_ Git - youtube-dl/blob - youtube_dl/YoutubeDL.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3
   4 from __future__ import absolute_import, unicode_literals
   5
   6 import collections
   7 import contextlib
   8 import datetime
   9 import errno
  10 import fileinput
  11 import io
  12 import itertools
  13 import json
  14 import locale
  15 import operator
  16 import os
  17 import platform
  18 import re
  19 import shutil
  20 import subprocess
  21 import socket
  22 import sys
  23 import time
  24 import traceback
  25
  26 if os.name == 'nt':
  27     import ctypes
  28
  29 from .compat import (
  30     compat_basestring,
  31     compat_cookiejar,
  32     compat_expanduser,
  33     compat_get_terminal_size,
  34     compat_http_client,
  35     compat_kwargs,
  36     compat_str,
  37     compat_urllib_error,
  38     compat_urllib_request,
  39 )
  40 from .utils import (
  41     escape_url,
  42     ContentTooShortError,
  43     date_from_str,
  44     DateRange,
  45     DEFAULT_OUTTMPL,
  46     determine_ext,
  47     DownloadError,
  48     encodeFilename,
  49     ExtractorError,
  50     format_bytes,
  51     formatSeconds,
  52     locked_file,
  53     make_HTTPS_handler,
  54     MaxDownloadsReached,
  55     PagedList,
  56     parse_filesize,
  57     PerRequestProxyHandler,
  58     PostProcessingError,
  59     platform_name,
  60     preferredencoding,
  61     render_table,
  62     SameFileError,
  63     sanitize_filename,
  64     sanitize_path,
  65     std_headers,
  66     subtitles_filename,
  67     takewhile_inclusive,
  68     UnavailableVideoError,
  69     url_basename,
  70     version_tuple,
  71     write_json_file,
  72     write_string,
  73     YoutubeDLHandler,
  74     prepend_extension,
  75     args_to_str,
  76     age_restricted,
  77 )
  78 from .cache import Cache
  79 from .extractor import get_info_extractor, gen_extractors
  80 from .downloader import get_suitable_downloader
  81 from .downloader.rtmp import rtmpdump_version
  82 from .postprocessor import (
  83     FFmpegFixupM4aPP,
  84     FFmpegFixupStretchedPP,
  85     FFmpegMergerPP,
  86     FFmpegPostProcessor,
  87     get_postprocessor,
  88 )
  89 from .version import __version__
  90
  91
  92 class YoutubeDL(object):
  93     """YoutubeDL class.
  94
  95     YoutubeDL objects are the ones responsible of downloading the
  96     actual video file and writing it to disk if the user has requested
  97     it, among some other tasks. In most cases there should be one per
  98     program. As, given a video URL, the downloader doesn't know how to
  99     extract all the needed information, task that InfoExtractors do, it
 100     has to pass the URL to one of them.
 101
 102     For this, YoutubeDL objects have a method that allows
 103     InfoExtractors to be registered in a given order. When it is passed
 104     a URL, the YoutubeDL object handles it to the first InfoExtractor it
 105     finds that reports being able to handle it. The InfoExtractor extracts
 106     all the information about the video or videos the URL refers to, and
 107     YoutubeDL process the extracted information, possibly using a File
 108     Downloader to download the video.
 109
 110     YoutubeDL objects accept a lot of parameters. In order not to saturate
 111     the object constructor with arguments, it receives a dictionary of
 112     options instead. These options are available through the params
 113     attribute for the InfoExtractors to use. The YoutubeDL also
 114     registers itself as the downloader in charge for the InfoExtractors
 115     that are added to it, so this is a "mutual registration".
 116
 117     Available options:
 118
 119     username:          Username for authentication purposes.
 120     password:          Password for authentication purposes.
 121     videopassword:     Password for acces a video.
 122     usenetrc:          Use netrc for authentication instead.
 123     verbose:           Print additional info to stdout.
 124     quiet:             Do not print messages to stdout.
