Merge remote-tracking branch 'origin/master'
[youtube-dl] / youtube_dl / YoutubeDL.py
index 095692bb3eb6db2228ad26d30043fbe82d51df2a..e2332f9b84ef2c41d6cf986553d4f999bac946a6 100644 (file)
@@ -91,7 +91,7 @@ class YoutubeDL(object):
     downloadarchive:   File name of a file where all downloads are recorded.
                        Videos already present in the file are not downloaded
                        again.
-    
+
     The following parameters are not used by YoutubeDL itself, they are used by
     the FileDownloader:
     nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
@@ -216,10 +216,10 @@ class YoutubeDL(object):
         If stderr is a tty file the 'WARNING:' will be colored
         '''
         if sys.stderr.isatty() and os.name != 'nt':
-            _msg_header=u'\033[0;33mWARNING:\033[0m'
+            _msg_header = u'\033[0;33mWARNING:\033[0m'
         else:
-            _msg_header=u'WARNING:'
-        warning_message=u'%s %s' % (_msg_header,message)
+            _msg_header = u'WARNING:'
+        warning_message = u'%s %s' % (_msg_header, message)
         self.to_stderr(warning_message)
 
     def report_error(self, message, tb=None):
@@ -234,19 +234,6 @@ class YoutubeDL(object):
         error_message = u'%s %s' % (_msg_header, message)
         self.trouble(error_message, tb)
 
-    def slow_down(self, start_time, byte_counter):
-        """Sleep if the download speed is over the rate limit."""
-        rate_limit = self.params.get('ratelimit', None)
-        if rate_limit is None or byte_counter == 0:
-            return
-        now = time.time()
-        elapsed = now - start_time
-        if elapsed <= 0.0:
-            return
-        speed = float(byte_counter) / elapsed
-        if speed > rate_limit:
-            time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
-
     def report_writedescription(self, descfn):
         """ Report that the description file is being written """
         self.to_screen(u'[info] Writing video description to: ' + descfn)
@@ -330,14 +317,14 @@ class YoutubeDL(object):
             return (u'%(title)s has already been recorded in archive'
                     % info_dict)
         return None
-        
+
     def extract_info(self, url, download=True, ie_key=None, extra_info={}):
         '''
         Returns a list with a dictionary for each video we find.
         If 'download', also downloads the videos.
         extra_info is a dict containing the extra values to add to each result
          '''
-        
+
         if ie_key:
             ies = [self.get_info_extractor(ie_key)]
         else:
@@ -379,7 +366,7 @@ class YoutubeDL(object):
                     raise
         else:
             self.report_error(u'no suitable InfoExtractor: %s' % url)
-        
+
     def process_ie_result(self, ie_result, download=True, extra_info={}):
         """
         Take the result of the ie(may be modified) and resolve all unresolved
@@ -403,7 +390,7 @@ class YoutubeDL(object):
         elif result_type == 'playlist':
             # We process each entry in the playlist
             playlist = ie_result.get('title', None) or ie_result.get('id', None)
-            self.to_screen(u'[download] Downloading playlist: %s'  % playlist)
+            self.to_screen(u'[download] Downloading playlist: %s' % playlist)
 
             playlist_results = []
 
@@ -421,12 +408,12 @@ class YoutubeDL(object):
             self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
                 (ie_result['extractor'], playlist, n_all_entries, n_entries))
 
-            for i,entry in enumerate(entries,1):
-                self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
+            for i, entry in enumerate(entries, 1):
+                self.to_screen(u'[download] Downloading video #%s of %s' % (i, n_entries))
                 extra = {
-                         'playlist': playlist, 
-                         'playlist_index': i + playliststart,
-                         }
+                    'playlist': playlist,
+                    'playlist_index': i + playliststart,
+                }
                 if not 'extractor' in entry:
                     # We set the extractor, if it's an url it will be set then to
                     # the new extractor, but if it's already a video we must make
@@ -450,6 +437,22 @@ class YoutubeDL(object):
         else:
             raise Exception('Invalid result type: %s' % result_type)
 
+    def select_format(self, format_spec, available_formats):
+        if format_spec == 'best' or format_spec is None:
+            return available_formats[-1]
+        elif format_spec == 'worst':
+            return available_formats[0]
+        else:
+            extensions = [u'mp4', u'flv', u'webm', u'3gp']
+            if format_spec in extensions:
+                filter_f = lambda f: f['ext'] == format_spec
+            else:
+                filter_f = lambda f: f['format_id'] == format_spec
+            matches = list(filter(filter_f, available_formats))
+            if matches:
+                return matches[-1]
+        return None
+
     def process_video_result(self, info_dict, download=True):
         assert info_dict.get('_type', 'video') == 'video'
 
@@ -460,7 +463,8 @@ class YoutubeDL(object):
 
         # This extractors handle format selection themselves
         if info_dict['extractor'] in [u'youtube', u'Youku', u'YouPorn', u'mixcloud']:
-            self.process_info(info_dict)
+            if download:
+                self.process_info(info_dict)
             return info_dict
 
