diff options
| -rwxr-xr-x | youtube-dl | 79 |
1 files changed, 56 insertions, 23 deletions
diff --git a/youtube-dl b/youtube-dl index 8dd03daf3..c8c2ea88e 100755 --- a/youtube-dl +++ b/youtube-dl @@ -5,6 +5,7 @@ # Author: Benjamin Johnson # Author: Vasyl' Vavrychuk # Author: Witold Baryluk +# Author: Paweł Paprota # License: Public domain code import cookielib import ctypes @@ -2159,9 +2160,11 @@ class YoutubePlaylistIE(InfoExtractor): class YoutubeUserIE(InfoExtractor): """Information Extractor for YouTube users.""" - _VALID_URL = r'(?:http://)?(?:\w+\.)?youtube.com/user/(.*)' + _VALID_URL = r'(?:(?:(?:http://)?(?:\w+\.)?youtube.com/user/)|ytuser:)([A-Za-z0-9_-]+)' _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s' - _VIDEO_INDICATOR = r'http://gdata.youtube.com/feeds/api/videos/(.*)' # XXX Fix this. + _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=(.+?)&' _youtube_ie = None def __init__(self, youtube_ie, downloader=None): @@ -2172,9 +2175,10 @@ class YoutubeUserIE(InfoExtractor): def suitable(url): return (re.match(YoutubeUserIE._VALID_URL, url) is not None) - def report_download_page(self, username): + def report_download_page(self, username, start_index): """Report attempt to download user page.""" - self._downloader.to_screen(u'[youtube] user %s: Downloading page ' % (username)) + self._downloader.to_screen(u'[youtube] user %s: Downloading video ids from %d to %d' % + (username, start_index, start_index + self._GDATA_PAGE_SIZE)) def _real_initialize(self): self._youtube_ie.initialize() @@ -2186,34 +2190,63 @@ class YoutubeUserIE(InfoExtractor): self._downloader.trouble(u'ERROR: invalid url: %s' % url) return - # Download user page username = mobj.group(1) + + # Download video ids using YouTube Data API. Result size per + # query is limited (currently to 50 videos) so we need to query + # page by page until there are no video ids - it means we got + # all of them. + video_ids = [] - pagenum = 1 + pagenum = 0 - self.report_download_page(username) - request = urllib2.Request(self._TEMPLATE_URL % (username)) - try: - page = urllib2.urlopen(request).read() - except (urllib2.URLError, httplib.HTTPException, socket.error), err: - self._downloader.trouble(u'ERROR: unable to download webpage: %s' % str(err)) - return + while True: + start_index = pagenum * self._GDATA_PAGE_SIZE + 1 + self.report_download_page(username, start_index) - # Extract video identifiers - ids_in_page = [] + request = urllib2.Request(self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)) - 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)) - video_ids.extend(ids_in_page) + try: + page = urllib2.urlopen(request).read() + except (urllib2.URLError, httplib.HTTPException, socket.error), err: + self._downloader.trouble(u'ERROR: unable to download webpage: %s' % str(err)) + return + # 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)) + + video_ids.extend(ids_in_page) + + # A little optimization - if current page is not + # "full", ie. does not contain PAGE_SIZE video ids then + # we can assume that this page is the last one - there + # are no more ids on further pages - no need to query + # again. + + if len(ids_in_page) < self._GDATA_PAGE_SIZE: + break + + pagenum += 1 + + all_ids_count = len(video_ids) playliststart = self._downloader.params.get('playliststart', 1) - 1 playlistend = self._downloader.params.get('playlistend', -1) - video_ids = video_ids[playliststart:playlistend] - for id in video_ids: - self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) - return + if playlistend == -1: + video_ids = video_ids[playliststart:] + else: + video_ids = video_ids[playliststart:playlistend] + + self._downloader.to_screen("[youtube] user %s: Collected %d video ids (downloading %d of them)" % + (username, all_ids_count, len(video_ids))) + + for video_id in video_ids: + self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % video_id) + class DepositFilesIE(InfoExtractor): """Information extractor for depositfiles.com""" |