[teamcoco] fix extraction(closes #16374)

youtube_dl/extractor/teamcoco.py

@@ -1,35 +1,34 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import binascii
-import re
 import json
 
 from .common import InfoExtractor
-from ..compat import (
-    compat_b64decode,
-    compat_ord,
-)
 from ..utils import (
+    determine_ext,
     ExtractorError,
+    int_or_none,
+    mimetype2ext,
+    parse_duration,
+    parse_iso8601,
     qualities,
-    determine_ext,
 )
 
 
 class TeamcocoIE(InfoExtractor):
-    _VALID_URL = r'https?://teamcoco\.com/video/(?P<video_id>[0-9]+)?/?(?P<display_id>.*)'
+    _VALID_URL = r'https?://teamcoco\.com/video/(?P<id>[^/?#]+)'
     _TESTS = [
         {
-            'url': 'http://teamcoco.com/video/80187/conan-becomes-a-mary-kay-beauty-consultant',
+            'url': 'http://teamcoco.com/video/mary-kay-remote',
-            'md5': '3f7746aa0dc86de18df7539903d399ea',
+            'md5': '55d532f81992f5c92046ad02fec34d7d',
             'info_dict': {
                 'id': '80187',
                 'ext': 'mp4',
                 'title': 'Conan Becomes A Mary Kay Beauty Consultant',
                 'description': 'Mary Kay is perhaps the most trusted name in female beauty, so of course Conan is a natural choice to sell their products.',
-                'duration': 504,
+                'duration': 495.0,
-                'age_limit': 0,
+                'upload_date': '20140402',
+                'timestamp': 1396407600,
             }
         }, {
             'url': 'http://teamcoco.com/video/louis-ck-interview-george-w-bush',
@@ -40,7 +39,8 @@ class TeamcocoIE(InfoExtractor):
                 'description': 'Louis C.K. got starstruck by George W. Bush, so what? Part one.',
                 'title': 'Louis C.K. Interview Pt. 1 11/3/11',
                 'duration': 288,
-                'age_limit': 0,
+                'upload_date': '20111104',
+                'timestamp': 1320405840,
             }
         }, {
             'url': 'http://teamcoco.com/video/timothy-olyphant-drinking-whiskey',
@@ -49,6 +49,8 @@ class TeamcocoIE(InfoExtractor):
                 'ext': 'mp4',
                 'title': 'Timothy Olyphant Raises A Toast To “Justified”',
                 'description': 'md5:15501f23f020e793aeca761205e42c24',
+                'upload_date': '20150415',
+                'timestamp': 1429088400,
             },
             'params': {
                 'skip_download': True,  # m3u8 downloads
@@ -63,110 +65,93 @@ class TeamcocoIE(InfoExtractor):
             },
             'params': {
                 'skip_download': True,  # m3u8 downloads
-            }
+            },
+            'skip': 'This video is no longer available.',
         }
     ]
-    _VIDEO_ID_REGEXES = (
-        r'"eVar42"\s*:\s*(\d+)',
-        r'Ginger\.TeamCoco\.openInApp\("video",\s*"([^"]+)"',
-        r'"id_not"\s*:\s*(\d+)'
-    )
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-
-        display_id = mobj.group('display_id')
-        webpage, urlh = self._download_webpage_handle(url, display_id)
-        if 'src=expired' in urlh.geturl():
-            raise ExtractorError('This video is expired.', expected=True)
-
-        video_id = mobj.group('video_id')
-        if not video_id:
-            video_id = self._html_search_regex(
-                self._VIDEO_ID_REGEXES, webpage, 'video id')
-
-        data = None
 
-        preload_codes = self._html_search_regex(
+    def _graphql_call(self, query_template, object_type, object_id):
-            r'(function.+)setTimeout\(function\(\)\{playlist',
+        find_object = 'find' + object_type
-            webpage, 'preload codes')
+        return self._download_json(
-        base64_fragments = re.findall(r'"([a-zA-Z0-9+/=]+)"', preload_codes)
+            'http://teamcoco.com/graphql/', object_id, data=json.dumps({
-        base64_fragments.remove('init')
+                'query': query_template % (find_object, object_id)
+            }))['data'][find_object]
 
