Include Danny Colligan's YouTube search InfoExtractor
authorRicardo Garcia <sarbalap+freshmeat@gmail.com>
Mon, 2 Feb 2009 18:59:48 +0000 (19:59 +0100)
committerRicardo Garcia <sarbalap+freshmeat@gmail.com>
Sun, 31 Oct 2010 10:23:47 +0000 (11:23 +0100)
youtube-dl

index 1ad297c26dd62eb484eb576cfb9711404636851e..29afd930f0782ccfc01b74150e9aefbe885307c7 100755 (executable)
@@ -721,6 +721,90 @@ class MetacafeIE(InfoExtractor):
                        'ext':          video_extension.decode('utf-8'),
                        }]
 
+
+class YoutubeSearchIE(InfoExtractor):
+       """Information Extractor for YouTube search queries."""
+       _VALID_QUERY = 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'>Next</a>'
+       _youtube_ie = None
+
+       def __init__(self, youtube_ie, downloader=None): 
+               InfoExtractor.__init__(self, downloader)
+               self._youtube_ie = youtube_ie
+       
+       @staticmethod
+       def suitable(url):
+               return (re.match(YoutubeSearchIE._VALID_QUERY, url) is not None)
+
+       def report_download_page(self, query, pagenum):
+               """Report attempt to download playlist page with given number."""
+               self.to_stdout(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
+
+       def _real_initialize(self):
+               self._youtube_ie.initialize()
+       
+       def _real_extract(self, query):
+               mobj = re.match(self._VALID_QUERY, query)
+               if mobj is None:
+                       self.to_stderr(u'ERROR: invalid search query "%s"' % query)
+                       return [None]
+
+               prefix, query = query.split(':')
+               prefix = prefix[8:]
+               if prefix == '': 
+                       return self._download_n_results(query, 1)
+               elif prefix == 'all': 
+                       return self._download_n_results(query, -1)
+               else: 
+                       try:
+                               n = int(prefix)
+                               if n <= 0:
+                                       self.to_stderr(u'ERROR: invalid download number %s for query "%s"' % (n, query))
+                                       return [None]
+                               return self._download_n_results(query, n)
+                       except ValueError: # parsing prefix as int fails
+                               return self._download_n_results(query, 1)
+
+       def _download_n_results(self, query, n):
+               """Downloads a specified number of results for a query"""
+
+               video_ids = []
+               already_seen = set()
+               pagenum = 1
+
+               while True:
+                       self.report_download_page(query, pagenum)
+                       result_url = self._TEMPLATE_URL % (urllib.quote(query.replace(' ', '+')), pagenum)
+                       request = urllib2.Request(result_url, None, std_headers)
+                       try:
+                               page = urllib2.urlopen(request).read()
+                       except (urllib2.URLError, httplib.HTTPException, socket.error), err:
+                               self.to_stderr(u'ERROR: unable to download webpage: %s' % str(err))
+                               return [None]
+
+                       # 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
+                                               information = []
+                                               for id in video_ids:
+                                                       information.extend(self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id))
+                                               return information
+
+                       if self._MORE_PAGES_INDICATOR not in page:
+                               information = []
+                               for id in video_ids:
+                                       information.extend(self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id))
+                               return information
+
+                       pagenum = pagenum + 1
+
 class YoutubePlaylistIE(InfoExtractor):
        """Information Extractor for YouTube playlists."""
 
@@ -925,6 +1009,7 @@ if __name__ == '__main__':
                youtube_ie = YoutubeIE()
                metacafe_ie = MetacafeIE(youtube_ie)
                youtube_pl_ie = YoutubePlaylistIE(youtube_ie)
+               youtube_search_ie = YoutubeSearchIE(youtube_ie)
 
                # File downloader
                charset = locale.getdefaultlocale()[1]
@@ -946,6 +1031,7 @@ if __name__ == '__main__':
                        'ignoreerrors': opts.ignoreerrors,
                        'ratelimit': opts.ratelimit,
                        })
+               fd.add_info_extractor(youtube_search_ie)
                fd.add_info_extractor(youtube_pl_ie)
                fd.add_info_extractor(metacafe_ie)
                fd.add_info_extractor(youtube_ie)