[rtve] improve extraction
authorRemita Amine <remitamine@gmail.com>
Sun, 14 Mar 2021 13:49:23 +0000 (14:49 +0100)
committerRemita Amine <remitamine@gmail.com>
Sun, 14 Mar 2021 14:05:25 +0000 (15:05 +0100)
- extract all formats
- fix RTVE Infantil extraction(closes #24851)
- extract is_live and series

youtube_dl/extractor/rtve.py

index ce9db0629459d23ea0cc870a1bde348d21e99e9a..d2fb754cf57e09b2e3c0957da1e59078b456fc57 100644 (file)
@@ -2,8 +2,9 @@
 from __future__ import unicode_literals
 
 import base64
+import io
 import re
-import time
+import sys
 
 from .common import InfoExtractor
 from ..compat import (
@@ -14,56 +15,13 @@ from ..utils import (
     determine_ext,
     ExtractorError,
     float_or_none,
+    qualities,
     remove_end,
     remove_start,
-    sanitized_Request,
     std_headers,
 )
 
-
-def _decrypt_url(png):
-    encrypted_data = compat_b64decode(png)
-    text_index = encrypted_data.find(b'tEXt')
-    text_chunk = encrypted_data[text_index - 4:]
-    length = compat_struct_unpack('!I', text_chunk[:4])[0]
-    # Use bytearray to get integers when iterating in both python 2.x and 3.x
-    data = bytearray(text_chunk[8:8 + length])
-    data = [chr(b) for b in data if b != 0]
-    hash_index = data.index('#')
-    alphabet_data = data[:hash_index]
-    url_data = data[hash_index + 1:]
-    if url_data[0] == 'H' and url_data[3] == '%':
-        # remove useless HQ%% at the start
-        url_data = url_data[4:]
-
-    alphabet = []
-    e = 0
-    d = 0
-    for l in alphabet_data:
-        if d == 0:
-            alphabet.append(l)
-            d = e = (e + 1) % 4
-        else:
-            d -= 1
-    url = ''
-    f = 0
-    e = 3
-    b = 1
-    for letter in url_data:
-        if f == 0:
-            l = int(letter) * 10
-            f = 1
-        else:
-            if e == 0:
-                l += int(letter)
-                url += alphabet[l]
-                e = (b + 3) % 4
-                f = 0
-                b += 1
-            else:
-                e -= 1
-
-    return url
+_bytes_to_chr = (lambda x: x) if sys.version_info[0] == 2 else (lambda x: map(chr, x))
 
 
 class RTVEALaCartaIE(InfoExtractor):
@@ -79,28 +37,31 @@ class RTVEALaCartaIE(InfoExtractor):
             'ext': 'mp4',
             'title': 'Balonmano - Swiss Cup masculina. Final: España-Suecia',
             'duration': 5024.566,
+            'series': 'Balonmano',
         },
+        'expected_warnings': ['Failed to download MPD manifest', 'Failed to download m3u8 information'],
     }, {
         'note': 'Live stream',
         'url': 'http://www.rtve.es/alacarta/videos/television/24h-live/1694255/',
         'info_dict': {
             'id': '1694255',
-            'ext': 'flv',
-            'title': 'TODO',
+            'ext': 'mp4',
+            'title': 're:^24H LIVE [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+            'is_live': True,
+        },
+        'params': {
+            'skip_download': 'live stream',
         },
-        'skip': 'The f4m manifest can\'t be used yet',
     }, {
         'url': 'http://www.rtve.es/alacarta/videos/servir-y-proteger/servir-proteger-capitulo-104/4236788/',
-        'md5': 'e55e162379ad587e9640eda4f7353c0f',
+        'md5': 'd850f3c8731ea53952ebab489cf81cbf',
         'info_dict': {
             'id': '4236788',
             'ext': 'mp4',
-            'title': 'Servir y proteger - Capítulo 104 ',
+            'title': 'Servir y proteger - Capítulo 104',
             'duration': 3222.0,
         },
-        'params': {
-            'skip_download': True,  # requires ffmpeg
-        },
+        'expected_warnings': ['Failed to download MPD manifest', 'Failed to download m3u8 information'],
     }, {
         'url': 'http://www.rtve.es/m/alacarta/videos/cuentame-como-paso/cuentame-como-paso-t16-ultimo-minuto-nuestra-vida-capitulo-276/2969138/?media=tve',
         'only_matching': True,
@@ -111,58 +72,102 @@ class RTVEALaCartaIE(InfoExtractor):
 
