Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/bliptv.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6
   7 from ..compat import (
   8     compat_urllib_request,
   9     compat_urlparse,
  10 )
  11 from ..utils import (
  12     clean_html,
  13     int_or_none,
  14     parse_iso8601,
  15     unescapeHTML,
  16     xpath_text,
  17     xpath_with_ns,
  18 )
  19
  20
  21 class BlipTVIE(InfoExtractor):
  22     _VALID_URL = r'https?://(?:\w+\.)?blip\.tv/(?:(?:.+-|rss/flash/)(?P<id>\d+)|((?:play/|api\.swf#)(?P<lookup_id>[\da-zA-Z+_]+)))'
  23
  24     _TESTS = [
  25         {
  26             'url': 'http://blip.tv/cbr/cbr-exclusive-gotham-city-imposters-bats-vs-jokerz-short-3-5796352',
  27             'md5': '80baf1ec5c3d2019037c1c707d676b9f',
  28             'info_dict': {
  29                 'id': '5779306',
  30                 'ext': 'm4v',
  31                 'title': 'CBR EXCLUSIVE: "Gotham City Imposters" Bats VS Jokerz Short 3',
  32                 'description': 'md5:9bc31f227219cde65e47eeec8d2dc596',
  33                 'timestamp': 1323138843,
  34                 'upload_date': '20111206',
  35                 'uploader': 'cbr',
  36                 'uploader_id': '679425',
  37                 'duration': 81,
  38             }
  39         },
  40         {
  41             # https://github.com/rg3/youtube-dl/pull/2274
  42             'note': 'Video with subtitles',
  43             'url': 'http://blip.tv/play/h6Uag5OEVgI.html',
  44             'md5': '309f9d25b820b086ca163ffac8031806',
  45             'info_dict': {
  46                 'id': '6586561',
  47                 'ext': 'mp4',
  48                 'title': 'Red vs. Blue Season 11 Episode 1',
  49                 'description': 'One-Zero-One',
  50                 'timestamp': 1371261608,
  51                 'upload_date': '20130615',
  52                 'uploader': 'redvsblue',
  53                 'uploader_id': '792887',
  54                 'duration': 279,
  55             }
  56         },
  57         {
  58             # https://bugzilla.redhat.com/show_bug.cgi?id=967465
  59             'url': 'http://a.blip.tv/api.swf#h6Uag5KbVwI',
  60             'md5': '314e87b1ebe7a48fcbfdd51b791ce5a6',
  61             'info_dict': {
  62                 'id': '6573122',
  63                 'ext': 'mov',
  64                 'upload_date': '20130520',
  65                 'description': 'Two hapless space marines argue over what to do when they realize they have an astronomically huge problem on their hands.',
  66                 'title': 'Red vs. Blue Season 11 Trailer',
  67                 'timestamp': 1369029609,
  68                 'uploader': 'redvsblue',
  69                 'uploader_id': '792887',
  70             }
  71         },
  72         {
  73             'url': 'http://blip.tv/play/gbk766dkj4Yn',
  74             'md5': 'fe0a33f022d49399a241e84a8ea8b8e3',
  75             'info_dict': {
  76                 'id': '1749452',
  77                 'ext': 'mp4',
  78                 'upload_date': '20090208',
  79                 'description': 'Witness the first appearance of the Nostalgia Critic character, as Doug reviews the movie Transformers.',
  80                 'title': 'Nostalgia Critic: Transformers',
  81                 'timestamp': 1234068723,
  82                 'uploader': 'NostalgiaCritic',
  83                 'uploader_id': '246467',
  84             }
  85         },
  86         {
  87             # https://github.com/rg3/youtube-dl/pull/4404
  88             'note': 'Audio only',
  89             'url': 'http://blip.tv/hilarios-productions/weekly-manga-recap-kingdom-7119982',
  90             'md5': '76c0a56f24e769ceaab21fbb6416a351',
  91             'info_dict': {
  92                 'id': '7103299',
  93                 'ext': 'flv',
  94                 'title': 'Weekly Manga Recap: Kingdom',
  95                 'description': 'And then Shin breaks the enemy line, and he&apos;s all like HWAH! And then he slices a guy and it&apos;s all like FWASHING! And... it&apos;s really hard to describe the best parts of this series without breaking down into sound effects, okay?',
  96                 'timestamp': 1417660321,
  97                 'upload_date': '20141204',
  98                 'uploader': 'The Rollo T',