 125     no_warnings:       Do not print out anything for warnings.
 126     forceurl:          Force printing final URL.
 127     forcetitle:        Force printing title.
 128     forceid:           Force printing ID.
 129     forcethumbnail:    Force printing thumbnail URL.
 130     forcedescription:  Force printing description.
 131     forcefilename:     Force printing final filename.
 132     forceduration:     Force printing duration.
 133     forcejson:         Force printing info_dict as JSON.
 134     dump_single_json:  Force printing the info_dict of the whole playlist
 135                        (or video) as a single JSON line.
 136     simulate:          Do not download the video files.
 137     format:            Video format code. See options.py for more information.
 138     format_limit:      Highest quality format to try.
 139     outtmpl:           Template for output names.
 140     restrictfilenames: Do not allow "&" and spaces in file names
 141     ignoreerrors:      Do not stop on download errors.
 142     nooverwrites:      Prevent overwriting files.
 143     playliststart:     Playlist item to start at.
 144     playlistend:       Playlist item to end at.
 145     playlist_items:    Specific indices of playlist to download.
 146     playlistreverse:   Download playlist items in reverse order.
 147     matchtitle:        Download only matching titles.
 148     rejecttitle:       Reject downloads for matching titles.
 149     logger:            Log messages to a logging.Logger instance.
 150     logtostderr:       Log messages to stderr instead of stdout.
 151     writedescription:  Write the video description to a .description file
 152     writeinfojson:     Write the video description to a .info.json file
 153     writeannotations:  Write the video annotations to a .annotations.xml file
 154     writethumbnail:    Write the thumbnail image to a file
 155     write_all_thumbnails:  Write all thumbnail formats to files
 156     writesubtitles:    Write the video subtitles to a file
 157     writeautomaticsub: Write the automatic subtitles to a file
 158     allsubtitles:      Downloads all the subtitles of the video
 159                        (requires writesubtitles or writeautomaticsub)
 160     listsubtitles:     Lists all available subtitles for the video
 161     subtitlesformat:   The format code for subtitles
 162     subtitleslangs:    List of languages of the subtitles to download
 163     keepvideo:         Keep the video file after post-processing
 164     daterange:         A DateRange object, download only if the upload_date is in the range.
 165     skip_download:     Skip the actual download of the video file
 166     cachedir:          Location of the cache files in the filesystem.
 167                        False to disable filesystem cache.
 168     noplaylist:        Download single video instead of a playlist if in doubt.
 169     age_limit:         An integer representing the user's age in years.
 170                        Unsuitable videos for the given age are skipped.
 171     min_views:         An integer representing the minimum view count the video
 172                        must have in order to not be skipped.
 173                        Videos without view count information are always
 174                        downloaded. None for no limit.
 175     max_views:         An integer representing the maximum view count.
 176                        Videos that are more popular than that are not
 177                        downloaded.
 178                        Videos without view count information are always
 179                        downloaded. None for no limit.
 180     download_archive:  File name of a file where all downloads are recorded.
 181                        Videos already present in the file are not downloaded
 182                        again.
 183     cookiefile:        File name where cookies should be read from and dumped to.
 184     nocheckcertificate:Do not verify SSL certificates
 185     prefer_insecure:   Use HTTP instead of HTTPS to retrieve information.
 186                        At the moment, this is only supported by YouTube.
 187     proxy:             URL of the proxy server to use
 188     cn_verification_proxy:  URL of the proxy to use for IP address verification
 189                        on Chinese sites. (Experimental)
 190     socket_timeout:    Time to wait for unresponsive hosts, in seconds
 191     bidi_workaround:   Work around buggy terminals without bidirectional text
 192                        support, using fridibi
 193     debug_printtraffic:Print out sent and received HTTP traffic
 194     include_ads:       Download ads as well
 195     default_search:    Prepend this string if an input url is not valid.
 196                        'auto' for elaborate guessing
 197     encoding:          Use this encoding instead of the system-specified.
