Option '--load-info': if the download fails, try extracting the info with the 'webpag...
[youtube-dl] / youtube_dl / YoutubeDL.py
index 87eb1a0b37fd5342b9a50d55109a2efe29a8c2fb..77339dddfd591f2a02bdba3e91414142ec9c1d7c 100644 (file)
@@ -33,6 +33,7 @@ from .utils import (
     DownloadError,
     encodeFilename,
     ExtractorError,
+    format_bytes,
     locked_file,
     make_HTTPS_handler,
     MaxDownloadsReached,
@@ -125,12 +126,13 @@ 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
     proxy:             URL of the proxy server to use
+    socket_timeout:    Time to wait for unresponsive hosts, in seconds
 
     The following parameters are not used by YoutubeDL itself, they are used by
     the FileDownloader:
@@ -145,7 +147,7 @@ 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 = {}
@@ -154,6 +156,7 @@ 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']
@@ -163,12 +166,11 @@ 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()
@@ -257,10 +259,6 @@ 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.
 
@@ -797,7 +795,9 @@ 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:
@@ -812,6 +812,21 @@ class YoutubeDL(object):
 
         return self._download_retcode
 
+    def download_with_info_file(self, info_filename):
+        with open(info_filename, 'r') as f:
+            # TODO: Check for errors
+            info = json.load(f)
+        try:
+            self.process_ie_result(info, download=True)
+        except DownloadError:
+            webpage_url = info.get('webpage_url')
+            if webpage_url is not None:
+                self.report_warning(u'The info failed to download, trying with "%s"' % webpage_url)
+                return self.download([webpage_url])
+            else:
+                raise
+        return self._download_retcode
+
     def post_process(self, filename, ie_info):
         """Run all the postprocessors on the given file."""
         info = dict(ie_info)
@@ -835,20 +850,26 @@ class YoutubeDL(object):
             except (IOError, OSError):
                 self.report_warning(u'Unable to remove downloaded video file')
 
-    def in_download_archive(self, info_dict):
-        fn = self.params.get('download_archive')
-        if fn is None:
-            return False
-        extractor = info_dict.get('extractor_id')
+    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 = self._make_archive_id(info_dict)
+        if vid_id is None:
             return False  # Incomplete video information
-        # Future-proof against any change in case
-        # and backwards compatibility with prior versions
-        extractor = extractor.lower()
-        vid_id = extractor + u' ' + info_dict['id']
+
         try:
             with locked_file(fn, 'r', encoding='utf-8') as archive_file:
                 for line in archive_file:
@@ -863,12 +884,15 @@ 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:
@@ -882,10 +906,11 @@ 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'
@@ -901,26 +926,31 @@ 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)))
 
@@ -955,7 +985,10 @@ 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_val = self.params.get('socket_timeout')
+        timeout = 600 if timeout_val is None else float(timeout_val)
+
         opts_cookiefile = self.params.get('cookiefile')
         opts_proxy = self.params.get('proxy')