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