]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube-dl
Imported Upstream version 2010.10.03
[youtubedl] / youtube-dl
index 48c13b701cce6d13382eae08e84b25cf60396600..51344f27be8932102674821016fb30506c155691 100755 (executable)
@@ -27,7 +27,7 @@ except ImportError:
        from cgi import parse_qs
 
 std_headers = {
-       'User-Agent': 'Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.7) Gecko/20100720 Firefox/3.6.7',
+       'User-Agent': 'Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.8) Gecko/20100723 Firefox/3.6.8',
        'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
        'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
        'Accept-Language': 'en-us,en;q=0.5',
@@ -194,7 +194,7 @@ class FileDownloader(object):
        ignoreerrors:   Do not stop on download errors.
        ratelimit:      Download speed limit, in bytes/sec.
        nooverwrites:   Prevent overwriting files.
-       retries:        Number of times to retry for HTTP error 503
+       retries:        Number of times to retry for HTTP error 5xx
        continuedl:     Try to continue downloads if possible.
        noprogress:     Do not print the progress bar.
        """
@@ -287,16 +287,6 @@ class FileDownloader(object):
                multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
                return long(round(number * multiplier))
 
-       @staticmethod
-       def verify_url(url):
-               """Verify a URL is valid and data could be downloaded. Return real data URL."""
-               request = urllib2.Request(url, None, std_headers)
-               data = urllib2.urlopen(request)
-               data.read(1)
-               url = data.geturl()
-               data.close()
-               return url
-
        def add_info_extractor(self, ie):
                """Add an InfoExtractor object to the end of the list."""
                self._ies.append(ie)
@@ -363,12 +353,12 @@ class FileDownloader(object):
                                (percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
 
        def report_resuming_byte(self, resume_len):
-               """Report attemtp to resume at given byte."""
+               """Report attempt to resume at given byte."""
                self.to_stdout(u'[download] Resuming download at byte %s' % resume_len)
        
        def report_retry(self, count, retries):
-               """Report retry in case of HTTP error 503"""
-               self.to_stdout(u'[download] Got HTTP error 503. Retrying (attempt %d of %d)...' % (count, retries))
+               """Report retry in case of HTTP error 5xx"""
+               self.to_stdout(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
        
        def report_file_already_downloaded(self, file_name):
                """Report file has already been fully downloaded."""
@@ -396,13 +386,6 @@ class FileDownloader(object):
                """Process a single dictionary returned by an InfoExtractor."""
                # Do nothing else if in simulate mode
                if self.params.get('simulate', False):
-                       # Verify URL if it's an HTTP one
-                       if info_dict['url'].startswith('http'):
-                               try:
-                                       self.verify_url(info_dict['url'].encode('utf-8')).decode('utf-8')
-                               except (OSError, IOError, urllib2.URLError, httplib.HTTPException, socket.error), err:
-                                       raise UnavailableVideoError
-
                        # Forced printings
                        if self.params.get('forcetitle', False):
                                print info_dict['title'].encode(preferredencoding(), 'xmlcharrefreplace')
@@ -421,15 +404,16 @@ class FileDownloader(object):
                        template_dict['ord'] = unicode('%05d' % self._num_downloads)
                        filename = self.params['outtmpl'] % template_dict
                except (ValueError, KeyError), err:
-                       self.trouble('ERROR: invalid output template or system charset: %s' % str(err))
+                       self.trouble(u'ERROR: invalid system charset or erroneous output template')
+                       return
                if self.params.get('nooverwrites', False) and os.path.exists(filename):
-                       self.to_stderr(u'WARNING: file exists: %s; skipping' % filename)
+                       self.to_stderr(u'WARNING: file exists and will be skipped')
                        return
 
                try:
                        self.pmkdir(filename)
                except (OSError, IOError), err:
-                       self.trouble('ERROR: unable to create directories: %s' % str(err))
+                       self.trouble(u'ERROR: unable to create directories: %s' % str(err))
                        return
 
                try:
@@ -437,17 +421,17 @@ class FileDownloader(object):
                except (OSError, IOError), err:
                        raise UnavailableVideoError
                except (urllib2.URLError, httplib.HTTPException, socket.error), err:
-                       self.trouble('ERROR: unable to download video data: %s' % str(err))
+                       self.trouble(u'ERROR: unable to download video data: %s' % str(err))
                        return
                except (ContentTooShortError, ), err:
-                       self.trouble('ERROR: content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
+                       self.trouble(u'ERROR: content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
                        return
 
                if success:
                        try:
                                self.post_process(filename, info_dict)
                        except (PostProcessingError), err:
-                               self.trouble('ERROR: postprocessing: %s' % str(err))
+                               self.trouble(u'ERROR: postprocessing: %s' % str(err))
                                return
 
        def download(self, url_list):
@@ -472,7 +456,7 @@ class FileDownloader(object):
                                break
 
