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_str,
   9     compat_urllib_request,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     clean_html,
  14     int_or_none,
  15     parse_iso8601,
  16     unescapeHTML,
  17 )
  18
  19
  20 class BlipTVIE(InfoExtractor):
  21     _VALID_URL = r'https?://(?:\w+\.)?blip\.tv/(?:(?:.+-|rss/flash/)(?P<id>\d+)|((?:play/|api\.swf#)(?P<lookup_id>[\da-zA-Z+_]+)))'
  22
  23     _TESTS = [
  24         {
  25             'url': 'http://blip.tv/cbr/cbr-exclusive-gotham-city-imposters-bats-vs-jokerz-short-3-5796352',
  26             'md5': 'c6934ad0b6acf2bd920720ec888eb812',
  27             'info_dict': {
  28                 'id': '5779306',
  29                 'ext': 'mov',
  30                 'title': 'CBR EXCLUSIVE: "Gotham City Imposters" Bats VS Jokerz Short 3',
  31                 'description': 'md5:9bc31f227219cde65e47eeec8d2dc596',
  32                 'timestamp': 1323138843,
  33                 'upload_date': '20111206',
  34                 'uploader': 'cbr',
  35                 'uploader_id': '679425',
  36                 'duration': 81,
  37             }
  38         },
  39         {
  40             # https://github.com/rg3/youtube-dl/pull/2274
  41             'note': 'Video with subtitles',
  42             'url': 'http://blip.tv/play/h6Uag5OEVgI.html',
  43             'md5': '309f9d25b820b086ca163ffac8031806',
  44             'info_dict': {
  45                 'id': '6586561',
  46                 'ext': 'mp4',
  47                 'title': 'Red vs. Blue Season 11 Episode 1',
  48                 'description': 'One-Zero-One',
  49                 'timestamp': 1371261608,
  50                 'upload_date': '20130615',
  51                 'uploader': 'redvsblue',
  52                 'uploader_id': '792887',
  53                 'duration': 279,
  54             }
  55         },
  56         {
  57             # https://bugzilla.redhat.com/show_bug.cgi?id=967465
  58             'url': 'http://a.blip.tv/api.swf#h6Uag5KbVwI',
  59             'md5': '314e87b1ebe7a48fcbfdd51b791ce5a6',
  60             'info_dict': {
  61                 'id': '6573122',
  62                 'ext': 'mov',
  63                 'upload_date': '20130520',
  64                 'description': 'Two hapless space marines argue over what to do when they realize they have an astronomically huge problem on their hands.',
  65                 'title': 'Red vs. Blue Season 11 Trailer',
  66                 'timestamp': 1369029609,
  67                 'uploader': 'redvsblue',
  68                 'uploader_id': '792887',
  69             }
  70         },
  71         {
  72             'url': 'http://blip.tv/play/gbk766dkj4Yn',
  73             'md5': 'fe0a33f022d49399a241e84a8ea8b8e3',
  74             'info_dict': {
  75                 'id': '1749452',
  76                 'ext': 'mp4',
  77                 'upload_date': '20090208',
  78                 'description': 'Witness the first appearance of the Nostalgia Critic character, as Doug reviews the movie Transformers.',
  79                 'title': 'Nostalgia Critic: Transformers',
  80                 'timestamp': 1234068723,
  81                 'uploader': 'NostalgiaCritic',
  82                 'uploader_id': '246467',
  83             }
  84         },
  85         {
  86             # https://github.com/rg3/youtube-dl/pull/4404
  87             'note': 'Audio only',
  88             'url': 'http://blip.tv/hilarios-productions/weekly-manga-recap-kingdom-7119982',
  89             'md5': '76c0a56f24e769ceaab21fbb6416a351',
  90             'info_dict': {
  91                 'id': '7103299',
  92                 'ext': 'flv',
  93                 'title': 'Weekly Manga Recap: Kingdom',
  94                 '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?',
  95                 'timestamp': 1417660321,
  96                 'upload_date': '20141204',
  97                 'uploader': 'The Rollo T',
  98                 'uploader_id': '407429',
  99                 'duration': 7251,
 100                 'vcodec': 'none',
 101             }
 102         },
 103     ]
 104
 105     @staticmethod
 106     def _extract_url(webpage):
 107         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 108         if mobj:
 109             return 'http://blip.tv/a/a-' + mobj.group(1)
 110         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 111         if mobj:
 112             return mobj.group(1)
 113
 114     def _real_extract(self, url):
 115         mobj = re.match(self._VALID_URL, url)
 116         lookup_id = mobj.group('lookup_id')
 117
 118         # See https://github.com/rg3/youtube-dl/issues/857 and
 119         # https://github.com/rg3/youtube-dl/issues/4197
 120         if lookup_id:
 121             urlh = self._request_webpage(
 122                 'http://blip.tv/play/%s' % lookup_id, lookup_id, 'Resolving lookup id')
 123             url = compat_urlparse.urlparse(urlh.geturl())
 124             qs = compat_urlparse.parse_qs(url.query)
 125             mobj = re.match(self._VALID_URL, qs['file'][0])
 126
 127         video_id = mobj.group('id')
 128
 129         rss = self._download_xml('http://blip.tv/rss/flash/%s' % video_id, video_id, 'Downloading video RSS')
 130
 131         def blip(s):
 132             return '{http://blip.tv/dtd/blip/1.0}%s' % s
 133
 134         def media(s):
 135             return '{http://search.yahoo.com/mrss/}%s' % s
 136
 137         def itunes(s):
 138             return '{http://www.itunes.com/dtds/podcast-1.0.dtd}%s' % s
 139
 140         item = rss.find('channel/item')
 141
 142         video_id = item.find(blip('item_id')).text
 143         title = item.find('./title').text
 144         description = clean_html(compat_str(item.find(blip('puredescription')).text))
 145         timestamp = parse_iso8601(item.find(blip('datestamp')).text)