  99                 'uploader_id': '407429',
 100                 'duration': 7251,
 101                 'vcodec': 'none',
 102             }
 103         },
 104         {
 105             # missing duration
 106             'url': 'http://blip.tv/rss/flash/6700880',
 107             'info_dict': {
 108                 'id': '6684191',
 109                 'ext': 'm4v',
 110                 'title': 'Cowboy Bebop: Gateway Shuffle Review',
 111                 'description': 'md5:3acc480c0f9ae157f5fe88547ecaf3f8',
 112                 'timestamp': 1386639757,
 113                 'upload_date': '20131210',
 114                 'uploader': 'sfdebris',
 115                 'uploader_id': '706520',
 116             }
 117         }
 118     ]
 119
 120     @staticmethod
 121     def _extract_url(webpage):
 122         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 123         if mobj:
 124             return 'http://blip.tv/a/a-' + mobj.group(1)
 125         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 126         if mobj:
 127             return mobj.group(1)
 128
 129     def _real_extract(self, url):
 130         mobj = re.match(self._VALID_URL, url)
 131         lookup_id = mobj.group('lookup_id')
 132
 133         # See https://github.com/rg3/youtube-dl/issues/857 and
 134         # https://github.com/rg3/youtube-dl/issues/4197
 135         if lookup_id:
 136             urlh = self._request_webpage(
 137                 'http://blip.tv/play/%s' % lookup_id, lookup_id, 'Resolving lookup id')
 138             url = compat_urlparse.urlparse(urlh.geturl())
 139             qs = compat_urlparse.parse_qs(url.query)
 140             mobj = re.match(self._VALID_URL, qs['file'][0])
 141
 142         video_id = mobj.group('id')
 143
 144         rss = self._download_xml('http://blip.tv/rss/flash/%s' % video_id, video_id, 'Downloading video RSS')
 145
 146         def _x(p):
 147             return xpath_with_ns(p, {
 148                 'blip': 'http://blip.tv/dtd/blip/1.0',
 149                 'media': 'http://search.yahoo.com/mrss/',
 150                 'itunes': 'http://www.itunes.com/dtds/podcast-1.0.dtd',
 151             })
 152
 153         item = rss.find('channel/item')
 154
 155         video_id = xpath_text(item, _x('blip:item_id'), 'video id') or lookup_id
 156         title = xpath_text(item, 'title', 'title', fatal=True)
 157         description = clean_html(xpath_text(item, _x('blip:puredescription'), 'description'))
 158         timestamp = parse_iso8601(xpath_text(item, _x('blip:datestamp'), 'timestamp'))
 159         uploader = xpath_text(item, _x('blip:user'), 'uploader')
 160         uploader_id = xpath_text(item, _x('blip:userid'), 'uploader id')
 161         duration = int_or_none(xpath_text(item, _x('blip:runtime'), 'duration'))
 162         media_thumbnail = item.find(_x('media:thumbnail'))
 163         thumbnail = (media_thumbnail.get('url') if media_thumbnail is not None
 164                      else xpath_text(item, 'image', 'thumbnail'))
 165         categories = [category.text for category in item.findall('category') if category is not None]
 166
 167         formats = []
 168         subtitles_urls = {}
 169
 170         media_group = item.find(_x('media:group'))
 171         for media_content in media_group.findall(_x('media:content')):
 172             url = media_content.get('url')
 173             role = media_content.get(_x('blip:role'))
 174             msg = self._download_webpage(
 175                 url + '?showplayer=20140425131715&referrer=http://blip.tv&mask=7&skin=flashvars&view=url',
 176                 video_id, 'Resolving URL for %s' % role)
 177             real_url = compat_urlparse.parse_qs(msg.strip())['message'][0]
 178
 179             media_type = media_content.get('type')
 180             if media_type == 'text/srt' or url.endswith('.srt'):
 181                 LANGS = {
 182                     'english': 'en',
 183                 }
 184                 lang = role.rpartition('-')[-1].strip().lower()
 185                 langcode = LANGS.get(lang, lang)
 186                 subtitles_urls[langcode] = url
 187             elif media_type.startswith('video/'):
 188                 formats.append({
 189                     'url': real_url,
 190                     'format_id': role,
