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

   1 from __future__ import unicode_literals
   2
   3 import datetime
   4 import json
   5 import re
   6 import socket
   7
   8 from .common import InfoExtractor
   9 from ..utils import (
  10     compat_http_client,
  11     compat_parse_qs,
  12     compat_str,
  13     compat_urllib_error,
  14     compat_urllib_parse_urlparse,
  15     compat_urllib_request,
  16
  17     ExtractorError,
  18     unescapeHTML,
  19 )
  20
  21
  22 class BlipTVIE(InfoExtractor):
  23     """Information extractor for blip.tv"""
  24
  25     _VALID_URL = r'^(?:https?://)?(?:www\.)?blip\.tv/((.+/)|(play/)|(api\.swf#))(.+)$'
  26     IE_NAME = 'blip.tv'
  27     _TEST = {
  28         'url': 'http://blip.tv/cbr/cbr-exclusive-gotham-city-imposters-bats-vs-jokerz-short-3-5796352',
  29         'file': '5779306.mov',
  30         'md5': 'c6934ad0b6acf2bd920720ec888eb812',
  31         'info_dict': {
  32             'upload_date': '20111205',
  33             'description': 'md5:9bc31f227219cde65e47eeec8d2dc596',
  34             'uploader': 'Comic Book Resources - CBR TV',
  35             'title': 'CBR EXCLUSIVE: "Gotham City Imposters" Bats VS Jokerz Short 3',
  36         }
  37     }
  38
  39     def report_direct_download(self, title):
  40         """Report information extraction."""
  41         self.to_screen('%s: Direct download detected' % title)
  42
  43     def _real_extract(self, url):
  44         mobj = re.match(self._VALID_URL, url)
  45         if mobj is None:
  46             raise ExtractorError('Invalid URL: %s' % url)
  47
  48         # See https://github.com/rg3/youtube-dl/issues/857
  49         api_mobj = re.match(r'http://a\.blip\.tv/api\.swf#(?P<video_id>[\d\w]+)', url)
  50         if api_mobj is not None:
  51             url = 'http://blip.tv/play/g_%s' % api_mobj.group('video_id')
  52         urlp = compat_urllib_parse_urlparse(url)
  53         if urlp.path.startswith('/play/'):
  54             response = self._request_webpage(url, None, False)
  55             redirecturl = response.geturl()
  56             rurlp = compat_urllib_parse_urlparse(redirecturl)
  57             file_id = compat_parse_qs(rurlp.fragment)['file'][0].rpartition('/')[2]
  58             url = 'http://blip.tv/a/a-' + file_id
  59             return self._real_extract(url)
  60
  61         if '?' in url:
  62             cchar = '&'
  63         else:
  64             cchar = '?'
  65         json_url = url + cchar + 'skin=json&version=2&no_wrap=1'
  66         request = compat_urllib_request.Request(json_url)
  67         request.add_header('User-Agent', 'iTunes/10.6.1')
  68         self.report_extraction(mobj.group(1))
  69         urlh = self._request_webpage(request, None, False,
  70             'unable to download video info webpage')
  71
  72         try:
  73             json_code_bytes = urlh.read()
  74             json_code = json_code_bytes.decode('utf-8')
  75         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
  76             raise ExtractorError('Unable to read video info webpage: %s' % compat_str(err))
  77
  78         try:
  79             json_data = json.loads(json_code)
  80             if 'Post' in json_data:
  81                 data = json_data['Post']
  82             else:
  83                 data = json_data
  84
  85             upload_date = datetime.datetime.strptime(data['datestamp'], '%m-%d-%y %H:%M%p').strftime('%Y%m%d')
  86             formats = []
  87             if 'additionalMedia' in data:
  88                 for f in sorted(data['additionalMedia'], key=lambda f: int(f['media_height'])):
  89                     if not int(f['media_width']): # filter m3u8
  90                         continue
  91                     formats.append({
  92                         'url': f['url'],
  93                         'format_id': f['role'],
  94                         'width': int(f['media_width']),
  95                         'height': int(f['media_height']),
  96                     })
  97             else:
  98                 formats.append({
  99                     'url': data['media']['url'],
 100                     'width': int(data['media']['width']),
 101                     'height': int(data['media']['height']),
 102                 })
 103
 104             self._sort_formats(formats)
 105
 106             return {
 107                 'id': compat_str(data['item_id']),
 108                 'uploader': data['display_name'],
 109                 'upload_date': upload_date,
 110                 'title': data['title'],
 111                 'thumbnail': data['thumbnailUrl'],
 112                 'description': data['description'],
 113                 'user_agent': 'iTunes/10.6.1',
 114                 'formats': formats,
 115             }
 116         except (ValueError, KeyError) as err:
 117             raise ExtractorError('Unable to parse video information: %s' % repr(err))
 118
 119
 120 class BlipTVUserIE(InfoExtractor):
 121     """Information Extractor for blip.tv users."""
 122
 123     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?blip\.tv/)|bliptvuser:)([^/]+)/*$'
 124     _PAGE_SIZE = 12
 125     IE_NAME = 'blip.tv:user'
 126
 127     def _real_extract(self, url):
 128         # Extract username
 129         mobj = re.match(self._VALID_URL, url)
 130         if mobj is None:
 131             raise ExtractorError('Invalid URL: %s' % url)
 132
 133         username = mobj.group(1)
 134
 135         page_base = 'http://m.blip.tv/pr/show_get_full_episode_list?users_id=%s&lite=0&esi=1'
 136
 137         page = self._download_webpage(url, username, 'Downloading user page')
 138         mobj = re.search(r'data-users-id="([^"]+)"', page)
 139         page_base = page_base % mobj.group(1)
 140
 141
 142         # Download video ids using BlipTV Ajax calls. Result size per
 143         # query is limited (currently to 12 videos) so we need to query
 144         # page by page until there are no video ids - it means we got
 145         # all of them.
 146
 147         video_ids = []
 148         pagenum = 1
 149
 150         while True:
 151             url = page_base + "&page=" + str(pagenum)
 152             page = self._download_webpage(url, username,
 153                                           'Downloading video ids from page %d' % pagenum)
 154
 155             # Extract video identifiers
 156             ids_in_page = []
 157
 158             for mobj in re.finditer(r'href="/([^"]+)"', page):
 159                 if mobj.group(1) not in ids_in_page:
 160                     ids_in_page.append(unescapeHTML(mobj.group(1)))
 161
 162             video_ids.extend(ids_in_page)
 163
 164             # A little optimization - if current page is not
 165             # "full", ie. does not contain PAGE_SIZE video ids then
 166             # we can assume that this page is the last one - there
 167             # are no more ids on further pages - no need to query
 168             # again.
 169
 170             if len(ids_in_page) < self._PAGE_SIZE:
 171                 break
 172
 173             pagenum += 1
 174
 175         urls = ['http://blip.tv/%s' % video_id for video_id in video_ids]
 176         url_entries = [self.url_result(vurl, 'BlipTV') for vurl in urls]
 177         return [self.playlist_result(url_entries, playlist_title = username)]