X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/71050031bb901b19a829cc44763bf68e67b1f648..99b46abf79266c2bf5eabc291dc4f366e61b5dde:/youtube_dl/extractor/teamcoco.py

diff --git a/youtube_dl/extractor/teamcoco.py b/youtube_dl/extractor/teamcoco.py
index bc48620..79a7789 100644
--- a/youtube_dl/extractor/teamcoco.py
+++ b/youtube_dl/extractor/teamcoco.py
@@ -1,71 +1,170 @@
+# -*- coding: utf-8 -*-
+from __future__ import unicode_literals
+
+import base64
+import binascii
 import re
-import xml.etree.ElementTree
+import json
 
 from .common import InfoExtractor
 from ..utils import (
     ExtractorError,
+    qualities,
+    determine_ext,
 )
+from ..compat import compat_ord
 
 
 class TeamcocoIE(InfoExtractor):
-    _VALID_URL = r'http://teamcoco\.com/video/(?P<url_title>.*)'
-    _TEST = {
-        u'url': u'http://teamcoco.com/video/louis-ck-interview-george-w-bush',
-        u'file': u'19705.mp4',
-        u'md5': u'cde9ba0fa3506f5f017ce11ead928f9a',
-        u'info_dict': {
-            u"description": u"Louis C.K. got starstruck by George W. Bush, so what? Part one.", 
-            u"title": u"Louis C.K. Interview Pt. 1 11/3/11"
+    _VALID_URL = r'https?://teamcoco\.com/video/(?P<video_id>[0-9]+)?/?(?P<display_id>.*)'
+    _TESTS = [
+        {
+            'url': 'http://teamcoco.com/video/80187/conan-becomes-a-mary-kay-beauty-consultant',
+            'md5': '3f7746aa0dc86de18df7539903d399ea',
+            '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,
+                'age_limit': 0,
+            }
+        }, {
+            '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,
+                'age_limit': 0,
+            }
+        }, {
+            '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',
+            },
+            '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
+            }
         }
-    }
+    ]
+    _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)
-        if mobj is None:
-            raise ExtractorError(u'Invalid URL: %s' % url)
-        url_title = mobj.group('url_title')
-        webpage = self._download_webpage(url, url_title)
 
-        video_id = self._html_search_regex(r'<article class="video" data-id="(\d+?)"',
-            webpage, u'video id')
+        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')
 
-        self.report_extraction(video_id)
+        data = None
 
-        data_url = 'http://teamcoco.com/cvp/2.0/%s.xml' % video_id
-        data_xml = self._download_webpage(data_url, video_id, 'Downloading data webpage')
-        data = xml.etree.ElementTree.fromstring(data_xml.encode('utf-8'))
+        preload_codes = self._html_search_regex(
+            r'(function.+)setTimeout\(function\(\)\{playlist',
+            webpage, 'preload codes')
+        base64_fragments = re.findall(r'"([a-zA-Z0-9+/=]+)"', preload_codes)
+        base64_fragments.remove('init')
 
+        def _check_sequence(cur_fragments):
+            if not cur_fragments:
+                return
+            for i in range(len(cur_fragments)):
+                cur_sequence = (''.join(cur_fragments[i:] + cur_fragments[:i])).encode('ascii')
+                try:
+                    raw_data = base64.b64decode(cur_sequence)
+                    if compat_ord(raw_data[0]) == compat_ord('{'):
+                        return json.loads(raw_data.decode('utf-8'))
+                except (TypeError, binascii.Error, UnicodeDecodeError, ValueError):
+                    continue
+
+        def _check_data():
+            for i in range(len(base64_fragments) + 1):
+                for j in range(i, len(base64_fragments) + 1):
+                    data = _check_sequence(base64_fragments[:i] + base64_fragments[j:])
+                    if data:
+                        return data
+
+        self.to_screen('Try to compute possible data sequence. This may take some time.')
+        data = _check_data()
+
+        if not data:
+            raise ExtractorError(
+                'Preload information could not be extracted', expected=True)
 
-        qualities = ['500k', '480p', '1000k', '720p', '1080p']
         formats = []
-        for file in data.findall('files/file'):
-            if file.attrib.get('playmode') == 'all':
-                # it just duplicates one of the entries
-                break
-            file_url = file.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(['500k', '480p', '1000k', '720p', '1080p'])
+        for filed in data['files']:
+            if determine_ext(filed['url']) == 'm3u8':
+                # 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
             else:
-                format_id = file.attrib['bitrate']
-            formats.append({
-                'url': file_url,
-                'ext': 'mp4',
-                'format_id': format_id,
-            })
-        def sort_key(f):
-            try:
-                return qualities.index(f['format_id'])
-            except ValueError:
-                return -1
-        formats.sort(key=sort_key)
-        if not formats:
-            raise RegexNotFoundError(u'Unable to extract video URL')
+                if filed['url'].startswith('/mp4:protected/'):
+                    # TODO Correct extraction for these files
+                    continue
+                m_format = re.search(r'(\d+(k|p))\.mp4', filed['url'])
+                if m_format is not 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'],
+                    'ext': 'mp4',
+                    'tbr': tbr,
+                    'format_id': format_id,
+                    'quality': get_quality(format_id),
+                })
+
+        self._sort_formats(formats)
 
         return {
-            'id':          video_id,
+            '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': data['title'],
+            'thumbnail': data.get('thumb', {}).get('href'),
+            'description': data.get('teaser'),
+            'duration': data.get('duration'),
+            'age_limit': self._family_friendly_search(webpage),
         }