         # We now pick which formats have to be downloaded
@@ -472,17 +476,14 @@ class YoutubeDL(object):
 
         # We check that all the formats have the format and format_id fields
         for (i, format) in enumerate(formats):
-            if format.get('format') is None:
-                if format.get('height') is not None:
-                    if format.get('width') is not None:
-                        format_desc = u'%sx%s' % (format['width'], format['height'])
-                    else:
-                        format_desc = u'%sp' % format['height']
-                else:
-                    format_desc = '???'
-                format['format'] = format_desc
             if format.get('format_id') is None:
                 format['format_id'] = compat_str(i)
+            if format.get('format') is None:
+                format['format'] = u'{id} - {res}{note}'.format(
+                    id=format['format_id'],
+                    res=self.format_resolution(format),
+                    note=u' ({})'.format(format['format_note']) if format.get('format_note') is not None else '',
+                )
 
         if self.params.get('listformats', None):
             self.list_formats(info_dict)
@@ -504,22 +505,20 @@ class YoutubeDL(object):
             formats = sorted(formats, key=_free_formats_key)
 
         req_format = self.params.get('format', 'best')
+        if req_format is None:
+            req_format = 'best'
         formats_to_download = []
-        if req_format == 'best' or req_format is None:
-            formats_to_download = [formats[-1]]
-        elif req_format == 'worst':
-            formats_to_download = [formats[0]]
         # The -1 is for supporting YoutubeIE
-        elif req_format in ('-1', 'all'):
+        if req_format in ('-1', 'all'):
             formats_to_download = formats
         else:
-            # We can accept formats requestd in the format: 34/10/5, we pick
+            # We can accept formats requestd in the format: 34/5/best, we pick
             # the first that is available, starting from left
             req_formats = req_format.split('/')
             for rf in req_formats:
-                matches = filter(lambda f:f['format_id'] == rf ,formats)
-                if matches:
-                    formats_to_download = [matches[0]]
+                selected_format = self.select_format(rf, formats)
+                if selected_format is not None:
+                    formats_to_download = [selected_format]
                     break
         if not formats_to_download:
             raise ExtractorError(u'requested format not available')
@@ -610,20 +609,20 @@ class YoutubeDL(object):
 
         if self.params.get('writeannotations', False):
             try:
-               annofn = filename + u'.annotations.xml'
-               self.report_writeannotations(annofn)
-               with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
-                   annofile.write(info_dict['annotations'])
+                annofn = filename + u'.annotations.xml'
+                self.report_writeannotations(annofn)
+                with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
+                    annofile.write(info_dict['annotations'])
             except (KeyError, TypeError):
                 self.report_warning(u'There are no annotations to write.')
             except (OSError, IOError):
-                 self.report_error(u'Cannot write annotations file: ' + annofn)
-                 return
+                self.report_error(u'Cannot write annotations file: ' + annofn)
+                return
 
         subtitles_are_requested = any([self.params.get('writesubtitles', False),
                                        self.params.get('writeautomaticsub')])
 
-        if  subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
+        if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
             # subtitles download errors are already managed as troubles in relevant IE
             # that way it will silently go on when used with unsupporting IE
             subtitles = info_dict['subtitles']
@@ -645,7 +644,7 @@ class YoutubeDL(object):
             infofn = filename + u'.info.json'
             self.report_writeinfojson(infofn)
             try:
-                json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
+                json_info_dict = dict((k, v) for k, v in info_dict.items() if not k in ['urlhandle'])
                 write_json_file(json_info_dict, encodeFilename(infofn))
             except (OSError, IOError):
                 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
@@ -715,7 +714,7 @@ class YoutubeDL(object):
         keep_video = None
         for pp in self._pps:
             try:
-                keep_video_wish,new_info = pp.run(info)
+                keep_video_wish, new_info = pp.run(info)
                 if keep_video_wish is not None:
                     if keep_video_wish:
                         keep_video = keep_video_wish
@@ -754,16 +753,31 @@ class YoutubeDL(object):
         with locked_file(fn, 'a', encoding='utf-8') as archive_file:
             archive_file.write(vid_id + u'\n')
 
+    @staticmethod
+    def format_resolution(format):
+        if format.get('height') is not None:
+            if format.get('width') is not None:
+                res = u'%sx%s' % (format['width'], format['height'])
+            else:
+                res = u'%sp' % format['height']
+        else:
+            res = '???'
+        return res
+
     def list_formats(self, info_dict):
         formats_s = []
         for format in info_dict.get('formats', [info_dict]):
-            formats_s.append("%s\t:\t%s\t[%s]" % (format['format_id'],
-                                                format['ext'],
-                                                format.get('format', '???'),
-                                                )
-                            )
+            formats_s.append(u'%-15s: %-5s     %-15s[%s]' % (
+                format['format_id'],
+                format['ext'],
+                format.get('format_note') or '-',
+                self.format_resolution(format),
+                )
+            )
         if len(formats_s) != 1:
-            formats_s[0]  += ' (worst)'
+            formats_s[0] += ' (worst)'
             formats_s[-1] += ' (best)'
         formats_s = "\n".join(formats_s)
-        self.to_screen(u"[info] Available formats for %s:\nformat code\textension\n%s" % (info_dict['id'], formats_s)) 
+        self.to_screen(u'[info] Available formats for %s:\n'
+            u'format code    extension   note           resolution\n%s' % (
+                info_dict['id'], formats_s))