Do not mutate default arguments
[youtube-dl] / youtube_dl / YoutubeDL.py
index 72ccfa2ae49c7e46cecac4235c67c94ed2614690..b822930cbe358226c70b514287509d391adc116c 100644 (file)
@@ -33,6 +33,7 @@ from .utils import (
     DownloadError,
     encodeFilename,
     ExtractorError,
+    format_bytes,
     locked_file,
     make_HTTPS_handler,
     MaxDownloadsReached,
@@ -104,6 +105,7 @@ class YoutubeDL(object):
     playlistend:       Playlist item to end at.
     matchtitle:        Download only matching titles.
     rejecttitle:       Reject downloads for matching titles.
+    logger:            Log messages to a logging.Logger instance.
     logtostderr:       Log messages to stderr instead of stdout.
     writedescription:  Write the video description to a .description file
     writeinfojson:     Write the video description to a .info.json file
@@ -124,11 +126,12 @@ class YoutubeDL(object):
     noplaylist:        Download single video instead of a playlist if in doubt.
     age_limit:         An integer representing the user's age in years.
                        Unsuitable videos for the given age are skipped.
-    downloadarchive:   File name of a file where all downloads are recorded.
+    download_archive:   File name of a file where all downloads are recorded.
                        Videos already present in the file are not downloaded
                        again.
     cookiefile:        File name where cookies should be read from and dumped to.
-    nocheckcertificate Do not verify SSL certificates
+    nocheckcertificate:Do not verify SSL certificates
+    proxy:             URL of the proxy server to use
 
     The following parameters are not used by YoutubeDL itself, they are used by
     the FileDownloader:
@@ -143,7 +146,7 @@ class YoutubeDL(object):
     _num_downloads = None
     _screen_file = None
 
-    def __init__(self, params):
+    def __init__(self, params=None):
         """Create a FileDownloader object with the given options."""
         self._ies = []
         self._ies_instances = {}
@@ -152,6 +155,7 @@ class YoutubeDL(object):
         self._download_retcode = 0
         self._num_downloads = 0
         self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
+        self.params = {} if params is None else params
 
         if (sys.version_info >= (3,) and sys.platform != 'win32' and
                 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
@@ -161,12 +165,11 @@ class YoutubeDL(object):
                 u'Assuming --restrict-filenames since file system encoding '
                 u'cannot encode all charactes. '
                 u'Set the LC_ALL environment variable to fix this.')
-            params['restrictfilenames'] = True
+            self.params['restrictfilenames'] = True
 
-        self.params = params
         self.fd = FileDownloader(self, self.params)
 
-        if '%(stitle)s' in self.params['outtmpl']:
+        if '%(stitle)s' in self.params.get('outtmpl', ''):
             self.report_warning(u'%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
 
         self._setup_opener()
@@ -203,7 +206,9 @@ class YoutubeDL(object):
 
     def to_screen(self, message, skip_eol=False):
         """Print message to stdout if not in quiet mode."""
-        if not self.params.get('quiet', False):
+        if self.params.get('logger'):
+            self.params['logger'].debug(message)
+        elif not self.params.get('quiet', False):
             terminator = [u'\n', u''][skip_eol]
             output = message + terminator
             write_string(output, self._screen_file)
@@ -211,10 +216,13 @@ class YoutubeDL(object):
     def to_stderr(self, message):
         """Print message to stderr."""
         assert type(message) == type(u'')
-        output = message + u'\n'
-        if 'b' in getattr(self._screen_file, 'mode', '') or sys.version_info[0] < 3: # Python 2 lies about the mode of sys.stdout/sys.stderr
-            output = output.encode(preferredencoding())
-        sys.stderr.write(output)
+        if self.params.get('logger'):
+            self.params['logger'].error(message)
+        else:
+            output = message + u'\n'
+            if 'b' in getattr(self._screen_file, 'mode', '') or sys.version_info[0] < 3: # Python 2 lies about the mode of sys.stdout/sys.stderr
+                output = output.encode(preferredencoding())
+            sys.stderr.write(output)
 
     def to_console_title(self, message):
         if not self.params.get('consoletitle', False):
@@ -250,10 +258,6 @@ class YoutubeDL(object):
         if self.params.get('cookiefile') is not None:
             self.cookiejar.save()
 
-    def fixed_template(self):
-        """Checks if the output template is fixed."""
-        return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
-
     def trouble(self, message=None, tb=None):
         """Determine action to take when a download problem appears.
 
