vtv_api v2 — multi-source auto-failover

vtv_api.py

@@ -1,6 +1,11 @@
 """
-VTV Channels API - Backend endpoints for VTV1-VTV10 + VTVPrime
-Fetches stream URLs from hd.xemtv.net PHP endpoints and VTVPrime
+VTV Channels API v2 — Multi-source with auto-failover
+Sources per channel (priority order):
+  1. xemtv.net scraping (fresh, changes dynamically)
+  2. VTVGo failover CDN (vtvgolive-failover.vtvdigital.vn)
+  3. fptplay CDN (live247 / fnxch2 / epzhd1)
+  4. fptplay ABR (.smil/chunklist)
+Every source is verified with HTTP HEAD before returning.
 """
 import re, time, threading
 import requests
@@ -15,69 +20,81 @@ UA = {
     "Referer": "https://hd.xemtv.net/",
 }
 
-# Channel ID -> xemtv.net PHP endpoint mapping
-XEMTV_PHP_ENDPOINTS = {
-    "vtv1":  "https://hd.xemtv.net/kenh/vtv1.php",
-    "vtv2":  "https://hd.xemtv.net/kenh/vtv2.php",
-    "vtv3":  "https://hd.xemtv.net/kenh/vtv3.php",
-    "vtv4":  "https://hd.xemtv.net/kenh/vtv4.php",
-    "vtv5":  "https://hd.xemtv.net/kenh/vtv5.php",
-    "vtv6":  "https://hd.xemtv.net/kenh/vtv6.php",
-    "vtv7":  "https://hd.xemtv.net/kenh/vtv7.php",
-    "vtv8":  "https://hd.xemtv.net/kenh/vtv8.php",
-    "vtv9":  "https://hd.xemtv.net/kenh/vtv9.php",
-    "vtv10": "https://hd.xemtv.net/kenh/vtv10.php",
-    "vtvprime": "https://hd.xemtv.net/kenh/vtvprime.php",
-}
-
-# Channel display names
 CHANNEL_NAMES = {
-    "vtv1":  "VTV1",
-    "vtv2":  "VTV2",
-    "vtv3":  "VTV3",
-    "vtv4":  "VTV4",
-    "vtv5":  "VTV5",
-    "vtv6":  "VTV6",
-    "vtv7":  "VTV7",
-    "vtv8":  "VTV8",
-    "vtv9":  "VTV9",
-    "vtv10": "VTV10",
-    "vtvprime": "VTVPrime",
+    "vtv1": "VTV1", "vtv2": "VTV2", "vtv3": "VTV3", "vtv4": "VTV4",
+    "vtv5": "VTV5", "vtv6": "VTV6", "vtv7": "VTV7", "vtv8": "VTV8",
+    "vtv9": "VTV9", "vtv10": "VTV10", "vtvprime": "VTVPrime",
 }
 