 198     extract_flat:      Do not resolve URLs, return the immediate result.
 199                        Pass in 'in_playlist' to only show this behavior for
 200                        playlist items.
 201     postprocessors:    A list of dictionaries, each with an entry
 202                        * key:  The name of the postprocessor. See
 203                                youtube_dl/postprocessor/__init__.py for a list.
 204                        as well as any further keyword arguments for the
 205                        postprocessor.
 206     progress_hooks:    A list of functions that get called on download
 207                        progress, with a dictionary with the entries
 208                        * status: One of "downloading", "error", or "finished".
 209                                  Check this first and ignore unknown values.
 210
 211                        If status is one of "downloading", or "finished", the
 212                        following properties may also be present:
 213                        * filename: The final filename (always present)
 214                        * tmpfilename: The filename we're currently writing to
 215                        * downloaded_bytes: Bytes on disk
 216                        * total_bytes: Size of the whole file, None if unknown
 217                        * total_bytes_estimate: Guess of the eventual file size,
 218                                                None if unavailable.
 219                        * elapsed: The number of seconds since download started.
 220                        * eta: The estimated time in seconds, None if unknown
 221                        * speed: The download speed in bytes/second, None if
 222                                 unknown
 223                        * fragment_index: The counter of the currently
 224                                          downloaded video fragment.
 225                        * fragment_count: The number of fragments (= individual
 226                                          files that will be merged)
 227
 228                        Progress hooks are guaranteed to be called at least once
 229                        (with status "finished") if the download is successful.
 230     merge_output_format: Extension to use when merging formats.
 231     fixup:             Automatically correct known faults of the file.
 232                        One of:
 233                        - "never": do nothing
 234                        - "warn": only emit a warning
 235                        - "detect_or_warn": check whether we can do anything
 236                                            about it, warn otherwise (default)
 237     source_address:    (Experimental) Client-side IP address to bind to.
 238     call_home:         Boolean, true iff we are allowed to contact the
 239                        youtube-dl servers for debugging.
 240     sleep_interval:    Number of seconds to sleep before each download.
 241     listformats:       Print an overview of available video formats and exit.
 242     list_thumbnails:   Print a table of all thumbnails and exit.
 243     match_filter:      A function that gets called with the info_dict of
 244                        every video.
 245                        If it returns a message, the video is ignored.
 246                        If it returns None, the video is downloaded.
 247                        match_filter_func in utils.py is one example for this.
 248     no_color:          Do not emit color codes in output.
 249
 250     The following options determine which downloader is picked:
 251     external_downloader: Executable of the external downloader to call.
 252                        None or unset for standard (built-in) downloader.
 253     hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
 254
 255     The following parameters are not used by YoutubeDL itself, they are used by
 256     the downloader (see youtube_dl/downloader/common.py):
 257     nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
 258     noresizebuffer, retries, continuedl, noprogress, consoletitle,
 259     xattr_set_filesize, external_downloader_args.
 260
 261     The following options are used by the post processors:
 262     prefer_ffmpeg:     If True, use ffmpeg instead of avconv if both are available,
 263                        otherwise prefer avconv.
 264     exec_cmd:          Arbitrary command to run after downloading
 265     """
 266
 267     params = None
 268     _ies = []
 269     _pps = []
 270     _download_retcode = None
 271     _num_downloads = None
 272     _screen_file = None
 273
 274     def __init__(self, params=None, auto_init=True):
 275         """Create a FileDownloader object with the given options."""
