_ Git - youtube-dl/blob - youtube_dl/extractor/teamcoco.py

   1 # -*- coding: utf-8 -*-
   2 from __future__ import unicode_literals
   3
   4 import base64
   5 import binascii
   6 import re
   7 import json
   8
   9 from .common import InfoExtractor
  10 from ..utils import (
  11     ExtractorError,
  12     qualities,
  13 )
  14 from ..compat import compat_ord
  15
  16
  17 class TeamcocoIE(InfoExtractor):
  18     _VALID_URL = r'http://teamcoco\.com/video/(?P<video_id>[0-9]+)?/?(?P<display_id>.*)'
  19     _TESTS = [
  20         {
  21             'url': 'http://teamcoco.com/video/80187/conan-becomes-a-mary-kay-beauty-consultant',
  22             'md5': '3f7746aa0dc86de18df7539903d399ea',
  23             'info_dict': {
  24                 'id': '80187',
  25                 'ext': 'mp4',
  26                 'title': 'Conan Becomes A Mary Kay Beauty Consultant',
  27                 'description': 'Mary Kay is perhaps the most trusted name in female beauty, so of course Conan is a natural choice to sell their products.',
  28                 'duration': 504,
  29                 'age_limit': 0,
  30             }
  31         }, {
  32             'url': 'http://teamcoco.com/video/louis-ck-interview-george-w-bush',
  33             'md5': 'cde9ba0fa3506f5f017ce11ead928f9a',
  34             'info_dict': {
  35                 'id': '19705',
  36                 'ext': 'mp4',
  37                 'description': 'Louis C.K. got starstruck by George W. Bush, so what? Part one.',
  38                 'title': 'Louis C.K. Interview Pt. 1 11/3/11',
  39                 'duration': 288,
  40                 'age_limit': 0,
  41             }
  42         }, {
  43             'url': 'http://teamcoco.com/video/timothy-olyphant-drinking-whiskey',
  44             'info_dict': {
  45                 'id': '88748',
  46                 'ext': 'mp4',
  47                 'title': 'Timothy Olyphant Raises A Toast To “Justified”',
  48                 'description': 'md5:15501f23f020e793aeca761205e42c24',
  49             },
  50             'params': {
  51                 'skip_download': True,  # m3u8 downloads
  52             }
  53         }
  54     ]
  55     _VIDEO_ID_REGEXES = (
  56         r'"eVar42"\s*:\s*(\d+)',
  57         r'Ginger\.TeamCoco\.openInApp\("video",\s*"([^"]+)"',
  58         r'"id_not"\s*:\s*(\d+)'
  59     )
  60
  61     def _real_extract(self, url):
  62         mobj = re.match(self._VALID_URL, url)
  63
  64         display_id = mobj.group('display_id')
  65         webpage = self._download_webpage(url, display_id)
  66
  67         video_id = mobj.group('video_id')
  68         if not video_id:
  69             video_id = self._html_search_regex(
  70                 self._VIDEO_ID_REGEXES, webpage, 'video id')
  71
  72         data = None
  73
  74         preload_codes = self._html_search_regex(
  75             r'(function.+)setTimeout\(function\(\)\{playlist',
  76             webpage, 'preload codes')
  77         base64_fragments = re.findall(r'"([a-zA-z0-9+/=]+)"', preload_codes)
  78         base64_fragments.remove('init')
  79
  80         def _check_sequence(cur_fragments):
  81             if not cur_fragments:
  82                 return
  83             for i in range(len(cur_fragments)):
  84                 cur_sequence = (''.join(cur_fragments[i:] + cur_fragments[:i])).encode('ascii')
  85                 try:
  86                     raw_data = base64.b64decode(cur_sequence)
  87                     if compat_ord(raw_data[0]) == compat_ord('{'):
  88                         return json.loads(raw_data.decode('utf-8'))
  89                 except (TypeError, binascii.Error, UnicodeDecodeError, ValueError):
  90                     continue
  91
  92         def _check_data():
  93             for i in range(len(base64_fragments) + 1):
  94                 for j in range(i, len(base64_fragments) + 1):
  95                     data = _check_sequence(base64_fragments[:i] + base64_fragments[j:])
  96                     if data:
  97                         return data
  98
  99         self.to_screen('Try to compute possible data sequence. This may take some time.')
 100         data = _check_data()
 101
 102         if not data:
 103             raise ExtractorError(
 104                 'Preload information could not be extracted', expected=True)
 105
 106         formats = []
 107         get_quality = qualities(['500k', '480p', '1000k', '720p', '1080p'])
 108         for filed in data['files']:
 109             if filed['type'] == 'hls':
 110                 formats.extend(self._extract_m3u8_formats(
 111                     filed['url'], video_id, ext='mp4'))
 112             else:
 113                 m_format = re.search(r'(\d+(k|p))\.mp4', filed['url'])
 114                 if m_format is not None:
 115                     format_id = m_format.group(1)
 116                 else:
 117                     format_id = filed['bitrate']
 118                 tbr = (
 119                     int(filed['bitrate'])
 120                     if filed['bitrate'].isdigit()
 121                     else None)
 122
 123                 formats.append({
 124                     'url': filed['url'],
 125                     'ext': 'mp4',
 126                     'tbr': tbr,
 127                     'format_id': format_id,
 128                     'quality': get_quality(format_id),
 129                 })
 130
 131         self._sort_formats(formats)
 132
 133         return {
 134             'id': video_id,
 135             'display_id': display_id,
 136             'formats': formats,
 137             'title': data['title'],
 138             'thumbnail': data.get('thumb', {}).get('href'),
 139             'description': data.get('teaser'),
 140             'duration': data.get('duration'),
 141             'age_limit': self._family_friendly_search(webpage),
 142         }