                        if not suitable_found:
-                               self.trouble('ERROR: no suitable InfoExtractor: %s' % url)
+                               self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url)
 
                return self._download_retcode
 
@@ -512,7 +496,7 @@ class FileDownloader(object):
                        self.to_stdout(u'\r[rtmpdump] %s bytes' % os.path.getsize(filename))
                        return True
                else:
-                       self.trouble('\nERROR: rtmpdump exited with code %d' % retval)
+                       self.trouble(u'\nERROR: rtmpdump exited with code %d' % retval)
                        return False
 
        def _do_download(self, filename, url, player_url):
@@ -539,32 +523,50 @@ class FileDownloader(object):
 
                count = 0
                retries = self.params.get('retries', 0)
-               while True:
+               while count <= retries:
                        # Establish connection
                        try:
                                data = urllib2.urlopen(request)
                                break
                        except (urllib2.HTTPError, ), err:
-                               if err.code == 503:
-                                       # Retry in case of HTTP error 503
-                                       count += 1
-                                       if count <= retries:
-                                               self.report_retry(count, retries)
-                                               continue
-                               if err.code != 416: #  416 is 'Requested range not satisfiable'
+                               if (err.code < 500 or err.code >= 600) and err.code != 416:
+                                       # Unexpected HTTP error
                                        raise
-                               # Unable to resume
-                               data = urllib2.urlopen(basic_request)
-                               content_length = data.info()['Content-Length']
-
-                               if content_length is not None and long(content_length) == resume_len:
-                                       # Because the file had already been fully downloaded
-                                       self.report_file_already_downloaded(filename)
-                                       return True
-                               else:
-                                       # Because the server didn't let us
-                                       self.report_unable_to_resume()
-                                       open_mode = 'wb'
+                               elif err.code == 416:
+                                       # Unable to resume (requested range not satisfiable)
+                                       try:
+                                               # Open the connection again without the range header
+                                               data = urllib2.urlopen(basic_request)
+                                               content_length = data.info()['Content-Length']
+                                       except (urllib2.HTTPError, ), err:
+                                               if err.code < 500 or err.code >= 600:
+                                                       raise
+                                       else:
+                                               # Examine the reported length
+                                               if (content_length is not None and
+                                                   (resume_len - 100 < long(content_length) < resume_len + 100)):
+                                                       # The file had already been fully downloaded.
+                                                       # Explanation to the above condition: in issue #175 it was revealed that
+                                                       # YouTube sometimes adds or removes a few bytes from the end of the file,
+                                                       # changing the file size slightly and causing problems for some users. So
+                                                       # I decided to implement a suggested change and consider the file
+                                                       # completely downloaded if the file size differs less than 100 bytes from
+                                                       # the one in the hard drive.
+                                                       self.report_file_already_downloaded(filename)
+                                                       return True
+                                               else:
+                                                       # The length does not match, we start the download over
+                                                       self.report_unable_to_resume()
+                                                       open_mode = 'wb'
+                                                       break
+                       # Retry
+                       count += 1
+                       if count <= retries:
+                               self.report_retry(count, retries)
+
+               if count > retries:
+                       self.trouble(u'ERROR: giving up after %s retries' % retries)
+                       return False
 
                data_len = data.info().get('Content-length', None)
                data_len_str = self.format_bytes(data_len)
@@ -587,12 +589,13 @@ class FileDownloader(object):
                                        (stream, filename) = sanitize_open(filename, open_mode)
                                        self.report_destination(filename)
                                except (OSError, IOError), err:
-                                       self.trouble('ERROR: unable to open for writing: %s' % str(err))
+                                       self.trouble(u'ERROR: unable to open for writing: %s' % str(err))
                                        return False
                        try:
                                stream.write(data_block)
                        except (IOError, OSError), err:
-                               self.trouble('\nERROR: unable to write data: %s' % str(err))
+                               self.trouble(u'\nERROR: unable to write data: %s' % str(err))
+                               return False
                        block_size = self.best_block_size(after - before, data_block_len)
 
                        # Progress message
@@ -683,7 +686,7 @@ class InfoExtractor(object):
 class YoutubeIE(InfoExtractor):
        """Information extractor for youtube.com."""
 