 276         if params is None:
 277             params = {}
 278         self._ies = []
 279         self._ies_instances = {}
 280         self._pps = []
 281         self._progress_hooks = []
 282         self._download_retcode = 0
 283         self._num_downloads = 0
 284         self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
 285         self._err_file = sys.stderr
 286         self.params = params
 287         self.cache = Cache(self)
 288
 289         if params.get('bidi_workaround', False):
 290             try:
 291                 import pty
 292                 master, slave = pty.openpty()
 293                 width = compat_get_terminal_size().columns
 294                 if width is None:
 295                     width_args = []
 296                 else:
 297                     width_args = ['-w', str(width)]
 298                 sp_kwargs = dict(
 299                     stdin=subprocess.PIPE,
 300                     stdout=slave,
 301                     stderr=self._err_file)
 302                 try:
 303                     self._output_process = subprocess.Popen(
 304                         ['bidiv'] + width_args, **sp_kwargs
 305                     )
 306                 except OSError:
 307                     self._output_process = subprocess.Popen(
 308                         ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
 309                 self._output_channel = os.fdopen(master, 'rb')
 310             except OSError as ose:
 311                 if ose.errno == 2:
 312                     self.report_warning('Could not find fribidi executable, ignoring --bidi-workaround . Make sure that  fribidi  is an executable file in one of the directories in your $PATH.')
 313                 else:
 314                     raise
 315
 316         if (sys.version_info >= (3,) and sys.platform != 'win32' and
 317                 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
 318                 not params.get('restrictfilenames', False)):
 319             # On Python 3, the Unicode filesystem API will throw errors (#1474)
 320             self.report_warning(
 321                 'Assuming --restrict-filenames since file system encoding '
 322                 'cannot encode all characters. '
 323                 'Set the LC_ALL environment variable to fix this.')
 324             self.params['restrictfilenames'] = True
 325
 326         if '%(stitle)s' in self.params.get('outtmpl', ''):
 327             self.report_warning('%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
 328
 329         self._setup_opener()
 330
 331         if auto_init:
 332             self.print_debug_header()
 333             self.add_default_info_extractors()
 334
 335         for pp_def_raw in self.params.get('postprocessors', []):
 336             pp_class = get_postprocessor(pp_def_raw['key'])
 337             pp_def = dict(pp_def_raw)
 338             del pp_def['key']
 339             pp = pp_class(self, **compat_kwargs(pp_def))
 340             self.add_post_processor(pp)
 341
 342         for ph in self.params.get('progress_hooks', []):
 343             self.add_progress_hook(ph)
 344
 345     def warn_if_short_id(self, argv):
 346         # short YouTube ID starting with dash?
 347         idxs = [
 348             i for i, a in enumerate(argv)
 349             if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
 350         if idxs:
 351             correct_argv = (
 352                 ['youtube-dl'] +
 353                 [a for i, a in enumerate(argv) if i not in idxs] +
 354                 ['--'] + [argv[i] for i in idxs]
 355             )
 356             self.report_warning(
 357                 'Long argument string detected. '
 358                 'Use -- to separate parameters and URLs, like this:\n%s\n' %
 359                 args_to_str(correct_argv))
 360
 361     def add_info_extractor(self, ie):
 362         """Add an InfoExtractor object to the end of the list."""
 363         self._ies.append(ie)
 364         self._ies_instances[ie.ie_key()] = ie
 365         ie.set_downloader(self)
 366
 367     def get_info_extractor(self, ie_key):
 368         """
 369         Get an instance of an IE with name ie_key, it will try to get one from
 370         the _ies list, if there's no instance it will create a new one and add
 371         it to the extractor list.
 372         """
 373         ie = self._ies_instances.get(ie_key)
 374         if ie is None:
 375             ie = get_info_extractor(ie_key)()
 376             self.add_info_extractor(ie)
 377         return ie
 378
 379     def add_default_info_extractors(self):
 380         """
 381         Add the InfoExtractors returned by gen_extractors to the end of the list
 382         """
 383         for ie in gen_extractors():
 384             self.add_info_extractor(ie)
 385
 386     def add_post_processor(self, pp):
 387         """Add a PostProcessor object to the end of the chain."""
