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

   1 import datetime
   2 import json
   3 import re
   4 import socket
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_http_client,
   9     compat_parse_qs,
  10     compat_str,
  11     compat_urllib_error,
  12     compat_urllib_parse_urlparse,
  13     compat_urllib_request,
  14
  15     ExtractorError,
  16     unescapeHTML,
  17     determine_ext,
  18 )
  19
  20
  21 class BlipTVIE(InfoExtractor):
  22     """Information extractor for blip.tv"""
  23
  24     _VALID_URL = r'^(?:https?://)?(?:www\.)?blip\.tv/((.+/)|(play/)|(api\.swf#))(.+)$'
  25     IE_NAME = u'blip.tv'
  26     _TEST = {
  27         u'url': u'http://blip.tv/cbr/cbr-exclusive-gotham-city-imposters-bats-vs-jokerz-short-3-5796352',
  28         u'file': u'5779306.m4v',
  29         u'md5': u'80baf1ec5c3d2019037c1c707d676b9f',
  30         u'info_dict': {
  31             u"upload_date": u"20111205",
  32             u"description": u"md5:9bc31f227219cde65e47eeec8d2dc596",
  33             u"uploader": u"Comic Book Resources - CBR TV",
  34             u"title": u"CBR EXCLUSIVE: \"Gotham City Imposters\" Bats VS Jokerz Short 3"
  35         }
  36     }
  37
  38     def report_direct_download(self, title):
  39         """Report information extraction."""
  40         self.to_screen(u'%s: Direct download detected' % title)
  41
  42     def _real_extract(self, url):
  43         mobj = re.match(self._VALID_URL, url)
  44         if mobj is None:
  45             raise ExtractorError(u'Invalid URL: %s' % url)
  46
  47         # See https://github.com/rg3/youtube-dl/issues/857
  48         api_mobj = re.match(r'http://a\.blip\.tv/api\.swf#(?P<video_id>[\d\w]+)', url)
  49         if api_mobj is not None:
  50             url = 'http://blip.tv/play/g_%s' % api_mobj.group('video_id')
  51         urlp = compat_urllib_parse_urlparse(url)
  52         if urlp.path.startswith('/play/'):
  53             response = self._request_webpage(url, None, False)
  54             redirecturl = response.geturl()
  55             rurlp = compat_urllib_parse_urlparse(redirecturl)
  56             file_id = compat_parse_qs(rurlp.fragment)['file'][0].rpartition('/')[2]
  57             url = 'http://blip.tv/a/a-' + file_id
  58             return self._real_extract(url)
  59
  60         if '?' in url:
  61             cchar = '&'
  62         else:
  63             cchar = '?'
  64         json_url = url + cchar + 'skin=json&version=2&no_wrap=1'
  65         request = compat_urllib_request.Request(json_url)
  66         request.add_header('User-Agent', 'iTunes/10.6.1')
  67         self.report_extraction(mobj.group(1))
  68         urlh = self._request_webpage(request, None, False,
  69             u'unable to download video info webpage')
  70
  71         try:
  72             json_code_bytes = urlh.read()
  73             json_code = json_code_bytes.decode('utf-8')
  74         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
  75             raise ExtractorError(u'Unable to read video info webpage: %s' % compat_str(err))
  76
  77         try:
  78             json_data = json.loads(json_code)
  79             if 'Post' in json_data:
  80                 data = json_data['Post']
  81             else:
  82                 data = json_data
  83
  84             upload_date = datetime.datetime.strptime(data['datestamp'], '%m-%d-%y %H:%M%p').strftime('%Y%m%d')
  85             formats = []
  86             if 'additionalMedia' in data:
  87                 for f in sorted(data['additionalMedia'], key=lambda f: int(f['media_height'])):
  88                     if not int(f['media_width']): # filter m3u8
  89                         continue
  90                     formats.append({
  91                         'url': f['url'],
  92                         'ext': determine_ext(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                     'ext': determine_ext(data['media']['url']),
 101                     'width': int(data['media']['width']),
 102                     'height': int(data['media']['height']),
 103                 })
 104
 105             return {
 106                 'id': compat_str(data['item_id']),
 107                 'uploader': data['display_name'],
 108                 'upload_date': upload_date,
 109                 'title': data['title'],
 110                 'thumbnail': data['thumbnailUrl'],
 111                 'description': data['description'],
 112                 'user_agent': 'iTunes/10.6.1',
 113                 'formats': formats,
 114             }
 115         except (ValueError, KeyError) as err:
 116             raise ExtractorError(u'Unable to parse video information: %s' % repr(err))
 117
 118
 119 class BlipTVUserIE(InfoExtractor):
 120     """Information Extractor for blip.tv users."""
 121
 122     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?blip\.tv/)|bliptvuser:)([^/]+)/*$'
 123     _PAGE_SIZE = 12
 124     IE_NAME = u'blip.tv:user'
 125
 126     def _real_extract(self, url):
 127         # Extract username
 128         mobj = re.match(self._VALID_URL, url)
 129         if mobj is None:
 130             raise ExtractorError(u'Invalid URL: %s' % url)
 131
 132         username = mobj.group(1)
 133
 134         page_base = 'http://m.blip.tv/pr/show_get_full_episode_list?users_id=%s&lite=0&esi=1'
 135
 136         page = self._download_webpage(url, username, u'Downloading user page')
 137         mobj = re.search(r'data-users-id="([^"]+)"', page)
 138         page_base = page_base % mobj.group(1)
 139
 140
 141         # Download video ids using BlipTV Ajax calls. Result size per
 142         # query is limited (currently to 12 videos) so we need to query
 143         # page by page until there are no video ids - it means we got
 144         # all of them.
 145
 146         video_ids = []
 147         pagenum = 1
 148
 149         while True:
 150             url = page_base + "&page=" + str(pagenum)
 151             page = self._download_webpage(url, username,
 152                                           u'Downloading video ids from page %d' % pagenum)
 153
 154             # Extract video identifiers
 155             ids_in_page = []
 156
 157             for mobj in re.finditer(r'href="/([^"]+)"', page):
 158                 if mobj.group(1) not in ids_in_page:
 159                     ids_in_page.append(unescapeHTML(mobj.group(1)))
 160
 161             video_ids.extend(ids_in_page)
 162
 163             # A little optimization - if current page is not
 164             # "full", ie. does not contain PAGE_SIZE video ids then
 165             # we can assume that this page is the last one - there
 166             # are no more ids on further pages - no need to query
 167             # again.
 168
 169             if len(ids_in_page) < self._PAGE_SIZE:
 170                 break
 171
 172             pagenum += 1
 173
 174         urls = [u'http://blip.tv/%s' % video_id for video_id in video_ids]
 175         url_entries = [self.url_result(vurl, 'BlipTV') for vurl in urls]
 176         return [self.playlist_result(url_entries, playlist_title = username)]