-       _VALID_URL = r'^((?:http://)?(?:youtu\.be/|(?:\w+\.)?youtube\.com/(?:(?:v/)|(?:(?:watch(?:_popup)?(?:\.php)?)?[\?#](?:.+&)?v=))))?([0-9A-Za-z_-]+)(?(1).+)?$'
+       _VALID_URL = r'^((?:http://)?(?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/(?:(?:v/)|(?:(?:watch(?:_popup)?(?:\.php)?)?(?:\?|#!?)(?:.+&)?v=))))?([0-9A-Za-z_-]+)(?(1).+)?$'
        _LANG_URL = r'http://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
        _LOGIN_URL = 'http://www.youtube.com/signup?next=/&gl=US&hl=en'
        _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
@@ -844,6 +847,14 @@ class YoutubeIE(InfoExtractor):
                        except (urllib2.URLError, httplib.HTTPException, socket.error), err:
                                self._downloader.trouble(u'ERROR: unable to download video info webpage: %s' % str(err))
                                return
+               if 'token' not in video_info:
+                       if 'reason' in video_info:
+                               self._downloader.trouble(u'ERROR: YouTube said: %s' % video_info['reason'][0].decode('utf-8'))
+                       else:
+                               self._downloader.trouble(u'ERROR: "token" parameter not in video info for unknown reason')
+                       return
+
+               # Start extracting information
                self.report_information_extraction(video_id)
 
                # uploader
@@ -1011,7 +1022,6 @@ class MetacafeIE(InfoExtractor):
                self._downloader.increment_downloads()
 
                simple_title = mobj.group(2).decode('utf-8')
-               video_extension = 'flv'
 
                # Retrieve video webpage to extract further information
                request = urllib2.Request('http://www.metacafe.com/watch/%s/' % video_id)
@@ -1025,20 +1035,33 @@ class MetacafeIE(InfoExtractor):
                # Extract URL, uploader and title from webpage
                self.report_extraction(video_id)
                mobj = re.search(r'(?m)&mediaURL=([^&]+)', webpage)
-               if mobj is None:
-                       self._downloader.trouble(u'ERROR: unable to extract media URL')
-                       return
-               mediaURL = urllib.unquote(mobj.group(1))
-
-               #mobj = re.search(r'(?m)&gdaKey=(.*?)&', webpage)
-               #if mobj is None:
-               #       self._downloader.trouble(u'ERROR: unable to extract gdaKey')
-               #       return
-               #gdaKey = mobj.group(1)
-               #
-               #video_url = '%s?__gda__=%s' % (mediaURL, gdaKey)
-
-               video_url = mediaURL
+               if mobj is not None:
+                       mediaURL = urllib.unquote(mobj.group(1))
+                       video_extension = mediaURL[-3:]
+                       
+                       # Extract gdaKey if available
+                       mobj = re.search(r'(?m)&gdaKey=(.*?)&', webpage)
+                       if mobj is None:
+                               video_url = mediaURL
+                       else:
+                               gdaKey = mobj.group(1)
+                               video_url = '%s?__gda__=%s' % (mediaURL, gdaKey)
+               else:
+                       mobj = re.search(r' name="flashvars" value="(.*?)"', webpage)
+                       if mobj is None:
+                               self._downloader.trouble(u'ERROR: unable to extract media URL')
+                               return
+                       vardict = parse_qs(mobj.group(1))
+                       if 'mediaData' not in vardict:
+                               self._downloader.trouble(u'ERROR: unable to extract media URL')
+                               return
+                       mobj = re.search(r'"mediaURL":"(http.*?)","key":"(.*?)"', vardict['mediaData'][0])
+                       if mobj is None:
+                               self._downloader.trouble(u'ERROR: unable to extract media URL')
+                               return
+                       mediaURL = mobj.group(1).replace('\\/', '/')
+                       video_extension = mediaURL[-3:]
+                       video_url = '%s?__gda__=%s' % (mediaURL, mobj.group(2))
 
                mobj = re.search(r'(?im)<title>(.*) - Video</title>', webpage)
                if mobj is None:
@@ -1135,7 +1158,7 @@ class DailymotionIE(InfoExtractor):
                video_title = mobj.group(1).decode('utf-8')
                video_title = sanitize_title(video_title)
 