 388         self._pps.append(pp)
 389         pp.set_downloader(self)
 390
 391     def add_progress_hook(self, ph):
 392         """Add the progress hook (currently only for the file downloader)"""
 393         self._progress_hooks.append(ph)
 394
 395     def _bidi_workaround(self, message):
 396         if not hasattr(self, '_output_channel'):
 397             return message
 398
 399         assert hasattr(self, '_output_process')
 400         assert isinstance(message, compat_str)
 401         line_count = message.count('\n') + 1
 402         self._output_process.stdin.write((message + '\n').encode('utf-8'))
 403         self._output_process.stdin.flush()
 404         res = ''.join(self._output_channel.readline().decode('utf-8')
 405                       for _ in range(line_count))
 406         return res[:-len('\n')]
 407
 408     def to_screen(self, message, skip_eol=False):
 409         """Print message to stdout if not in quiet mode."""
 410         return self.to_stdout(message, skip_eol, check_quiet=True)
 411
 412     def _write_string(self, s, out=None):
 413         write_string(s, out=out, encoding=self.params.get('encoding'))
 414
 415     def to_stdout(self, message, skip_eol=False, check_quiet=False):
 416         """Print message to stdout if not in quiet mode."""
 417         if self.params.get('logger'):
 418             self.params['logger'].debug(message)
 419         elif not check_quiet or not self.params.get('quiet', False):
 420             message = self._bidi_workaround(message)
 421             terminator = ['\n', ''][skip_eol]
 422             output = message + terminator
 423
 424             self._write_string(output, self._screen_file)
 425
 426     def to_stderr(self, message):
 427         """Print message to stderr."""
 428         assert isinstance(message, compat_str)
 429         if self.params.get('logger'):
 430             self.params['logger'].error(message)
 431         else:
 432             message = self._bidi_workaround(message)
 433             output = message + '\n'
 434             self._write_string(output, self._err_file)
 435
 436     def to_console_title(self, message):
 437         if not self.params.get('consoletitle', False):
 438             return
 439         if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
 440             # c_wchar_p() might not be necessary if `message` is
 441             # already of type unicode()
 442             ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
 443         elif 'TERM' in os.environ:
 444             self._write_string('\033]0;%s\007' % message, self._screen_file)
 445
 446     def save_console_title(self):
 447         if not self.params.get('consoletitle', False):
 448             return
 449         if 'TERM' in os.environ:
 450             # Save the title on stack
 451             self._write_string('\033[22;0t', self._screen_file)
 452
 453     def restore_console_title(self):
 454         if not self.params.get('consoletitle', False):
 455             return
 456         if 'TERM' in os.environ:
 457             # Restore the title from stack
 458             self._write_string('\033[23;0t', self._screen_file)
 459
 460     def __enter__(self):
 461         self.save_console_title()
 462         return self
 463
 464     def __exit__(self, *args):
 465         self.restore_console_title()
 466
 467         if self.params.get('cookiefile') is not None:
 468             self.cookiejar.save()
 469
 470     def trouble(self, message=None, tb=None):
 471         """Determine action to take when a download problem appears.
 472
 473         Depending on if the downloader has been configured to ignore
 474         download errors or not, this method may throw an exception or
 475         not when errors are found, after printing the message.
 476
 477         tb, if given, is additional traceback information.
 478         """
 479         if message is not None:
 480             self.to_stderr(message)
 481         if self.params.get('verbose'):
 482             if tb is None:
 483                 if sys.exc_info()[0]:  # if .trouble has been called from an except block
 484                     tb = ''
 485                     if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
 486                         tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
 487                     tb += compat_str(traceback.format_exc())
 488                 else:
 489                     tb_data = traceback.format_list(traceback.extract_stack())
 490                     tb = ''.join(tb_data)
 491             self.to_stderr(tb)
 492         if not self.params.get('ignoreerrors', False):
 493             if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
 494                 exc_info = sys.exc_info()[1].exc_info
 495             else:
 496                 exc_info = sys.exc_info()
 497             raise DownloadError(message, exc_info)
 498         self._download_retcode = 1
 499
 500     def report_warning(self, message):
 501         '''
 502         Print the message to stderr, it will be prefixed with 'WARNING:'
 503         If stderr is a tty file the 'WARNING:' will be colored
 504         '''
 505         if self.params.get('logger') is not None:
 506             self.params['logger'].warning(message)
 507         else:
 508             if self.params.get('no_warnings'):
 509                 return
 510             if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
 511                 _msg_header = '\033[0;33mWARNING:\033[0m'
 512             else:
 513                 _msg_header = 'WARNING:'
 514             warning_message = '%s %s' % (_msg_header, message)
 515             self.to_stderr(warning_message)
 516
 517     def report_error(self, message, tb=None):
 518         '''
 519         Do the same as trouble, but prefixes the message with 'ERROR:', colored
 520         in red if stderr is a tty file.