@@ -369,15 +373,17 @@ class YoutubeDL(object):
     def _match_entry(self, info_dict):
         """ Returns None iff the file should be downloaded """
 
-        title = info_dict['title']
-        matchtitle = self.params.get('matchtitle', False)
-        if matchtitle:
-            if not re.search(matchtitle, title, re.IGNORECASE):
-                return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
-        rejecttitle = self.params.get('rejecttitle', False)
-        if rejecttitle:
-            if re.search(rejecttitle, title, re.IGNORECASE):
-                return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
+        if 'title' in info_dict:
+            # This can happen when we're just evaluating the playlist
+            title = info_dict['title']
+            matchtitle = self.params.get('matchtitle', False)
+            if matchtitle:
+                if not re.search(matchtitle, title, re.IGNORECASE):
+                    return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
+            rejecttitle = self.params.get('rejecttitle', False)
+            if rejecttitle:
+                if re.search(rejecttitle, title, re.IGNORECASE):
+                    return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
         date = info_dict.get('upload_date', None)
         if date is not None:
             dateRange = self.params.get('daterange', DateRange())
@@ -388,8 +394,8 @@ class YoutubeDL(object):
             if age_limit < info_dict.get('age_limit', 0):
                 return u'Skipping "' + title + '" because it is age restricted'
         if self.in_download_archive(info_dict):
-            return (u'%(title)s has already been recorded in archive'
-                    % info_dict)
+            return (u'%s has already been recorded in archive'
+                    % info_dict.get('title', info_dict.get('id', u'video')))
         return None
 
     @staticmethod
@@ -468,7 +474,7 @@ class YoutubeDL(object):
                                      ie_key=ie_result.get('ie_key'),
                                      extra_info=extra_info)
         elif result_type == 'playlist':
-            self.add_extra_info(ie_result, extra_info)
+
             # 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)
@@ -498,6 +504,12 @@ class YoutubeDL(object):
                     'webpage_url': ie_result['webpage_url'],
                     'extractor_key': ie_result['extractor_key'],
                 }
+
+                reason = self._match_entry(entry)
+                if reason is not None:
+                    self.to_screen(u'[download] ' + reason)
+                    continue
+
                 entry_result = self.process_ie_result(entry,
                                                       download=download,
                                                       extra_info=extra)
@@ -653,7 +665,7 @@ class YoutubeDL(object):
 
         # Forced printings
         if self.params.get('forcetitle', False):
-            compat_print(info_dict['title'])
+            compat_print(info_dict['fulltitle'])
         if self.params.get('forceid', False):
             compat_print(info_dict['id'])
         if self.params.get('forceurl', False):
@@ -782,7 +794,9 @@ class YoutubeDL(object):
 
     def download(self, url_list):
         """Download a given list of URLs."""
-        if len(url_list) > 1 and self.fixed_template():
+        if (len(url_list) > 1 and
+                '%' not in self.params['outtmpl']
+                and self.params.get('max_downloads') != 1):
             raise SameFileError(self.params['outtmpl'])
 
         for url in url_list:
@@ -820,11 +834,26 @@ class YoutubeDL(object):
             except (IOError, OSError):
                 self.report_warning(u'Unable to remove downloaded video file')
 