-# ===== PRIMARY: VTVGo failover CDN (verified working 2026) =====
-VTVGO_FAILOVER = {
-    "vtv1":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv1-manifest.m3u8",
-    "vtv2":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv2-manifest.m3u8",
-    "vtv3":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv3-manifest.m3u8",
-    "vtv4":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv4-manifest.m3u8",
-    "vtv5":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv5-manifest.m3u8",
-    "vtv7":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv7-manifest.m3u8",
-    "vtv8":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv8-manifest.m3u8",
-    "vtv9":  "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv9-manifest.m3u8",
+# ===== SOURCE LISTS — multiple backups per channel =====
+# Each channel has a list of URLs to try in order
+CHANNEL_SOURCES = {
+    "vtv1": [
+        "https://live-a.fptplay53.net/live/media/vtv1/live247-hls-avc/index.m3u8",
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv1-manifest.m3u8",
+        "https://live.fptplay53.net/fnxch2/vtv1hd_abr.smil/chunklist.m3u8",
+    ],
+    "vtv2": [
+        "https://live-a.fptplay53.net/live/media/vtv2/live247-hls-avc/index.m3u8",
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv2-manifest.m3u8",
+        "https://live.fptplay53.net/fnxch2/vtv2hd_abr.smil/chunklist.m3u8",
+    ],
+    "vtv3": [
+        "https://live-a.fptplay53.net/live/media/vtv3/live247-hls-avc/index.m3u8",
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv3-manifest.m3u8",
+        "https://live.fptplay53.net/fnxch2/vtv3hd_abr.smil/chunklist.m3u8",
+    ],
+    "vtv4": [
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv4-manifest.m3u8",
+        "https://live.fptplay53.net/fnxch2/vtv4hd_abr.smil/chunklist.m3u8",
+    ],
+    "vtv5": [
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv5-manifest.m3u8",
+        "https://live-a.fptplay53.net/live/media/VTV5HD/live_hls_avc/index.m3u8",
+    ],
+    "vtv6": [
+        "https://live-a.fptplay53.net/live/media/vtv6/live247-hls-avc/vtv6-avc1_5600000=10000-mp4a_131600=20000.m3u8",
+        "https://live-a.fptplay53.net/live/media/vtv6/live247-hls-avc/index.m3u8",
+    ],
+    "vtv7": [
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv7-manifest.m3u8",
+        "https://live.fptplay53.net/fnxhd1/vtv7hd_vhls.smil/chunklist_b5000000.m3u8",
+    ],
+    "vtv8": [
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv8-manifest.m3u8",
+        "https://live.fptplay53.net/epzhd1/vtv8hd_vhls.smil/chunklist.m3u8",
+    ],
+    "vtv9": [
+        "https://vtvgolive-failover.vtvdigital.vn/vtvgo/vtv9-manifest.m3u8",
+        "https://live.fptplay53.net/fnxhd1/vtv9hd_vhls.smil/chunklist.m3u8",
+    ],
+    "vtv10": [
+        "https://live-a.fptplay53.net/live/media/vtv10/live247-hls-avc/vtv10-avc1_5600000=10000-mp4a_131600=20000.m3u8",
+        "https://live-a.fptplay53.net/live/media/vtv10/live247-hls-avc/index.m3u8",
+    ],
+    "vtvprime": [
+        "https://live.fptplay53.net/fnxch2/vtvprime_abr.smil/chunklist.m3u8",
+    ],
 }
 
-# ===== Channels that should use xemtv.net scraping (not VTVGo failover) =====
-# VTV6, VTV10: xemtv.net scraping (returns fptplay CDN URLs that need proxying)
-XEMTV_ONLY_CHANNELS = {"vtv6", "vtv10"}
-
-# ===== LAST RESORT: fptplay CDN URLs (need proxy for referer) =====
-FPTPLAY_URLS = {
-    "vtv1":  "https://live.fptplay53.net/fnxch2/vtv1hd_abr.smil/chunklist.m3u8",
-    "vtv2":  "https://live.fptplay53.net/fnxch2/vtv2hd_abr.smil/chunklist.m3u8",
-    "vtv3":  "https://live.fptplay53.net/fnxch2/vtv3hd_abr.smil/chunklist.m3u8",
-    "vtv4":  "https://live.fptplay53.net/fnxch2/vtv4hd_abr.smil/chunklist.m3u8",
-    "vtv5":  "https://live-a.fptplay53.net/live/media/VTV5HD/live_hls_avc/index.m3u8",
-    "vtv6":  "https://live-a.fptplay53.net/live/media/vtv6/live247-hls-avc/vtv6-avc1_5600000=10000-mp4a_131600=20000.m3u8",
-    "vtv7":  "https://live.fptplay53.net/fnxhd1/vtv7hd_vhls.smil/chunklist_b5000000.m3u8",
-    "vtv8":  "https://live.fptplay53.net/epzhd1/vtv8hd_vhls.smil/chunklist.m3u8",
-    "vtv9":  "https://live.fptplay53.net/fnxhd1/vtv9hd_vhls.smil/chunklist.m3u8",
-    "vtv10": "https://live-a.fptplay53.net/live/media/vtv10/live247-hls-avc/vtv10-avc1_5600000=10000-mp4a_131600=20000.m3u8",
+# xemtv.net PHP endpoints for dynamic scraping
+XEMTV_PHP = {
+    "vtv1": "https://hd.xemtv.net/kenh/vtv1.php",
+    "vtv2": "https://hd.xemtv.net/kenh/vtv2.php",
+    "vtv3": "https://hd.xemtv.net/kenh/vtv3.php",
+    "vtv4": "https://hd.xemtv.net/kenh/vtv4.php",
+    "vtv5": "https://hd.xemtv.net/kenh/vtv5.php",
+    "vtv6": "https://hd.xemtv.net/kenh/vtv6.php",
+    "vtv7": "https://hd.xemtv.net/kenh/vtv7.php",
+    "vtv8": "https://hd.xemtv.net/kenh/vtv8.php",
+    "vtv9": "https://hd.xemtv.net/kenh/vtv9.php",
+    "vtv10": "https://hd.xemtv.net/kenh/vtv10.php",
+    "vtvprime": "https://hd.xemtv.net/kenh/vtvprime.php",
 }
 