     def _real_initialize(self):
         user_agent_b64 = base64.b64encode(std_headers['User-Agent'].encode('utf-8')).decode('utf-8')
-        manager_info = self._download_json(
+        self._manager = self._download_json(
             'http://www.rtve.es/odin/loki/' + user_agent_b64,
-            None, 'Fetching manager info')
-        self._manager = manager_info['manager']
+            None, 'Fetching manager info')['manager']
+
+    @staticmethod
+    def _decrypt_url(png):
+        encrypted_data = io.BytesIO(compat_b64decode(png)[8:])
+        while True:
+            length = compat_struct_unpack('!I', encrypted_data.read(4))[0]
+            chunk_type = encrypted_data.read(4)
+            if chunk_type == b'IEND':
+                break
+            data = encrypted_data.read(length)
+            if chunk_type == b'tEXt':
+                alphabet_data, text = data.split(b'\0')
+                quality, url_data = text.split(b'%%')
+                alphabet = []
+                e = 0
+                d = 0
+                for l in _bytes_to_chr(alphabet_data):
+                    if d == 0:
+                        alphabet.append(l)
+                        d = e = (e + 1) % 4
+                    else:
+                        d -= 1
+                url = ''
+                f = 0
+                e = 3
+                b = 1
+                for letter in _bytes_to_chr(url_data):
+                    if f == 0:
+                        l = int(letter) * 10
+                        f = 1
+                    else:
+                        if e == 0:
+                            l += int(letter)
+                            url += alphabet[l]
+                            e = (b + 3) % 4
+                            f = 0
+                            b += 1
+                        else:
+                            e -= 1
+
+                yield quality.decode(), url
+            encrypted_data.read(4)  # CRC
+
+    def _extract_png_formats(self, video_id):
+        png = self._download_webpage(
+            'http://www.rtve.es/ztnr/movil/thumbnail/%s/videos/%s.png' % (self._manager, video_id),
+            video_id, 'Downloading url information', query={'q': 'v2'})
+        q = qualities(['Media', 'Alta', 'HQ', 'HD_READY', 'HD_FULL'])
+        formats = []
+        for quality, video_url in self._decrypt_url(png):
+            ext = determine_ext(video_url)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    video_url, video_id, 'mp4', 'm3u8_native',
+                    m3u8_id='hls', fatal=False))
+            elif ext == 'mpd':
+                formats.extend(self._extract_mpd_formats(
+                    video_url, video_id, 'dash', fatal=False))
+            else:
+                formats.append({
+                    'format_id': quality,
+                    'quality': q(quality),
+                    'url': video_url,
+                })
+        self._sort_formats(formats)
+        return formats
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
+        video_id = self._match_id(url)
         info = self._download_json(
             'http://www.rtve.es/api/videos/%s/config/alacarta_videos.json' % video_id,
             video_id)['page']['items'][0]
         if info['state'] == 'DESPU':
             raise ExtractorError('The video is no longer available', expected=True)
-        title = info['title']
-        png_url = 'http://www.rtve.es/ztnr/movil/thumbnail/%s/videos/%s.png' % (self._manager, video_id)
-        png_request = sanitized_Request(png_url)
-        png_request.add_header('Referer', url)
-        png = self._download_webpage(png_request, video_id, 'Downloading url information')
-        video_url = _decrypt_url(png)
-        ext = determine_ext(video_url)
-
-        formats = []
-        if not video_url.endswith('.f4m') and ext != 'm3u8':
-            if '?' not in video_url:
-                video_url = video_url.replace('resources/', 'auth/resources/')
-            video_url = video_url.replace('.net.rtve', '.multimedia.cdn.rtve')
-
-        if ext == 'm3u8':
-            formats.extend(self._extract_m3u8_formats(
-                video_url, video_id, ext='mp4', entry_protocol='m3u8_native',
-                m3u8_id='hls', fatal=False))
-        elif ext == 'f4m':
-            formats.extend(self._extract_f4m_formats(
-                video_url, video_id, f4m_id='hds', fatal=False))
-        else:
-            formats.append({
-                'url': video_url,
-            })
-        self._sort_formats(formats)
+        title = info['title'].strip()
+        formats = self._extract_png_formats(video_id)
 
         subtitles = None
-        if info.get('sbtFile') is not None:
-            subtitles = self.extract_subtitles(video_id, info['sbtFile'])
+        sbt_file = info.get('sbtFile')
+        if sbt_file:
+            subtitles = self.extract_subtitles(video_id, sbt_file)
+
+        is_live = info.get('live') is True
 
         return {
             'id': video_id,
-            'title': title,
+            'title': self._live_title(title) if is_live else title,
             'formats': formats,
             'thumbnail': info.get('image'),
-            'page_url': url,
             'subtitles': subtitles,
-            'duration': float_or_none(info.get('duration'), scale=1000),
+            'duration': float_or_none(info.get('duration'), 1000),
+            'is_live': is_live,
+            'series': info.get('programTitle'),
         }
 
     def _get_subtitles(self, video_id, sub_file):
@@ -174,48 +179,26 @@ class RTVEALaCartaIE(InfoExtractor):
             for s in subs)
 