-               mobj = re.search(r'(?im)<div class="dmco_html owner">.*?<a class="name" href="/.+?">(.+?)</a></div>', webpage)
+               mobj = re.search(r'(?im)<div class="dmco_html owner">.*?<a class="name" href="/.+?">(.+?)</a>', webpage)
                if mobj is None:
                        self._downloader.trouble(u'ERROR: unable to extract uploader nickname')
                        return
@@ -1932,6 +1955,11 @@ class YoutubePlaylistIE(InfoExtractor):
                                break
                        pagenum = pagenum + 1
 
+               playliststart = self._downloader.params.get('playliststart', 1)
+               playliststart -= 1 #our arrays are zero-based but the playlist is 1-based
+               if playliststart > 0:
+                       video_ids = video_ids[playliststart:]
+                       
                for id in video_ids:
                        self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id)
                return
@@ -1987,6 +2015,11 @@ class YoutubeUserIE(InfoExtractor):
                                ids_in_page.append(mobj.group(1))
                video_ids.extend(ids_in_page)
 
+               playliststart = self._downloader.params.get('playliststart', 1)
+               playliststart = playliststart-1 #our arrays are zero-based but the playlist is 1-based
+               if playliststart > 0:
+                       video_ids = video_ids[playliststart:]   
+
                for id in video_ids:
                        self._youtube_ie.extract('http://www.youtube.com/watch?v=%s' % id)
                return
@@ -2068,7 +2101,7 @@ if __name__ == '__main__':
                # Parse command line
                parser = optparse.OptionParser(
                        usage='Usage: %prog [options] url...',
-                       version='2010.07.24',
+                       version='2010.10.03',
                        conflict_handler='resolve',
                )
 
@@ -2084,6 +2117,8 @@ if __name__ == '__main__':
                                dest='ratelimit', metavar='LIMIT', help='download rate limit (e.g. 50k or 44.6m)')
                parser.add_option('-R', '--retries',
                                dest='retries', metavar='RETRIES', help='number of retries (default is 10)', default=10)
+               parser.add_option('--playlist-start',
+                               dest='playliststart', metavar='NUMBER', help='playlist video to start at (default is 1)', default=1)
 
                authentication = optparse.OptionGroup(parser, 'Authentication Options')
                authentication.add_option('-u', '--username',
@@ -2103,6 +2138,8 @@ if __name__ == '__main__':
                                action='store_const', dest='format', help='download all available video formats', const='-1')
                video_format.add_option('--max-quality',
                                action='store', dest='format_limit', metavar='FORMAT', help='highest quality format to download')
+               video_format.add_option('-b', '--best-quality',
+                               action='store_true', dest='bestquality', help='download the best video quality (DEPRECATED)')
                parser.add_option_group(video_format)
 
                verbosity = optparse.OptionGroup(parser, 'Verbosity / Simulation Options')
@@ -2155,6 +2192,8 @@ if __name__ == '__main__':
                all_urls = batchurls + args
 
                # Conflicting, missing and erroneous options
+               if opts.bestquality:
+                       print >>sys.stderr, u'\nWARNING: -b/--best-quality IS DEPRECATED AS IT IS THE DEFAULT BEHAVIOR NOW\n'
                if opts.usenetrc and (opts.username is not None or opts.password is not None):
                        parser.error(u'using .netrc conflicts with giving username/password')
                if opts.password is not None and opts.username is None:
@@ -2175,6 +2214,11 @@ if __name__ == '__main__':
                                opts.retries = long(opts.retries)
                        except (TypeError, ValueError), err:
                                parser.error(u'invalid retry count specified')
+               if opts.playliststart is not None:
+                       try:
+                               opts.playliststart = long(opts.playliststart)
+                       except (TypeError, ValueError), err:
+                               parser.error(u'invalid playlist page specified')
 
                # Information extractors
                youtube_ie = YoutubeIE()
@@ -2216,6 +2260,7 @@ if __name__ == '__main__':
                        'retries': opts.retries,
                        'continuedl': opts.continue_dl,
                        'noprogress': opts.noprogress,
+                       'playliststart': opts.playliststart,
                        })
                fd.add_info_extractor(youtube_search_ie)
                fd.add_info_extractor(youtube_pl_ie)