 _vtv_cache = {}
 _vtv_lock = threading.Lock()
-_CACHE_TTL = 180
+_CACHE_TTL = 120  # 2 min cache
 
 
 def _cached(key):
@@ -92,16 +109,42 @@ def _set_cache(key, data):
         _vtv_cache[key] = {'t': time.time(), 'd': data}
 
 
+def _verify_url(url, timeout=5):
+    """Verify a stream URL is reachable with HEAD request."""
+    try:
+        headers = {"User-Agent": UA["User-Agent"]}
+        if "fptplay" in url:
+            headers["Referer"] = "https://fptplay.vn/"
+            headers["Origin"] = "https://fptplay.vn"
+        r = requests.head(url, headers=headers, timeout=timeout, allow_redirects=True)
+        if r.status_code == 200:
+            return True
+        # Some servers don't support HEAD, try GET with range
+        headers["Range"] = "bytes=0-0"
+        r = requests.get(url, headers=headers, timeout=timeout, allow_redirects=True)
+        return r.status_code in (200, 206)
+    except:
+        return False
+
+
 def extract_m3u8_from_html(html):
-    """Extract m3u8 URL from xemtv PHP page (jwplayer config)."""
+    """Extract m3u8/smil URL from xemtv PHP page."""
     if not html:
         return None
-    m = re.search(r"file\s*:\s*['\"]([^'\"]*\.m3u8[^'\"]*)['\"]", html, re.IGNORECASE)
+    # jwplayer file: 'URL'
+    m = re.search(r"file\s*:\s*['\"]([^'\"]+\.(?:m3u8|smil)[^'\"]*)['\"]", html, re.IGNORECASE)
     if m:
         url = m.group(1).strip()
         if len(url) > 20:
             return url
+    # Generic m3u8 URL
     m = re.search(r"(https?://[^\s\"'<>\\]+\.m3u8[^\s\"'<>\\]*)", html, re.IGNORECASE)
+    if m:
+        url = m.group(1).strip()
+        if len(url) > 20:
+            return url
+    # Generic smil URL
+    m = re.search(r"(https?://[^\s\"'<>\\]+\.smil[^\s\"'<>\\]*)", html, re.IGNORECASE)
     if m:
         url = m.group(1).strip()
         if len(url) > 20:
@@ -111,34 +154,15 @@ def extract_m3u8_from_html(html):
 
 def fetch_xemtv_stream(channel_id):
     """Scrape xemtv.net PHP page for fresh stream URL."""
-    php_url = XEMTV_PHP_ENDPOINTS.get(channel_id)
+    php_url = XEMTV_PHP.get(channel_id)
     if not php_url:
         return None
     try:
-        r = requests.get(php_url, headers=UA, timeout=15, allow_redirects=True)
+        r = requests.get(php_url, headers=UA, timeout=10, allow_redirects=True)
         if r.status_code == 200:
-            m3u8 = extract_m3u8_from_html(r.text)
-            if m3u8:
-                return m3u8
-    except:
-        pass
-    return None
-
-
-def fetch_fptplay_stream(channel_id):
-    """Fetch fptplay CDN stream URL with proper Referer header."""
-    url = FPTPLAY_URLS.get(channel_id)
-    if not url:
-        return None
-    try:
-        headers = {
-            "User-Agent": UA["User-Agent"],
-            "Referer": "https://fptplay.vn/",
-            "Origin": "https://fptplay.vn",
-        }
-        r = requests.get(url, headers=headers, timeout=15, allow_redirects=True)
-        if r.status_code == 200:
-            return url
+            url = extract_m3u8_from_html(r.text)
+            if url and _verify_url(url):
+                return url
     except:
         pass
     return None
@@ -146,19 +170,13 @@ def fetch_fptplay_stream(channel_id):
 
