Initial Exctractor work

2025-10-20 19:56:37 -04:00
parent 9ad5a2bc53
commit e5d769d59b
3 changed files with 201 additions and 174 deletions
--- a/go.mod
+++ b/go.mod
@@ -1,11 +1,13 @@
 module github.com/Salastil/streamed-tui

-go 1.22
+go 1.24

 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 (
@@ -14,7 +16,12 @@ 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
+	github.com/gobwas/pool v0.2.1 // indirect
+	github.com/gobwas/ws v1.4.0 // indirect
 	github.com/lucasb-eyer/go-colorful v1.2.0 // indirect
 	github.com/mattn/go-isatty v0.0.20 // indirect
 	github.com/mattn/go-localereader v0.0.1 // indirect
@@ -25,6 +32,6 @@ require (
 	github.com/rivo/uniseg v0.4.7 // indirect
 	github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e // indirect
 	golang.org/x/sync v0.7.0 // indirect
-	golang.org/x/sys v0.21.0 // indirect
+	golang.org/x/sys v0.34.0 // indirect
 	golang.org/x/text v0.3.8 // indirect
 )
--- a/internal/app.go
+++ b/internal/app.go
@@ -3,8 +3,6 @@ package internal
 import (
 	"context"
 	"fmt"
-	"os"
-	"os/exec"
 	"strings"
 	"time"

@@ -14,6 +12,10 @@ import (
 	"github.com/charmbracelet/lipgloss"
 )

+// ────────────────────────────────
+// KEYMAP
+// ────────────────────────────────
+
 type keyMap struct {
 	Up, Down, Left, Right key.Binding
 	Enter, Quit, Refresh  key.Binding
@@ -40,6 +42,7 @@ func defaultKeys() keyMap {
 func (k keyMap) ShortHelp() []key.Binding {
 	return []key.Binding{k.Up, k.Down, k.Left, k.Right, k.Enter, k.OpenBrowser, k.OpenMPV, k.Quit}
 }
+
 func (k keyMap) FullHelp() [][]key.Binding {
 	return [][]key.Binding{
 		{k.Up, k.Down, k.Left, k.Right},
@@ -47,6 +50,10 @@ func (k keyMap) FullHelp() [][]key.Binding {
 	}
 }

+// ────────────────────────────────
+// TYPES & CONSTANTS
+// ────────────────────────────────
+
 type (
 	sportsLoadedMsg  []Sport
 	matchesLoadedMsg struct {
@@ -74,6 +81,10 @@ const (
 	viewDebug
 )

+// ────────────────────────────────
+// MODEL
+// ────────────────────────────────
+
 type Model struct {
 	apiClient   *Client
 	styles      Styles
@@ -92,6 +103,10 @@ type Model struct {
 	TerminalWidth int
 }

+// ────────────────────────────────
+// ENTRY POINT
+// ────────────────────────────────
+
 func Run() error {
 	p := tea.NewProgram(New(), tea.WithAltScreen())
 	_, err := p.Run()
@@ -134,6 +149,10 @@ func New() Model {
 	return m
 }

+// ────────────────────────────────
+// VIEW MANAGEMENT
+// ────────────────────────────────
+
 func (m Model) Init() tea.Cmd {
 	return tea.Batch(m.fetchSports(), m.fetchPopularMatches())
 }
@@ -212,6 +231,10 @@ func (m Model) renderDebugPanel() string {
 	return panel
 }

+// ────────────────────────────────
+// UPDATE LOOP
+// ────────────────────────────────
+
 func (m Model) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
 	switch msg := msg.(type) {

@@ -368,6 +391,10 @@ func (m Model) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
 	return m, nil
 }

+// ────────────────────────────────
+// FETCHERS
+// ────────────────────────────────
+
 func (m Model) fetchSports() tea.Cmd {
 	return func() tea.Msg {
 		sports, err := m.apiClient.GetSports(context.Background())
@@ -408,33 +435,52 @@ func (m Model) fetchStreamsForMatch(mt Match) tea.Cmd {
 	}
 }