-        def _check_sequence(cur_fragments):
+    def _real_extract(self, url):
-            if not cur_fragments:
+        display_id = self._match_id(url)
-                return
+
-            for i in range(len(cur_fragments)):
+        response = self._graphql_call('''{
-                cur_sequence = (''.join(cur_fragments[i:] + cur_fragments[:i])).encode('ascii')
+  %s(slug: "video/%s") {
-                try:
+    ... on RecordSlug {
-                    raw_data = compat_b64decode(cur_sequence)
+      record {
-                    if compat_ord(raw_data[0]) == compat_ord('{'):
+        id
-                        return json.loads(raw_data.decode('utf-8'))
+        title
-                except (TypeError, binascii.Error, UnicodeDecodeError, ValueError):
+        teaser
-                    continue
+        publishOn
-
+        thumb {
-        def _check_data():
+          preview
-            for i in range(len(base64_fragments) + 1):
+        }
-                for j in range(i, len(base64_fragments) + 1):
+        tags {
-                    data = _check_sequence(base64_fragments[:i] + base64_fragments[j:])
+          name
-                    if data:
+        }
-                        return data
+        duration
+      }
+    }
+    ... on NotFoundSlug {
+      status
+    }
+  }
+}''', 'Slug', display_id)
+        if response.get('status'):
+            raise ExtractorError('This video is no longer available.', expected=True)
 
-        self.to_screen('Try to compute possible data sequence. This may take some time.')
+        record = response['record']
-        data = _check_data()
+        video_id = record['id']
 
-        if not data:
+        srcs = self._graphql_call('''{
-            raise ExtractorError(
+  %s(id: "%s") {
-                'Preload information could not be extracted', expected=True)
+    src
+  }
+}''', 'RecordVideoSource', video_id)['src']
 
         formats = []
-        get_quality = qualities(['500k', '480p', '1000k', '720p', '1080p'])
+        get_quality = qualities(['low', 'sd', 'hd', 'uhd'])
-        for filed in data['files']:
+        for format_id, src in srcs.items():
-            if determine_ext(filed['url']) == 'm3u8':
+            if not isinstance(src, dict):
-                # compat_urllib_parse.urljoin does not work here
-                if filed['url'].startswith('/'):
-                    m3u8_url = 'http://ht.cdn.turner.com/tbs/big/teamcoco' + filed['url']
-                else:
-                    m3u8_url = filed['url']
-                m3u8_formats = self._extract_m3u8_formats(
-                    m3u8_url, video_id, ext='mp4')
-                for m3u8_format in m3u8_formats:
-                    if m3u8_format not in formats:
-                        formats.append(m3u8_format)
-            elif determine_ext(filed['url']) == 'f4m':
-                # TODO Correct f4m extraction
                 continue
+            src_url = src.get('src')
+            if not src_url:
+                continue
+            ext = determine_ext(src_url, mimetype2ext(src.get('type')))
+            if format_id == 'hls' or ext == 'm3u8':
+                # compat_urllib_parse.urljoin does not work here
+                if src_url.startswith('/'):
+                    src_url = 'http://ht.cdn.turner.com/tbs/big/teamcoco' + src_url
+                formats.extend(self._extract_m3u8_formats(
+                    src_url, video_id, 'mp4', m3u8_id=format_id, fatal=False))
             else:
-                if filed['url'].startswith('/mp4:protected/'):
+                if src_url.startswith('/mp4:protected/'):
                     # TODO Correct extraction for these files
                     continue
-                m_format = re.search(r'(\d+(k|p))\.mp4', filed['url'])
+                tbr = int_or_none(self._search_regex(
-                if m_format is not None:
+                    r'(\d+)k\.mp4', src_url, 'tbr', default=None))
-                    format_id = m_format.group(1)
-                else:
-                    format_id = filed['bitrate']
-                tbr = (
-                    int(filed['bitrate'])
-                    if filed['bitrate'].isdigit()
-                    else None)
 
                 formats.append({
-                    'url': filed['url'],
+                    'url': src_url,
-                    'ext': 'mp4',
+                    'ext': ext,
                     'tbr': tbr,
                     'format_id': format_id,
                     'quality': get_quality(format_id),
                 })
-
         self._sort_formats(formats)
 
         return {
             'id': video_id,
             'display_id': display_id,
             'formats': formats,
-            'title': data['title'],
+            'title': record['title'],
-            'thumbnail': data.get('thumb', {}).get('href'),
+            'thumbnail': record.get('thumb', {}).get('preview'),
-            'description': data.get('teaser'),
+            'description': record.get('teaser'),
-            'duration': data.get('duration'),
+            'duration': parse_duration(record.get('duration')),
-            'age_limit': self._family_friendly_search(webpage),
+            'timestamp': parse_iso8601(record.get('publishOn')),
         }