diff --git a/youtube-dl b/youtube-dl index 5c81973cd..689427f5b 100755 --- a/youtube-dl +++ b/youtube-dl @@ -2314,9 +2314,7 @@ class GenericIE(InfoExtractor): class YoutubeSearchIE(InfoExtractor): """Information Extractor for YouTube search queries.""" _VALID_URL = r'ytsearch(\d+|all)?:[\s\S]+' - _TEMPLATE_URL = 'http://www.youtube.com/results?search_query=%s&page=%s&gl=US&hl=en' - _VIDEO_INDICATOR = r'href="/watch\?v=.+?"' - _MORE_PAGES_INDICATOR = r'(?m)>\s*Next\s*' + _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc' _youtube_ie = None _max_youtube_results = 1000 IE_NAME = u'youtube:search' @@ -2367,37 +2365,31 @@ class YoutubeSearchIE(InfoExtractor): """Downloads a specified number of results for a query""" video_ids = [] - already_seen = set() - pagenum = 1 + pagenum = 0 + limit = n - while True: - self.report_download_page(query, pagenum) - result_url = self._TEMPLATE_URL % (urllib.quote_plus(query), pagenum) + while (50 * pagenum) < limit: + self.report_download_page(query, pagenum+1) + result_url = self._API_URL % (urllib.quote_plus(query), (50*pagenum)+1) request = urllib2.Request(result_url) try: - page = urllib2.urlopen(request).read() + data = urllib2.urlopen(request).read() except (urllib2.URLError, httplib.HTTPException, socket.error), err: - self._downloader.trouble(u'ERROR: unable to download webpage: %s' % str(err)) + self._downloader.trouble(u'ERROR: unable to download API page: %s' % str(err)) return + api_response = json.loads(data)['data'] - # Extract video identifiers - for mobj in re.finditer(self._VIDEO_INDICATOR, page): - video_id = page[mobj.span()[0]:mobj.span()[1]].split('=')[2][:-1] - if video_id not in already_seen: - video_ids.append(video_id) - already_seen.add(video_id) - if len(video_ids) == n: - # Specified n videos reached - for id in video_ids: - self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) - return + new_ids = list(video['id'] for video in api_response['items']) + video_ids += new_ids - if re.search(self._MORE_PAGES_INDICATOR, page) is None: - for id in video_ids: - self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) - return + limit = min(n, api_response['totalItems']) + pagenum += 1 - pagenum = pagenum + 1 + if len(video_ids) > n: + video_ids = video_ids[:n] + for id in video_ids: + self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) + return class GoogleSearchIE(InfoExtractor): @@ -2581,7 +2573,7 @@ class YoutubePlaylistIE(InfoExtractor): _VALID_URL = r'(?:https?://)?(?:\w+\.)?youtube\.com/(?:(?:course|view_play_list|my_playlists|artist|playlist)\?.*?(p|a|list)=|user/.*?/user/|p/|user/.*?#[pg]/c/)(?:PL)?([0-9A-Za-z-_]+)(?:/.*?/([0-9A-Za-z_-]+))?.*' _TEMPLATE_URL = 'http://www.youtube.com/%s?%s=%s&page=%s&gl=US&hl=en' - _VIDEO_INDICATOR = r'/watch\?v=(.+?)&' + _VIDEO_INDICATOR_TEMPLATE = r'/watch\?v=(.+?)&list=PL%s&' _MORE_PAGES_INDICATOR = r'(?m)>\s*Next\s*' _youtube_ie = None IE_NAME = u'youtube:playlist' @@ -2633,7 +2625,7 @@ class YoutubePlaylistIE(InfoExtractor): # Extract video identifiers ids_in_page = [] - for mobj in re.finditer(self._VIDEO_INDICATOR, page): + for mobj in re.finditer(self._VIDEO_INDICATOR_TEMPLATE % playlist_id, page): if mobj.group(1) not in ids_in_page: ids_in_page.append(mobj.group(1)) video_ids.extend(ids_in_page) @@ -2644,7 +2636,10 @@ class YoutubePlaylistIE(InfoExtractor): playliststart = self._downloader.params.get('playliststart', 1) - 1 playlistend = self._downloader.params.get('playlistend', -1) - video_ids = video_ids[playliststart:playlistend] + if playlistend == -1: + video_ids = video_ids[playliststart:] + else: + video_ids = video_ids[playliststart:playlistend] for id in video_ids: self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) diff --git a/youtube_dl/__init__.py b/youtube_dl/__init__.py index 789724041..3b9e2d70e 100755 --- a/youtube_dl/__init__.py +++ b/youtube_dl/__init__.py @@ -2058,7 +2058,7 @@ def _real_extract(self, url, new_video=True): video_id = mobj.group(1) # Retrieve video webpage to extract further information - request = urllib2.Request("http://vimeo.com/moogaloop/load/clip:%s" % video_id, None, std_headers) + request = urllib2.Request(url, None, std_headers) try: self.report_download_webpage(video_id) webpage = urllib2.urlopen(request).read() @@ -2071,77 +2071,75 @@ def _real_extract(self, url, new_video=True): # and latter we extract those that are Vimeo specific. self.report_extraction(video_id) - # Extract title - mobj = re.search(r'(.