+// ────────────────────────────────
+// EXTRACTOR (chromedp integration)
+// ────────────────────────────────
+
 func (m Model) runExtractor(st Stream) tea.Cmd {
 	return func() tea.Msg {
 		if st.EmbedURL == "" {
 			return debugLogMsg("Extractor aborted: empty embed URL")
 		}
-		m3u8, err := extractM3U8Lite(st.EmbedURL, func(line string) {
-			// each extractor log line will flow back to the UI
-			return
+
+		logcb := func(line string) {
+			m.debugLines = append(m.debugLines, line)
+			if len(m.debugLines) > 200 {
+				m.debugLines = m.debugLines[len(m.debugLines)-200:]
+			}
+		}
+
+		logcb(fmt.Sprintf("[extractor] Starting Chrome-based extractor for %s", st.EmbedURL))
+
+		m3u8, hdrs, err := extractM3U8Lite(st.EmbedURL, func(line string) {
+			m.debugLines = append(m.debugLines, line)
 		})
 		if err != nil {
+			logcb(fmt.Sprintf("[extractor] ❌ %v", err))
 			return debugLogMsg(fmt.Sprintf("Extractor failed: %v", err))
 		}
-		cmd := exec.Command("mpv",
-			"--no-terminal",
-			"--really-quiet",
-			fmt.Sprintf("--http-header-fields=User-Agent: %s", "Mozilla/5.0 (X11; Linux x86_64) Gecko/20100101 Firefox/144.0"),
-			fmt.Sprintf("--http-header-fields=Origin: %s", "https://embedsports.top"),
-			fmt.Sprintf("--http-header-fields=Referer: %s", "https://embedsports.top/"),
-			m3u8,
-		)
-		cmd.Stdout = os.Stdout
-		cmd.Stderr = os.Stderr
-		_ = cmd.Start()
-		return debugLogMsg(fmt.Sprintf("Extractor success, launched MPV: %s", m3u8))
+
+		logcb(fmt.Sprintf("[extractor] ✅ Found M3U8: %s", m3u8))
+		if len(hdrs) > 0 {
+			logcb(fmt.Sprintf("[extractor] Captured %d headers", len(hdrs)))
+		}
+
+		if err := LaunchMPVWithHeaders(m3u8, hdrs, logcb); err != nil {
+			logcb(fmt.Sprintf("[mpv] ❌ %v", err))
+			return debugLogMsg(fmt.Sprintf("MPV error: %v", err))
+		}
+
+		logcb(fmt.Sprintf("[mpv] ▶ Streaming started for %s", st.EmbedURL))
+		return debugLogMsg("Extractor completed successfully")
 	}
 }

+// ────────────────────────────────
+// LOG TO UI
+// ────────────────────────────────
+
 func (m Model) logToUI(line string) tea.Cmd {
 	return func() tea.Msg {
 		return debugLogMsg(line)
--- a/internal/extractor.go
+++ b/internal/extractor.go
@@ -1,179 +1,153 @@
 package internal

 import (
-	"encoding/base64"
+	"context"
 	"errors"
 	"fmt"
-	"net/http"
-	"regexp"
+	"os"
+	"os/exec"
+	"strings"
 	"time"
+
+	"github.com/chromedp/cdproto/network"
+	"github.com/chromedp/chromedp"
 )

-// extractM3U8Lite performs static extraction like yt-dlp,
-// logging each step to the provided debug callback (for the F12 panel).
-func extractM3U8Lite(embedURL string, log func(string)) (string, error) {
-	const origin = "https://embedsports.top"
-	const referer = "https://embedsports.top/"
-	const ua = "Mozilla/5.0 (X11; Linux x86_64; rv:144.0) Gecko/20100101 Firefox/144.0"
-
-	if log != nil {
-		log(fmt.Sprintf("[extractor] Fetching embed page: %s", embedURL))
+// 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.
+func extractM3U8Lite(embedURL string, log func(string)) (string, map[string]string, error) {
+	if log == nil {
+		log = func(string) {}
 	}

