// from the command line ("-e <embedURL>"). When debug is true, verbose output
+// from the Puppeteer runner and mpv launch is printed to stdout.
+func RunExtractorCLI(embedURL string, debug bool) error {
+	if strings.TrimSpace(embedURL) == "" {
+		return errors.New("missing embed URL")
+	}
+
+	logger := func(string) {}
+	if debug {
+		logger = func(line string) { fmt.Println(line) }
+	}
+
+	fmt.Printf("[extractor] starting for %s\n", embedURL)
+	m3u8, hdrs, err := extractM3U8Lite(embedURL, logger)
+	if err != nil {
+		fmt.Printf("[extractor] ❌ %v\n", err)
+		return err
+	}
+
+	fmt.Printf("[extractor] ✅ found M3U8: %s\n", m3u8)
+	if len(hdrs) > 0 && debug {
+		fmt.Printf("[extractor] captured %d headers\n", len(hdrs))
+	}
+
+	if err := LaunchMPVWithHeaders(m3u8, hdrs, logger); err != nil {
+		fmt.Printf("[mpv] ❌ %v\n", err)
+		return err
+	}
+
+	fmt.Println("[mpv] ▶ streaming started")
+	return nil
+}
--- a/main.go
+++ b/main.go
@@ -1,6 +1,7 @@
 package main

 import (
+	"flag"
 	"log"
 	"os"

@@ -8,7 +9,19 @@ import (
 )

 func main() {
-	if err := internal.Run(); err != nil {
+	embedURL := flag.String("e", "", "extract a single embed URL and launch mpv")
+	debug := flag.Bool("debug", false, "enable verbose extractor/debug output")
+	flag.Parse()
+
+	if *embedURL != "" {
+		if err := internal.RunExtractorCLI(*embedURL, *debug); err != nil {
+			log.Println("error:", err)
+			os.Exit(1)
+		}
+		return
+	}
+
+	if err := internal.Run(*debug); err != nil {
 		log.Println("error:", err)
 		os.Exit(1)
 	}