 
-class RTVEInfantilIE(InfoExtractor):
+class RTVEInfantilIE(RTVEALaCartaIE):
     IE_NAME = 'rtve.es:infantil'
     IE_DESC = 'RTVE infantil'
-    _VALID_URL = r'https?://(?:www\.)?rtve\.es/infantil/serie/(?P<show>[^/]*)/video/(?P<short_title>[^/]*)/(?P<id>[0-9]+)/'
+    _VALID_URL = r'https?://(?:www\.)?rtve\.es/infantil/serie/[^/]+/video/[^/]+/(?P<id>[0-9]+)/'
 
     _TESTS = [{
         'url': 'http://www.rtve.es/infantil/serie/cleo/video/maneras-vivir/3040283/',
-        'md5': '915319587b33720b8e0357caaa6617e6',
+        'md5': '5747454717aedf9f9fdf212d1bcfc48d',
         'info_dict': {
             'id': '3040283',
             'ext': 'mp4',
             'title': 'Maneras de vivir',
-            'thumbnail': 'http://www.rtve.es/resources/jpg/6/5/1426182947956.JPG',
+            'thumbnail': r're:https?://.+/1426182947956\.JPG',
             'duration': 357.958,
         },
+        'expected_warnings': ['Failed to download MPD manifest', 'Failed to download m3u8 information'],
     }]
 
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-        info = self._download_json(
-            'http://www.rtve.es/api/videos/%s/config/alacarta_videos.json' % video_id,
-            video_id)['page']['items'][0]
-
-        webpage = self._download_webpage(url, video_id)
-        vidplayer_id = self._search_regex(
-            r' id="vidplayer([0-9]+)"', webpage, 'internal video ID')
 
-        png_url = 'http://www.rtve.es/ztnr/movil/thumbnail/default/videos/%s.png' % vidplayer_id
-        png = self._download_webpage(png_url, video_id, 'Downloading url information')
-        video_url = _decrypt_url(png)
-
-        return {
-            'id': video_id,
-            'ext': 'mp4',
-            'title': info['title'],
-            'url': video_url,
-            'thumbnail': info.get('image'),
-            'duration': float_or_none(info.get('duration'), scale=1000),
-        }
-
-
-class RTVELiveIE(InfoExtractor):
+class RTVELiveIE(RTVEALaCartaIE):
     IE_NAME = 'rtve.es:live'
     IE_DESC = 'RTVE.es live streams'
     _VALID_URL = r'https?://(?:www\.)?rtve\.es/directo/(?P<id>[a-zA-Z0-9-]+)'
@@ -225,7 +208,7 @@ class RTVELiveIE(InfoExtractor):
         'info_dict': {
             'id': 'la-1',
             'ext': 'mp4',
-            'title': 're:^La 1 [0-9]{4}-[0-9]{2}-[0-9]{2}Z[0-9]{6}$',
+            'title': 're:^La 1 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
         },
         'params': {
             'skip_download': 'live stream',
@@ -234,29 +217,22 @@ class RTVELiveIE(InfoExtractor):
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        start_time = time.gmtime()
         video_id = mobj.group('id')
 
         webpage = self._download_webpage(url, video_id)
         title = remove_end(self._og_search_title(webpage), ' en directo en RTVE.es')
         title = remove_start(title, 'Estoy viendo ')
-        title += ' ' + time.strftime('%Y-%m-%dZ%H%M%S', start_time)
 
         vidplayer_id = self._search_regex(
             (r'playerId=player([0-9]+)',
              r'class=["\'].*?\blive_mod\b.*?["\'][^>]+data-assetid=["\'](\d+)',
              r'data-id=["\'](\d+)'),
             webpage, 'internal video ID')
-        png_url = 'http://www.rtve.es/ztnr/movil/thumbnail/amonet/videos/%s.png' % vidplayer_id
-        png = self._download_webpage(png_url, video_id, 'Downloading url information')
-        m3u8_url = _decrypt_url(png)
-        formats = self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4')
-        self._sort_formats(formats)
 
         return {
             'id': video_id,
-            'title': title,
-            'formats': formats,
+            'title': self._live_title(title),
+            'formats': self._extract_png_formats(vidplayer_id),
             'is_live': True,
         }