+    def _make_archive_id(self, info_dict):
+        # Future-proof against any change in case
+        # and backwards compatibility with prior versions
+        extractor = info_dict.get('extractor_key')
+        if extractor is None:
+            if 'id' in info_dict:
+                extractor = info_dict.get('ie_key')  # key in a playlist
+        if extractor is None:
+            return None  # Incomplete video information
+        return extractor.lower() + u' ' + info_dict['id']
+
     def in_download_archive(self, info_dict):
         fn = self.params.get('download_archive')
         if fn is None:
             return False
-        vid_id = info_dict['extractor'] + u' ' + info_dict['id']
+
+        vid_id = self._make_archive_id(info_dict)
+        if vid_id is None:
+            return False  # Incomplete video information
+
         try:
             with locked_file(fn, 'r', encoding='utf-8') as archive_file:
                 for line in archive_file:
@@ -839,12 +868,15 @@ class YoutubeDL(object):
         fn = self.params.get('download_archive')
         if fn is None:
             return
-        vid_id = info_dict['extractor'] + u' ' + info_dict['id']
+        vid_id = self._make_archive_id(info_dict)
+        assert vid_id
         with locked_file(fn, 'a', encoding='utf-8') as archive_file:
             archive_file.write(vid_id + u'\n')
 
     @staticmethod
     def format_resolution(format, default='unknown'):
+        if format.get('vcodec') == 'none':
+            return 'audio only'
         if format.get('_resolution') is not None:
             return format['_resolution']
         if format.get('height') is not None:
@@ -858,10 +890,11 @@ class YoutubeDL(object):
 
     def list_formats(self, info_dict):
         def format_note(fdict):
-            if fdict.get('format_note') is not None:
-                return fdict['format_note']
             res = u''
-            if fdict.get('vcodec') is not None:
+            if fdict.get('format_note') is not None:
+                res += fdict['format_note'] + u' '
+            if (fdict.get('vcodec') is not None and
+                    fdict.get('vcodec') != 'none'):
                 res += u'%-5s' % fdict['vcodec']
             elif fdict.get('vbr') is not None:
                 res += u'video'
@@ -877,26 +910,31 @@ class YoutubeDL(object):
                 res += 'audio'
             if fdict.get('abr') is not None:
                 res += u'@%3dk' % fdict['abr']
+            if fdict.get('filesize') is not None:
+                if res:
+                    res += u', '
+                res += format_bytes(fdict['filesize'])
             return res
 
-        def line(format):
-            return (u'%-20s%-10s%-12s%s' % (
+        def line(format, idlen=20):
+            return ((u'%-' + compat_str(idlen + 1) + u's%-10s%-12s%s') % (
                 format['format_id'],
                 format['ext'],
                 self.format_resolution(format),
                 format_note(format),
-                )
-            )
+            ))
 
         formats = info_dict.get('formats', [info_dict])
-        formats_s = list(map(line, formats))
+        idlen = max(len(u'format code'),
+                    max(len(f['format_id']) for f in formats))
+        formats_s = [line(f, idlen) for f in formats]
         if len(formats) > 1:
             formats_s[0] += (' ' if format_note(formats[0]) else '') + '(worst)'
             formats_s[-1] += (' ' if format_note(formats[-1]) else '') + '(best)'
 
         header_line = line({
             'format_id': u'format code', 'ext': u'extension',
-            '_resolution': u'resolution', 'format_note': u'note'})
+            '_resolution': u'resolution', 'format_note': u'note'}, idlen=idlen)
         self.to_screen(u'[info] Available formats for %s:\n%s\n%s' %
                        (info_dict['id'], header_line, u"\n".join(formats_s)))
 
@@ -931,7 +969,7 @@ class YoutubeDL(object):
                 proxy_map.update(handler.proxies)
         write_string(u'[debug] Proxy map: ' + compat_str(proxy_map) + u'\n')
 
-    def _setup_opener(self, timeout=300):
+    def _setup_opener(self, timeout=20):
         opts_cookiefile = self.params.get('cookiefile')
         opts_proxy = self.params.get('proxy')