 521         '''
 522         if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
 523             _msg_header = '\033[0;31mERROR:\033[0m'
 524         else:
 525             _msg_header = 'ERROR:'
 526         error_message = '%s %s' % (_msg_header, message)
 527         self.trouble(error_message, tb)
 528
 529     def report_file_already_downloaded(self, file_name):
 530         """Report file has already been fully downloaded."""
 531         try:
 532             self.to_screen('[download] %s has already been downloaded' % file_name)
 533         except UnicodeEncodeError:
 534             self.to_screen('[download] The file has already been downloaded')
 535
 536     def prepare_filename(self, info_dict):
 537         """Generate the output filename."""
 538         try:
 539             template_dict = dict(info_dict)
 540
 541             template_dict['epoch'] = int(time.time())
 542             autonumber_size = self.params.get('autonumber_size')
 543             if autonumber_size is None:
 544                 autonumber_size = 5
 545             autonumber_templ = '%0' + str(autonumber_size) + 'd'
 546             template_dict['autonumber'] = autonumber_templ % self._num_downloads
 547             if template_dict.get('playlist_index') is not None:
 548                 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
 549             if template_dict.get('resolution') is None:
 550                 if template_dict.get('width') and template_dict.get('height'):
 551                     template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
 552                 elif template_dict.get('height'):
 553                     template_dict['resolution'] = '%sp' % template_dict['height']
 554                 elif template_dict.get('width'):
 555                     template_dict['resolution'] = '?x%d' % template_dict['width']
 556
 557             sanitize = lambda k, v: sanitize_filename(
 558                 compat_str(v),
 559                 restricted=self.params.get('restrictfilenames'),
 560                 is_id=(k == 'id'))
 561             template_dict = dict((k, sanitize(k, v))
 562                                  for k, v in template_dict.items()
 563                                  if v is not None)
 564             template_dict = collections.defaultdict(lambda: 'NA', template_dict)
 565
 566             outtmpl = sanitize_path(self.params.get('outtmpl', DEFAULT_OUTTMPL))
 567             tmpl = compat_expanduser(outtmpl)
 568             filename = tmpl % template_dict
 569             # Temporary fix for #4787
 570             # 'Treat' all problem characters by passing filename through preferredencoding
 571             # to workaround encoding issues with subprocess on python2 @ Windows
 572             if sys.version_info < (3, 0) and sys.platform == 'win32':
 573                 filename = encodeFilename(filename, True).decode(preferredencoding())
 574             return filename
 575         except ValueError as err:
 576             self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
 577             return None
 578
 579     def _match_entry(self, info_dict, incomplete):
 580         """ Returns None iff the file should be downloaded """
 581
 582         video_title = info_dict.get('title', info_dict.get('id', 'video'))
 583         if 'title' in info_dict:
 584             # This can happen when we're just evaluating the playlist
 585             title = info_dict['title']
 586             matchtitle = self.params.get('matchtitle', False)
 587             if matchtitle:
 588                 if not re.search(matchtitle, title, re.IGNORECASE):
 589                     return '"' + title + '" title did not match pattern "' + matchtitle + '"'
 590             rejecttitle = self.params.get('rejecttitle', False)
 591             if rejecttitle:
 592                 if re.search(rejecttitle, title, re.IGNORECASE):
 593                     return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
 594         date = info_dict.get('upload_date', None)
 595         if date is not None:
 596             dateRange = self.params.get('daterange', DateRange())
 597             if date not in dateRange:
 598                 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
 599         view_count = info_dict.get('view_count', None)
 600         if view_count is not None:
 601             min_views = self.params.get('min_views')
 602             if min_views is not None and view_count < min_views:
 603                 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
 604             max_views = self.params.get('max_views')
 605             if max_views is not None and view_count > max_views:
 606                 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
 607         if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
 608             return 'Skipping "%s" because it is age restricted' % video_title
 609         if self.in_download_archive(info_dict):
 610             return '%s has already been recorded in archive' % video_title
 611
 612         if not incomplete:
 613             match_filter = self.params.get('match_filter')
 614             if match_filter is not None:
 615                 ret = match_filter(info_dict)
 616                 if ret is not None:
 617                     return ret
 618
 619         return None
 620
 621     @staticmethod
 622     def add_extra_info(info_dict, extra_info):
 623         '''Set the keys from extra_info in info dict if they are missing'''
 624         for key, value in extra_info.items():
 625             info_dict.setdefault(key, value)
 626
 627     def extract_info(self, url, download=True, ie_key=None, extra_info={},
 628                      process=True):
 629         '''
 630         Returns a list with a dictionary for each video we find.
 631         If 'download', also downloads the videos.
 632         extra_info is a dict containing the extra values to add to each result
 633          '''
 634
 635         if ie_key:
 636             ies = [self.get_info_extractor(ie_key)]
 637         else:
 638             ies = self._ies
 639
 640         for ie in ies:
 641             if not ie.suitable(url):
 642                 continue
 643
 644             if not ie.working():
 645                 self.report_warning('The program functionality for this site has been marked as broken, '
 646                                     'and will probably not work.')
 647
 648             try:
 649                 ie_result = ie.extract(url)
 650                 if ie_result is None:  # Finished already (backwards compatibility; listformats and friends should be moved here)
 651                     break
 652                 if isinstance(ie_result, list):
 653                     # Backwards compatibility: old IE result format
 654                     ie_result = {
 655                         '_type': 'compat_list',
 656                         'entries': ie_result,
 657                     }
 658                 self.add_default_extra_info(ie_result, ie, url)
 659                 if process:
 660                     return self.process_ie_result(ie_result, download, extra_info)
 661                 else:
 662                     return ie_result
 663             except ExtractorError as de:  # An error we somewhat expected
 664                 self.report_error(compat_str(de), de.format_traceback())
 665                 break
 666             except MaxDownloadsReached:
 667                 raise
 668             except Exception as e:
 669                 if self.params.get('ignoreerrors', False):
 670                     self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
 671                     break
 672                 else:
 673                     raise
 674         else:
 675             self.report_error('no suitable InfoExtractor for URL %s' % url)
 676
 677     def add_default_extra_info(self, ie_result, ie, url):
 678         self.add_extra_info(ie_result, {
 679             'extractor': ie.IE_NAME,
 680             'webpage_url': url,
 681             'webpage_url_basename': url_basename(url),
 682             'extractor_key': ie.ie_key(),
 683         })
 684
 685     def process_ie_result(self, ie_result, download=True, extra_info={}):
 686         """
 687         Take the result of the ie(may be modified) and resolve all unresolved
 688         references (URLs, playlist items).