*?)', webpage) - if mobj is None: - self._downloader.trouble(u'ERROR: unable to extract video title') + # Extract the config JSON + config = webpage.split(' = {config:')[1].split(',assets:')[0] + try: + config = json.loads(config) + except: + self._downloader.trouble(u'ERROR: unable to extract info section') return - video_title = mobj.group(1).decode('utf-8') + + # Extract title + video_title = config["video"]["title"] simple_title = _simplify_title(video_title) # Extract uploader - mobj = re.search(r'http://vimeo.com/(.*?)', webpage) - if mobj is None: - self._downloader.trouble(u'ERROR: unable to extract video uploader') - return - video_uploader = mobj.group(1).decode('utf-8') + video_uploader = config["video"]["owner"]["name"] # Extract video thumbnail - mobj = re.search(r'(.*?)', webpage) - if mobj is None: - self._downloader.trouble(u'ERROR: unable to extract video thumbnail') - return - video_thumbnail = mobj.group(1).decode('utf-8') + video_thumbnail = config["video"]["thumbnail"] - # # Extract video description - # mobj = re.search(r'', webpage) - # if mobj is None: - # self._downloader.trouble(u'ERROR: unable to extract video description') - # return - # video_description = mobj.group(1).decode('utf-8') - # if not video_description: video_description = 'No description available.' - video_description = 'Foo.' - - # Vimeo specific: extract request signature - mobj = re.search(r'(.*?)', webpage) - if mobj is None: - self._downloader.trouble(u'ERROR: unable to extract request signature') - return - sig = mobj.group(1).decode('utf-8') - - # Vimeo specific: extract video quality information - mobj = re.search(r'(\d+)', webpage) - if mobj is None: - self._downloader.trouble(u'ERROR: unable to extract video quality information') - return - quality = mobj.group(1).decode('utf-8') - - if int(quality) == 1: - quality = 'hd' + # Extract video description + try: + lxml.etree + except NameError: + video_description = u'No description available.' + mobj = re.search(r'', webpage, re.MULTILINE) + if mobj is not None: + video_description = mobj.group(1) else: - quality = 'sd' + html_parser = lxml.etree.HTMLParser() + vwebpage_doc = lxml.etree.parse(StringIO.StringIO(webpage), html_parser) + video_description = u''.join(vwebpage_doc.xpath('id("description")//text()')).strip() + # TODO use another parser - # Vimeo specific: Extract request signature expiration - mobj = re.search(r'(.*?)', webpage) - if mobj is None: - self._downloader.trouble(u'ERROR: unable to extract request signature expiration') + # Extract upload date + video_upload_date = u'NA' + mobj = re.search(r'', webpage) + if mobj is not None: + video_upload_date = mobj.group(1) + + # Vimeo specific: extract request signature and timestamp + sig = config['request']['signature'] + timestamp = config['request']['timestamp'] + + # Vimeo specific: extract video codec and quality information + # TODO bind to format param + codecs = [('h264', 'mp4'), ('vp8', 'flv'), ('vp6', 'flv')] + for codec in codecs: + if codec[0] in config["video"]["files"]: + video_codec = codec[0] + video_extension = codec[1] + if 'hd' in config["video"]["files"][codec[0]]: quality = 'hd' + else: quality = 'sd' + break + else: + self._downloader.trouble(u'ERROR: no known codec found') return - sig_exp = mobj.group(1).decode('utf-8') - video_url = "http://vimeo.com/moogaloop/play/clip:%s/%s/%s/?q=%s" % (video_id, sig, sig_exp, quality) + video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \ + %(video_id, sig, timestamp, quality, video_codec.upper()) try: # Process video information self._downloader.process_info({ - 'id': video_id.decode('utf-8'), + 'id': video_id, 'url': video_url, 'uploader': video_uploader, - 'upload_date': u'NA', + 'upload_date': video_upload_date, 'title': video_title, 'stitle': simple_title, - 'ext': u'mp4', - 'thumbnail': video_thumbnail.decode('utf-8'), - 'description': video_description, + 'ext': video_extension, 'thumbnail': video_thumbnail, 'description': video_description, 'player_url': None, @@ -2250,9 +2248,7 @@ def _real_extract(self, url): class YoutubeSearchIE(InfoExtractor): """Information Extractor for YouTube search queries.""" _VALID_URL = r'ytsearch(\d+|all)?:[\s\S]+' - _TEMPLATE_URL = 'http://www.youtube.com/results?search_query=%s&page=%s&gl=US&hl=en' - _VIDEO_INDICATOR = r'href="/watch\?v=.+?"' - _MORE_PAGES_INDICATOR = r'(?m)>\s*Next\s*' + _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc' _youtube_ie = None _max_youtube_results = 1000 IE_NAME = u'youtube:search' @@ -2303,45 +2299,39 @@ def _download_n_results(self, query, n): """Downloads a specified number of results for a query""" video_ids = [] - already_seen = set() - pagenum = 1 + pagenum = 0 + limit = n - while True: - self.report_download_page(query, pagenum) - result_url = self._TEMPLATE_URL % (urllib.quote_plus(query), pagenum) + while (50 * pagenum) < limit: + self.report_download_page(query, pagenum+1) + result_url = self._API_URL % (urllib.quote_plus(query), (50*pagenum)+1) request = urllib2.Request(result_url) try: - page = urllib2.urlopen(request).read() + data = urllib2.urlopen(request).read() except (urllib2.URLError, httplib.HTTPException, socket.error), err: - self._downloader.trouble(u'ERROR: unable to download webpage: %s' % str(err)) + self._downloader.trouble(u'ERROR: unable to download API page: %s' % str(err)) return + api_response = json.loads(data)['data'] - # Extract video identifiers - for mobj in re.finditer(self._VIDEO_INDICATOR, page): - video_id = page[mobj.span()[0]:mobj.span()[1]].split('=')[2][:-1] - if video_id not in already_seen: - video_ids.append(video_id) - already_seen.add(video_id) - if len(video_ids) == n: - # Specified n videos reached - for id in video_ids: - self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) - return + new_ids = list(video['id'] for video in api_response['items']) + video_ids += new_ids - if re.search(self._MORE_PAGES_INDICATOR, page) is None: - for id in video_ids: - self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) - return + limit = min(n, api_response['totalItems']) + pagenum += 1 - pagenum = pagenum + 1 + if len(video_ids) > n: + video_ids = video_ids[:n] + for id in video_ids: + self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id) + return class GoogleSearchIE(InfoExtractor): """Information Extractor for Google Video search queries.""" _VALID_URL = r'gvsearch(\d+|all)?:[\s\S]+' _TEMPLATE_URL = 'http://video.google.com/videosearch?q=%s+site:video.google.com&start=%s&hl=en' - _VIDEO_INDICATOR = r'videoplay\?docid=([^\&>]+)\&' - _MORE_PAGES_INDICATOR = r'Next' + _VIDEO_INDICATOR = r'\s*Next\s*' _youtube_ie = None IE_NAME = u'youtube:playlist' @@ -2571,7 +2559,7 @@ def _real_extract(self, url): # Extract video identifiers ids_in_page = [] - for mobj in re.finditer(self._VIDEO_INDICATOR, page): + for mobj in re.finditer(self._VIDEO_INDICATOR_TEMPLATE % playlist_id, page): if mobj.group(1) not in ids_in_page: ids_in_page.append(mobj.group(1)) video_ids.extend(ids_in_page) @@ -2582,7 +2570,10 @@ def _real_extract(self, url): playliststart = self._downloader.params.get('playliststart', 1) - 1 playlistend = self._downloader.params.get('playlistend', -1) - video_ids = video_ids[playliststart:playlistend] + if playlistend == -1: + video_ids = video_ids[playliststart:] + else: + video_ids = video_ids[playliststart:playlistend] for id in video_ids: self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id)