X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fteamcoco.py;h=f06e5b19ac675a975ecc0a1f5cfca23a93a64bfa;hb=54fc90aabfb71968f28af68dfe3f7a3544cc2f0b;hp=e0fc3e60fcdf593b58216c215b87f1ff87ce909a;hpb=04ee53eca1345df36a8c4a251c22a29f797ab798;p=youtube-dl

diff --git a/youtube_dl/extractor/teamcoco.py b/youtube_dl/extractor/teamcoco.py
index e0fc3e60f..f06e5b19a 100644
--- a/youtube_dl/extractor/teamcoco.py
+++ b/youtube_dl/extractor/teamcoco.py
@@ -1,78 +1,157 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
-import re
+import json
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
     ExtractorError,
+    int_or_none,
+    mimetype2ext,
+    parse_duration,
+    parse_iso8601,
+    qualities,
 )
 
 
 class TeamcocoIE(InfoExtractor):
-    _VALID_URL = r'http://teamcoco\.com/video/(?P<video_id>\d*)?/?(?P<url_title>.*)'
-    _TEST = {
-        'url': 'http://teamcoco.com/video/louis-ck-interview-george-w-bush',
-        'file': '19705.mp4',
-        'md5': 'cde9ba0fa3506f5f017ce11ead928f9a',
-        'info_dict': {
-            "description": "Louis C.K. got starstruck by George W. Bush, so what? Part one.",
-            "title": "Louis C.K. Interview Pt. 1 11/3/11"
+    _VALID_URL = r'https?://teamcoco\.com/video/(?P<id>[^/?#]+)'
+    _TESTS = [
+        {
+            'url': 'http://teamcoco.com/video/mary-kay-remote',
+            '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': 495.0,
+                'upload_date': '20140402',
+                'timestamp': 1396407600,
+            }
+        }, {
+            'url': 'http://teamcoco.com/video/louis-ck-interview-george-w-bush',
+            'md5': 'cde9ba0fa3506f5f017ce11ead928f9a',
+            'info_dict': {
+                'id': '19705',
+                'ext': 'mp4',
+                '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,
+                'upload_date': '20111104',
+                'timestamp': 1320405840,
+            }
+        }, {
+            'url': 'http://teamcoco.com/video/timothy-olyphant-drinking-whiskey',
+            'info_dict': {
+                'id': '88748',
+                'ext': 'mp4',
+                'title': 'Timothy Olyphant Raises A Toast To “Justified”',
+                'description': 'md5:15501f23f020e793aeca761205e42c24',
+                'upload_date': '20150415',
+                'timestamp': 1429088400,
+            },
+            'params': {
+                'skip_download': True,  # m3u8 downloads
+            }
+        }, {
+            'url': 'http://teamcoco.com/video/full-episode-mon-6-1-joel-mchale-jake-tapper-and-musical-guest-courtney-barnett?playlist=x;eyJ0eXBlIjoidGFnIiwiaWQiOjl9',
+            'info_dict': {
+                'id': '89341',
+                'ext': 'mp4',
+                'title': 'Full Episode - Mon. 6/1 - Joel McHale, Jake Tapper, And Musical Guest Courtney Barnett',
+                'description': 'Guests: Joel McHale, Jake Tapper, And Musical Guest Courtney Barnett',
+            },
+            'params': {
+                'skip_download': True,  # m3u8 downloads
+            },
+            'skip': 'This video is no longer available.',
         }
-    }
+    ]
+
+    def _graphql_call(self, query_template, object_type, object_id):
+        find_object = 'find' + object_type
+        return self._download_json(
+            'http://teamcoco.com/graphql/', object_id, data=json.dumps({
+                'query': query_template % (find_object, object_id)
+            }))['data'][find_object]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        if mobj is None:
-            raise ExtractorError('Invalid URL: %s' % url)
-        url_title = mobj.group('url_title')
-        webpage = self._download_webpage(url, url_title)
-        
-        video_id = mobj.group("video_id")
-        if video_id == '':
-            video_id = self._html_search_regex(
-                r'<article class="video" data-id="(\d+?)"',
-                webpage, 'video id')
-        
-        self.report_extraction(video_id)
+        display_id = self._match_id(url)
+
+        response = self._graphql_call('''{
+  %s(slug: "video/%s") {
+    ... on RecordSlug {
+      record {
+        id
+        title
+        teaser
+        publishOn
+        thumb {
+          preview
+        }
+        tags {
+          name
+        }
+        duration
+      }
+    }
+    ... on NotFoundSlug {
+      status
+    }
+  }
+}''', 'Slug', display_id)
+        if response.get('status'):
+            raise ExtractorError('This video is no longer available.', expected=True)
 
-        data_url = 'http://teamcoco.com/cvp/2.0/%s.xml' % video_id
-        data = self._download_xml(data_url, video_id, 'Downloading data webpage')
+        record = response['record']
+        video_id = record['id']
+
+        srcs = self._graphql_call('''{
+  %s(id: "%s") {
+    src
+  }
+}''', 'RecordVideoSource', video_id)['src']
 
-        qualities = ['500k', '480p', '1000k', '720p', '1080p']
         formats = []
-        for filed in data.findall('files/file'):
-            if filed.attrib.get('playmode') == 'all':
-                # it just duplicates one of the entries
-                break
-            file_url = filed.text
-            m_format = re.search(r'(\d+(k|p))\.mp4', file_url)
-            if m_format is not None:
-                format_id = m_format.group(1)
+        get_quality = qualities(['low', 'sd', 'hd', 'uhd'])
+        for format_id, src in srcs.items():
+            if not isinstance(src, dict):
+                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:
-                format_id = filed.attrib['bitrate']
-            tbr = (
-                int(filed.attrib['bitrate'])
-                if filed.attrib['bitrate'].isdigit()
-                else None)
-
-            try:
-                quality = qualities.index(format_id)
-            except ValueError:
-                quality = -1
-            formats.append({
-                'url': file_url,
-                'ext': 'mp4',
-                'tbr': tbr,
-                'format_id': format_id,
-                'quality': quality,
-            })
+                if src_url.startswith('/mp4:protected/'):
+                    # TODO Correct extraction for these files
+                    continue
+                tbr = int_or_none(self._search_regex(
+                    r'(\d+)k\.mp4', src_url, 'tbr', default=None))
 
+                formats.append({
+                    'url': src_url,
+                    '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': self._og_search_title(webpage),
-            'thumbnail': self._og_search_thumbnail(webpage),
-            'description': self._og_search_description(webpage),
+            'title': record['title'],
+            'thumbnail': record.get('thumb', {}).get('preview'),
+            'description': record.get('teaser'),
+            'duration': parse_duration(record.get('duration')),
+            'timestamp': parse_iso8601(record.get('publishOn')),
         }