-	body, err := fetchHTML(embedURL, ua, origin, referer, 10*time.Second)
-	if err != nil {
-		if log != nil {
-			log(fmt.Sprintf("[extractor] ❌ fetch failed: %v", err))
+	ctx, cancel := chromedp.NewContext(context.Background())
+	defer cancel()
+
+	// Capture the first .m3u8 request Chrome makes
+	type capture struct {
+		URL     string
+		Headers map[string]string
 	}
-		return "", fmt.Errorf("fetch embed: %w", err)
+	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))
+		return "", nil, err
 	}

-	k, i, s := parseEmbedVars(body)
-	if log != nil {
-		log(fmt.Sprintf("[extractor] Parsed vars → k=%q i=%q s=%q", k, i, s))
+	log("[chromedp] page loaded, waiting for .m3u8 network requests...")
+
+	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...")
 	}

-	if k == "" || i == "" || s == "" {
-		if log != nil {
-			log("[extractor] ❌ missing vars (k,i,s) in embed HTML")
+	// DOM fallback: look for <video> src or inline JS with a URL
+	var candidate string
+	if err := chromedp.Run(ctx,
+		chromedp.EvaluateAsDevTools(`(function(){
+			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;
 				}
-		return "", errors.New("missing vars (k,i,s) in embed HTML")
+				const html = document.documentElement.innerHTML;
+				const match = html.match(/https?:\/\/[^'"\s]+\.m3u8[^'"\s]*/i);
+				if(match) return match[0];
+			}catch(e){}
+			return '';
+		})()`, &candidate),
+	); err != nil {
+		log(fmt.Sprintf("[chromedp] DOM evaluation error: %v", err))
 	}

-	if log != nil {
-		log("[extractor] Trying to extract secure token from bundle.js")
-	}
-	token, ok := extractTokenFromJS(origin+"/js/bundle.js", ua, origin, referer, log)
-	if ok {
-		if log != nil {
-			log(fmt.Sprintf("[extractor] Found token: %s", token))
-		}
-		for lb := 1; lb <= 9; lb++ {
-			url := fmt.Sprintf("https://lb%d.strmd.top/secure/%s/%s/stream/%s/%s/playlist.m3u8",
-				lb, token, k, i, s)
-			if log != nil {
-				log(fmt.Sprintf("[extractor] Probing %s", url))
-			}
-			if probeURL(url) {
-				if log != nil {
-					log(fmt.Sprintf("[extractor] ✅ Found working m3u8: %s", url))
-				}
-				return url, nil
-			}
-		}
-		if log != nil {
-			log("[extractor] ⚠️ No working load balancer found with token")
-		}
+	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 log != nil {
-		log("[extractor] Fallback: probing default tokens")
-	}
-	for lb := 1; lb <= 9; lb++ {
-		testURL := fmt.Sprintf("https://lb%d.strmd.top/secure/test/%s/stream/%s/%s/playlist.m3u8",
-			lb, k, i, s)
-		if log != nil {
-			log(fmt.Sprintf("[extractor] Probing fallback %s", testURL))
-		}
-		if probeURL(testURL) {
-			if log != nil {
-				log(fmt.Sprintf("[extractor] ✅ Fallback succeeded: %s", testURL))
-			}
-			return testURL, nil
-		}
+	log("[chromedp] ❌ failed to find .m3u8 via network or DOM")
+	return "", nil, errors.New("m3u8 not found")
 }

-	if log != nil {
-		log("[extractor] ❌ Unable to derive .m3u8 from embed")
+// LaunchMPVWithHeaders spawns mpv to play the given M3U8 URL,
+// reusing all captured HTTP headers to mimic browser playback.
+// Logs are streamed via the provided callback.
+func LaunchMPVWithHeaders(m3u8 string, hdrs map[string]string, log func(string)) error {
+	if log == nil {
+		log = func(string) {}
 	}
-	return "", errors.New("unable to derive .m3u8")
+	if m3u8 == "" {
+		return fmt.Errorf("empty m3u8 URL")
 	}