 def fetch_vtv_stream(channel_id):
     """
-    Fetch m3u8 stream URL for a VTV channel.
-    Priority:
-      1. VTV1-VTV5, VTV7-VTV9: VTVGo failover CDN (no referer needed)
-      2. VTV6, VTV10: xemtv.net scraping (user requirement) -> fptplay URL (proxied)
-      3. VTVPrime: try xemtv, else None (paid channel)
+    Fetch m3u8 stream URL with multi-source failover.
+    Returns (url, num_sources_tried) or (None, 0).
     """
     channel_id = channel_id.lower().strip()
     name_map = {
         'vtvct': 'vtv10', 'vtv-can-tho': 'vtv10', 'vtv can tho': 'vtv10',
         'vtv_can_tho': 'vtv10', 'cantho': 'vtv10',
-        'vietnam_vtv1': 'vtv1', 'vietnam_vtv2': 'vtv2', 'vietnam_vtv3': 'vtv3',
-        'vietnam_vtv4': 'vtv4', 'vietnam_vtv5': 'vtv5', 'vietnam_vtv6': 'vtv6',
-        'vietnam_vtv7': 'vtv7', 'vietnam_vtv8': 'vtv8', 'vietnam_vtv9': 'vtv9',
     }
     channel_id = name_map.get(channel_id, channel_id)
 
@@ -166,69 +184,42 @@ def fetch_vtv_stream(channel_id):
     if cached is not None:
         return cached
 
-    # --- VTV6: xemtv.net scraping (returns working fptplay URL) ---
-    # --- VTV10: xemtv returns vtvcantho (403), skip to fptplay CDN directly ---
-    if channel_id in XEMTV_ONLY_CHANNELS:
-        if channel_id == "vtv10":
-            # VTV10: xemtv.net returns dead vtvcantho URL (403), use fptplay CDN directly
-            fpt_url = FPTPLAY_URLS.get(channel_id)
-            if fpt_url:
-                _set_cache(channel_id, fpt_url)
-                return fpt_url
-            _set_cache(channel_id, None)
-            return None
-        xemtv_url = fetch_xemtv_stream(channel_id)
-        if xemtv_url:
-            _set_cache(channel_id, xemtv_url)
-            return xemtv_url
-        fpt_url = FPTPLAY_URLS.get(channel_id)
-        if fpt_url:
-            _set_cache(channel_id, fpt_url)
-            return fpt_url
-        _set_cache(channel_id, None)
-        return None
-
-    # --- VTV1-VTV5, VTV7-VTV9: VTVGo failover ---
-    if channel_id in VTVGO_FAILOVER:
-        result = VTVGO_FAILOVER[channel_id]
-        _set_cache(channel_id, result)
-        return result
-
-    # --- VTVPrime ---
-    if channel_id == 'vtvprime':
-        xemtv_url = fetch_xemtv_stream('vtvprime')
-        if xemtv_url:
-            _set_cache(channel_id, xemtv_url)
-            return xemtv_url
-        _set_cache(channel_id, None)
-        return None
+    sources_tried = []
 
-    # --- Fallback: xemtv scraping ---
+    # 1. Try xemtv.net scraping first (fresh, dynamic)
     xemtv_url = fetch_xemtv_stream(channel_id)
     if xemtv_url:
-        _set_cache(channel_id, xemtv_url)
-        return xemtv_url
+        result = {"url": xemtv_url, "source": "xemtv", "all_sources": [xemtv_url]}
+        _set_cache(channel_id, result)
+        return result
 
-    # --- Last resort: fptplay ---
-    fpt_url = fetch_fptplay_stream(channel_id)
-    if fpt_url:
-        _set_cache(channel_id, fpt_url)
-        return fpt_url
+    # 2. Try all static sources in order
+    static_sources = CHANNEL_SOURCES.get(channel_id, [])
+    for url in static_sources:
+        sources_tried.append(url)
+        if _verify_url(url):
+            result = {"url": url, "source": "static", "all_sources": static_sources}
+            _set_cache(channel_id, result)
+            return result
 
-    _set_cache(channel_id, None)
-    return None
+    # 3. VTVPrime special: no free sources
+    result = {"url": None, "source": "none", "all_sources": []}
+    _set_cache(channel_id, result)
+    return result
 
 
 @router.get("/api/vtv/streams")
 def api_vtv_streams():
-    """Get all VTV channel stream URLs."""
+    """Get all VTV channel stream URLs with multi-source info."""
     result = {}
     for ch_id in CHANNEL_NAMES:
-        stream_url = fetch_vtv_stream(ch_id)
+        stream = fetch_vtv_stream(ch_id)
         result[ch_id] = {
             "name": CHANNEL_NAMES[ch_id],
-            "stream_url": stream_url,
-            "status": "ok" if stream_url else "offline",
+            "stream_url": stream.get("url") if stream else None,
+            "source": stream.get("source", "none") if stream else "none",
+            "all_sources": stream.get("all_sources", []) if stream else [],
+            "status": "ok" if stream and stream.get("url") else "offline",
         }
     return JSONResponse(result)
 
