_VALID_URL = r"""^
(
(?:https?://)? # http(s):// (optional)
- (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/|
+ (?:(?:(?:(?:\w+\.)?youtube(?:-nocookie)?\.com/|
tube\.majestyc\.net/) # the various hostnames, with wildcard subdomains
(?:.*?\#/)? # handle anchor (#/) redirect urls
(?: # the various things that can precede the ID:
(?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx)
v=
)
- )? # optional -> youtube.com/xxxx is OK
+ ))
+ |youtu\.be/ # just youtu.be/xxxx
+ )
)? # all until now is optional -> you can pass the naked ID
([0-9A-Za-z_-]+) # here is it! the YouTube video ID
(?(1).+)? # if we found the ID, everything can follow
elif len(s) == 87:
return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
elif len(s) == 86:
- return s[81:73:-1] + s[84] + s[72:58:-1] + s[0] + s[57:35:-1] + s[85] + s[34:0:-1]
+ return s[5:34] + s[0] + s[35:38] + s[3] + s[39:45] + s[38] + s[46:53] + s[73] + s[54:73] + s[85] + s[74:85] + s[53]
elif len(s) == 85:
return s[83:34:-1] + s[0] + s[33:27:-1] + s[3] + s[26:19:-1] + s[34] + s[18:3:-1] + s[27]
elif len(s) == 84:
manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
formats_urls = _get_urls(manifest)
for format_url in formats_urls:
- itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
+ itag = self._search_regex(r'itag%3D(\d+?)/', format_url, 'itag')
url_map[itag] = format_url
return url_map
for entry in response['feed']['entry']:
index = entry['yt$position']['$t']
- if 'media$group' in entry and 'media$player' in entry['media$group']:
- videos.append((index, entry['media$group']['media$player']['url']))
+ if 'media$group' in entry and 'yt$videoid' in entry['media$group']:
+ videos.append((
+ index,
+ 'https://www.youtube.com/watch?v=' + entry['media$group']['yt$videoid']['$t']
+ ))
videos = [v[1] for v in sorted(videos)]
class YoutubeUserIE(InfoExtractor):
IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
- _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
+ _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?)|ytuser:)([A-Za-z0-9_-]+)'
_TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
_GDATA_PAGE_SIZE = 50
- _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
- _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
+ _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json'
IE_NAME = u'youtube:user'
+ def suitable(cls, url):
+ if YoutubeIE.suitable(url): return False
+ else: return super(YoutubeUserIE, cls).suitable(url)
+
def _real_extract(self, url):
# Extract username
mobj = re.match(self._VALID_URL, url)
page = self._download_webpage(gdata_url, username,
u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
+ try:
+ response = json.loads(page)
+ except ValueError as err:
+ raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
+
# Extract video identifiers
ids_in_page = []
-
- for mobj in re.finditer(self._VIDEO_INDICATOR, page):
- if mobj.group(1) not in ids_in_page:
- ids_in_page.append(mobj.group(1))
-
+ for entry in response['feed']['entry']:
+ ids_in_page.append(entry['id']['$t'].split('/')[-1])
video_ids.extend(ids_in_page)
# A little optimization - if current page is not