-// extractTokenFromJS is modified to accept a logger callback
-func extractTokenFromJS(url, ua, origin, referer string, log func(string)) (string, bool) {
-	if log != nil {
-		log(fmt.Sprintf("[extractor] Fetching JS: %s", url))
+	args := []string{"--no-terminal", "--really-quiet"}
+
+	for k, v := range hdrs {
+		if k == "" || v == "" {
+			continue
 		}
-	body, err := fetchHTML(url, ua, origin, referer, 8*time.Second)
-	if err != nil {
-		if log != nil {
-			log(fmt.Sprintf("[extractor] ⚠️ fetch JS failed: %v", err))
+		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
 		}
-		return "", false
+		args = append(args, fmt.Sprintf("--http-header-fields=%s: %s", k, v))
 	}

-	if m := regexp.MustCompile(`https://lb\d+\.strmd\.top/secure/([A-Za-z0-9]+)/`).FindStringSubmatch(body); len(m) > 1 {
-		if log != nil {
-			log("[extractor] Found literal /secure/ token in JS")
-		}
-		return m[1], true
+	args = append(args, m3u8)
+	log(fmt.Sprintf("[mpv] launching with %d headers: %s", len(hdrs), m3u8))
+
+	cmd := exec.Command("mpv", args...)
+	cmd.Stdout = os.Stdout
+	cmd.Stderr = os.Stderr
+
+	if err := cmd.Start(); err != nil {
+		log(fmt.Sprintf("[mpv] launch error: %v", err))
+		return err
 	}

-	if m := regexp.MustCompile(`atob\("([A-Za-z0-9+/=]+)"\)`).FindStringSubmatch(body); len(m) > 1 {
-		if log != nil {
-			log("[extractor] Found base64 token pattern, decoding…")
-		}
-		decoded, err := base64.StdEncoding.DecodeString(m[1])
-		if err == nil {
-			tok := reverseString(string(decoded))
-			if len(tok) > 16 {
-				if log != nil {
-					log(fmt.Sprintf("[extractor] Extracted token (decoded+reversed): %s", tok))
-				}
-				return tok, true
-			}
-		}
-	}
-
-	if m := regexp.MustCompile(`"([A-Za-z0-9]{24,})"\.split\(""\)\.reverse`).FindStringSubmatch(body); len(m) > 1 {
-		if log != nil {
-			log("[extractor] Found reversed string token")
-		}
-		return reverseString(m[1]), true
-	}
-
-	if log != nil {
-		log("[extractor] No token found in JS")
-	}
-	return "", false
-}
-
-// parseEmbedVars extracts var k, i, s from HTML
-func parseEmbedVars(html string) (k, i, s string) {
-	reK := regexp.MustCompile(`var k\s*=\s*"([^"]+)"`)
-	reI := regexp.MustCompile(`var i\s*=\s*"([^"]+)"`)
-	reS := regexp.MustCompile(`var s\s*=\s*"([^"]+)"`)
-
-	if m := reK.FindStringSubmatch(html); len(m) > 1 {
-		k = m[1]
-	}
-	if m := reI.FindStringSubmatch(html); len(m) > 1 {
-		i = m[1]
-	}
-	if m := reS.FindStringSubmatch(html); len(m) > 1 {
-		s = m[1]
-	}
-	return
-}
-
-// probeURL sends a quick HEAD to see if the URL exists
-func probeURL(u string) bool {
-	client := &http.Client{Timeout: 4 * time.Second}
-	req, _ := http.NewRequest("HEAD", u, nil)
-	req.Header.Set("User-Agent", "Mozilla/5.0")
-	resp, err := client.Do(req)
-	if err != nil {
-		return false
-	}
-	defer resp.Body.Close()
-	return resp.StatusCode == 200
-}
-
-// reverseString helper
-func reverseString(s string) string {
-	r := []rune(s)
-	for i, j := 0, len(r)-1; i < j; i, j = i+1, j-1 {
-		r[i], r[j] = r[j], r[i]
-	}
-	return string(r)
+	log(fmt.Sprintf("[mpv] started (pid %d)", cmd.Process.Pid))
+	return nil
 }