 191                     'format_note': media_type,
 192                     'vcodec': media_content.get(_x('blip:vcodec')) or 'none',
 193                     'acodec': media_content.get(_x('blip:acodec')),
 194                     'filesize': media_content.get('filesize'),
 195                     'width': int_or_none(media_content.get('width')),
 196                     'height': int_or_none(media_content.get('height')),
 197                 })
 198         self._check_formats(formats, video_id)
 199         self._sort_formats(formats)
 200
 201         subtitles = self.extract_subtitles(video_id, subtitles_urls)
 202
 203         return {
 204             'id': video_id,
 205             'title': title,
 206             'description': description,
 207             'timestamp': timestamp,
 208             'uploader': uploader,
 209             'uploader_id': uploader_id,
 210             'duration': duration,
 211             'thumbnail': thumbnail,
 212             'categories': categories,
 213             'formats': formats,
 214             'subtitles': subtitles,
 215         }
 216
 217     def _get_subtitles(self, video_id, subtitles_urls):
 218         subtitles = {}
 219         for lang, url in subtitles_urls.items():
 220             # For some weird reason, blip.tv serves a video instead of subtitles
 221             # when we request with a common UA
 222             req = compat_urllib_request.Request(url)
 223             req.add_header('User-Agent', 'youtube-dl')
 224             subtitles[lang] = [{
 225                 # The extension is 'srt' but it's actually an 'ass' file
 226                 'ext': 'ass',
 227                 'data': self._download_webpage(req, None, note=False),
 228             }]
 229         return subtitles
 230
 231
 232 class BlipTVUserIE(InfoExtractor):
 233     _VALID_URL = r'(?:(?:https?://(?:\w+\.)?blip\.tv/)|bliptvuser:)(?!api\.swf)([^/]+)/*$'
 234     _PAGE_SIZE = 12
 235     IE_NAME = 'blip.tv:user'
 236     _TEST = {
 237         'url': 'http://blip.tv/actone',
 238         'info_dict': {
 239             'id': 'actone',
 240             'title': 'Act One: The Series',
 241         },
 242         'playlist_count': 5,
 243     }
 244
 245     def _real_extract(self, url):
 246         mobj = re.match(self._VALID_URL, url)
 247         username = mobj.group(1)
 248
 249         page_base = 'http://m.blip.tv/pr/show_get_full_episode_list?users_id=%s&lite=0&esi=1'
 250
 251         page = self._download_webpage(url, username, 'Downloading user page')
 252         mobj = re.search(r'data-users-id="([^"]+)"', page)
 253         page_base = page_base % mobj.group(1)
 254         title = self._og_search_title(page)
 255
 256         # Download video ids using BlipTV Ajax calls. Result size per
 257         # query is limited (currently to 12 videos) so we need to query
 258         # page by page until there are no video ids - it means we got
 259         # all of them.
 260
 261         video_ids = []
 262         pagenum = 1
 263
 264         while True:
 265             url = page_base + "&page=" + str(pagenum)
 266             page = self._download_webpage(
 267                 url, username, 'Downloading video ids from page %d' % pagenum)
 268
 269             # Extract video identifiers
 270             ids_in_page = []
 271
 272             for mobj in re.finditer(r'href="/([^"]+)"', page):
 273                 if mobj.group(1) not in ids_in_page:
 274                     ids_in_page.append(unescapeHTML(mobj.group(1)))
 275
 276             video_ids.extend(ids_in_page)
 277
 278             # A little optimization - if current page is not
 279             # "full", ie. does not contain PAGE_SIZE video ids then
 280             # we can assume that this page is the last one - there
 281             # are no more ids on further pages - no need to query
 282             # again.
 283
 284             if len(ids_in_page) < self._PAGE_SIZE:
 285                 break
 286
 287             pagenum += 1
 288
 289         urls = ['http://blip.tv/%s' % video_id for video_id in video_ids]
 290         url_entries = [self.url_result(vurl, 'BlipTV') for vurl in urls]
 291         return self.playlist_result(
 292             url_entries, playlist_title=title, playlist_id=username)