@@ -236,48 +227,34 @@ def api_vtv_streams():
 @router.get("/api/vtv/stream/{channel_id}")
 def api_vtv_stream(channel_id: str):
     """Get stream URL for a specific VTV channel."""
-    stream_url = fetch_vtv_stream(channel_id)
-    if stream_url:
-        return JSONResponse({"stream_url": stream_url, "status": "ok"})
+    stream = fetch_vtv_stream(channel_id)
+    if stream and stream.get("url"):
+        return JSONResponse({"stream_url": stream["url"], "source": stream["source"], "status": "ok"})
     return JSONResponse({"error": "stream not found", "status": "offline"}, status_code=404)
 
 
-@router.get("/api/proxy/page")
-def proxy_page(url: str = Query(...)):
-    """Proxy a web page."""
-    try:
-        headers = {**UA}
-        if "xemtv.net" in url:
-            headers["Referer"] = "https://hd.xemtv.net/"
-        r = requests.get(url, headers=headers, timeout=15, allow_redirects=True)
-        if r.status_code != 200:
-            return Response(status_code=502, content="upstream error")
-        return Response(
-            content=r.text.encode("utf-8"),
-            media_type="text/html; charset=utf-8",
-            headers={"Access-Control-Allow-Origin": "*"}
-        )
-    except:
-        return Response(status_code=502, content="proxy error")
-
-
 @router.get("/api/proxy/m3u8/vtv")
 def proxy_vtv_m3u8(url: str = Query(...)):
-    """Proxy an m3u8 stream URL with proper headers for fptplay CDN."""
+    """Proxy an m3u8 stream URL with proper headers."""
     try:
         headers = {"User-Agent": UA["User-Agent"], "Accept": "*/*"}
         if "fptplay" in url:
             headers["Referer"] = "https://fptplay.vn/"
             headers["Origin"] = "https://fptplay.vn"
+        elif "vtvgolive" in url or "vtvdigital" in url:
+            headers["Referer"] = "https://vtvgo.vn/"
         elif "xemtv" in url:
             headers["Referer"] = "https://hd.xemtv.net/"
+
         r = requests.get(url, headers=headers, timeout=15, allow_redirects=True)
         if r.status_code != 200:
-            return Response(status_code=502, content="upstream error")
+            return Response(status_code=502, content=f"upstream error: {r.status_code}")
+
         content = r.text
         lines = content.split('\n')
         rewritten = []
         base_url = url.rsplit('/', 1)[0] + '/'
+
         for line in lines:
             line = line.strip()
             if not line or line.startswith('#'):
@@ -287,6 +264,7 @@ def proxy_vtv_m3u8(url: str = Query(...)):
                 if not seg_url.startswith('http'):
                     seg_url = base_url + seg_url
                 rewritten.append("/api/proxy/seg/vtv?url=" + requests.utils.quote(seg_url, safe=""))
+
         return Response(
             content='\n'.join(rewritten).encode('utf-8'),
             media_type="application/vnd.apple.mpegurl",
@@ -298,18 +276,24 @@ def proxy_vtv_m3u8(url: str = Query(...)):
 
 @router.get("/api/proxy/seg/vtv")
 def proxy_vtv_segment(url: str = Query(...)):
-    """Proxy a video segment with proper headers for fptplay CDN."""
+    """Proxy a video segment with proper headers."""
     try:
         headers = {"User-Agent": UA["User-Agent"], "Accept": "*/*"}
         if "fptplay" in url:
             headers["Referer"] = "https://fptplay.vn/"
             headers["Origin"] = "https://fptplay.vn"
-        r = requests.get(url, headers=headers, timeout=30, allow_redirects=True)
+        elif "vtvgolive" in url or "vtvdigital" in url:
+            headers["Referer"] = "https://vtvgo.vn/"
+
+        r = requests.get(url, headers=headers, timeout=30, allow_redirects=True, stream=True)
         if r.status_code != 200:
-            return Response(status_code=502, content="upstream error")
+            return Response(status_code=502, content=f"upstream error: {r.status_code}")
+
         data = r.content
+        # Strip PNG header if present (some CDNs prepend it)
         if len(data) > 188 and data[0:4] == b'\x89PNG' and data[188] == 0x47:
             data = data[188:]
+
         return Response(
             content=data,
             media_type="video/mp2t",