 146         uploader = item.find(blip('user')).text
 147         uploader_id = item.find(blip('userid')).text
 148         duration = int(item.find(blip('runtime')).text)
 149         media_thumbnail = item.find(media('thumbnail'))
 150         thumbnail = media_thumbnail.get('url') if media_thumbnail is not None else item.find(itunes('image')).text
 151         categories = [category.text for category in item.findall('category')]
 152
 153         formats = []
 154         subtitles_urls = {}
 155
 156         media_group = item.find(media('group'))
 157         for media_content in media_group.findall(media('content')):
 158             url = media_content.get('url')
 159             role = media_content.get(blip('role'))
 160             msg = self._download_webpage(
 161                 url + '?showplayer=20140425131715&referrer=http://blip.tv&mask=7&skin=flashvars&view=url',
 162                 video_id, 'Resolving URL for %s' % role)
 163             real_url = compat_urlparse.parse_qs(msg.strip())['message'][0]
 164
 165             media_type = media_content.get('type')
 166             if media_type == 'text/srt' or url.endswith('.srt'):
 167                 LANGS = {
 168                     'english': 'en',
 169                 }
 170                 lang = role.rpartition('-')[-1].strip().lower()
 171                 langcode = LANGS.get(lang, lang)
 172                 subtitles_urls[langcode] = url
 173             elif media_type.startswith('video/'):
 174                 formats.append({
 175                     'url': real_url,
 176                     'format_id': role,
 177                     'format_note': media_type,
 178                     'vcodec': media_content.get(blip('vcodec')) or 'none',
 179                     'acodec': media_content.get(blip('acodec')),
 180                     'filesize': media_content.get('filesize'),
 181                     'width': int_or_none(media_content.get('width')),
 182                     'height': int_or_none(media_content.get('height')),
 183                 })
 184         self._check_formats(formats, video_id)
 185         self._sort_formats(formats)
 186
 187         subtitles = self.extract_subtitles(video_id, subtitles_urls)
 188
 189         return {
 190             'id': video_id,
 191             'title': title,
 192             'description': description,
 193             'timestamp': timestamp,
 194             'uploader': uploader,
 195             'uploader_id': uploader_id,
 196             'duration': duration,
 197             'thumbnail': thumbnail,
 198             'categories': categories,
 199             'formats': formats,
 200             'subtitles': subtitles,
 201         }
 202
 203     def _get_subtitles(self, video_id, subtitles_urls):
 204         subtitles = {}
 205         for lang, url in subtitles_urls.items():
 206             # For some weird reason, blip.tv serves a video instead of subtitles
 207             # when we request with a common UA
 208             req = compat_urllib_request.Request(url)
 209             req.add_header('User-Agent', 'youtube-dl')
 210             subtitles[lang] = [{
 211                 # The extension is 'srt' but it's actually an 'ass' file
 212                 'ext': 'ass',
 213                 'data': self._download_webpage(req, None, note=False),
 214             }]
 215         return subtitles
 216
 217
 218 class BlipTVUserIE(InfoExtractor):
 219     _VALID_URL = r'(?:(?:https?://(?:\w+\.)?blip\.tv/)|bliptvuser:)(?!api\.swf)([^/]+)/*$'
 220     _PAGE_SIZE = 12
 221     IE_NAME = 'blip.tv:user'
 222     _TEST = {
 223         'url': 'http://blip.tv/actone',
 224         'info_dict': {
 225             'id': 'actone',
 226             'title': 'Act One: The Series',
 227         },
 228         'playlist_count': 5,
 229     }
 230
 231     def _real_extract(self, url):
 232         mobj = re.match(self._VALID_URL, url)
 233         username = mobj.group(1)
 234
 235         page_base = 'http://m.blip.tv/pr/show_get_full_episode_list?users_id=%s&lite=0&esi=1'
 236
 237         page = self._download_webpage(url, username, 'Downloading user page')
 238         mobj = re.search(r'data-users-id="([^"]+)"', page)
 239         page_base = page_base % mobj.group(1)
 240         title = self._og_search_title(page)
 241
 242         # Download video ids using BlipTV Ajax calls. Result size per
 243         # query is limited (currently to 12 videos) so we need to query
 244         # page by page until there are no video ids - it means we got
 245         # all of them.
 246
 247         video_ids = []
 248         pagenum = 1
 249
 250         while True:
 251             url = page_base + "&page=" + str(pagenum)
 252             page = self._download_webpage(
 253                 url, username, 'Downloading video ids from page %d' % pagenum)
 254
 255             # Extract video identifiers
 256             ids_in_page = []
 257
 258             for mobj in re.finditer(r'href="/([^"]+)"', page):
 259                 if mobj.group(1) not in ids_in_page:
 260                     ids_in_page.append(unescapeHTML(mobj.group(1)))
 261
 262             video_ids.extend(ids_in_page)
 263
 264             # A little optimization - if current page is not
 265             # "full", ie. does not contain PAGE_SIZE video ids then
 266             # we can assume that this page is the last one - there
 267             # are no more ids on further pages - no need to query
 268             # again.
 269
 270             if len(ids_in_page) < self._PAGE_SIZE:
 271                 break
 272
 273             pagenum += 1
 274
 275         urls = ['http://blip.tv/%s' % video_id for video_id in video_ids]
 276         url_entries = [self.url_result(vurl, 'BlipTV') for vurl in urls]
 277         return self.playlist_result(
 278             url_entries, playlist_title=title, playlist_id=username)