 689
 690         It will also download the videos if 'download'.
 691         Returns the resolved ie_result.
 692         """
 693
 694         result_type = ie_result.get('_type', 'video')
 695
 696         if result_type in ('url', 'url_transparent'):
 697             extract_flat = self.params.get('extract_flat', False)
 698             if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
 699                     extract_flat is True):
 700                 if self.params.get('forcejson', False):
 701                     self.to_stdout(json.dumps(ie_result))
 702                 return ie_result
 703
 704         if result_type == 'video':
 705             self.add_extra_info(ie_result, extra_info)
 706             return self.process_video_result(ie_result, download=download)
 707         elif result_type == 'url':
 708             # We have to add extra_info to the results because it may be
 709             # contained in a playlist
 710             return self.extract_info(ie_result['url'],
 711                                      download,
 712                                      ie_key=ie_result.get('ie_key'),
 713                                      extra_info=extra_info)
 714         elif result_type == 'url_transparent':
 715             # Use the information from the embedding page
 716             info = self.extract_info(
 717                 ie_result['url'], ie_key=ie_result.get('ie_key'),
 718                 extra_info=extra_info, download=False, process=False)
 719
 720             force_properties = dict(
 721                 (k, v) for k, v in ie_result.items() if v is not None)
 722             for f in ('_type', 'url'):
 723                 if f in force_properties:
 724                     del force_properties[f]
 725             new_result = info.copy()
 726             new_result.update(force_properties)
 727
 728             assert new_result.get('_type') != 'url_transparent'
 729
 730             return self.process_ie_result(
 731                 new_result, download=download, extra_info=extra_info)
 732         elif result_type == 'playlist' or result_type == 'multi_video':
 733             # We process each entry in the playlist
 734             playlist = ie_result.get('title', None) or ie_result.get('id', None)
 735             self.to_screen('[download] Downloading playlist: %s' % playlist)
 736
 737             playlist_results = []
 738
 739             playliststart = self.params.get('playliststart', 1) - 1
 740             playlistend = self.params.get('playlistend', None)
 741             # For backwards compatibility, interpret -1 as whole list
 742             if playlistend == -1:
 743                 playlistend = None
 744
 745             playlistitems_str = self.params.get('playlist_items', None)
 746             playlistitems = None
 747             if playlistitems_str is not None:
 748                 def iter_playlistitems(format):
 749                     for string_segment in format.split(','):
 750                         if '-' in string_segment:
 751                             start, end = string_segment.split('-')
 752                             for item in range(int(start), int(end) + 1):
 753                                 yield int(item)
 754                         else:
 755                             yield int(string_segment)
 756                 playlistitems = iter_playlistitems(playlistitems_str)
 757
 758             ie_entries = ie_result['entries']
 759             if isinstance(ie_entries, list):
 760                 n_all_entries = len(ie_entries)
 761                 if playlistitems:
 762                     entries = [ie_entries[i - 1] for i in playlistitems]
 763                 else:
 764                     entries = ie_entries[playliststart:playlistend]
 765                 n_entries = len(entries)
 766                 self.to_screen(
 767                     "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
 768                     (ie_result['extractor'], playlist, n_all_entries, n_entries))
 769             elif isinstance(ie_entries, PagedList):
 770                 if playlistitems:
 771                     entries = []
 772                     for item in playlistitems:
 773                         entries.extend(ie_entries.getslice(
 774                             item - 1, item
 775                         ))
 776                 else:
 777                     entries = ie_entries.getslice(
 778                         playliststart, playlistend)
 779                 n_entries = len(entries)
 780                 self.to_screen(
 781                     "[%s] playlist %s: Downloading %d videos" %
 782                     (ie_result['extractor'], playlist, n_entries))
 783             else:  # iterable
 784                 if playlistitems:
 785                     entry_list = list(ie_entries)
 786                     entries = [entry_list[i - 1] for i in playlistitems]
 787                 else:
 788                     entries = list(itertools.islice(
 789                         ie_entries, playliststart, playlistend))
 790                 n_entries = len(entries)
 791                 self.to_screen(
 792                     "[%s] playlist %s: Downloading %d videos" %
 793                     (ie_result['extractor'], playlist, n_entries))
 794
 795             if self.params.get('playlistreverse', False):
 796                 entries = entries[::-1]
 797
 798             for i, entry in enumerate(entries, 1):
 799                 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
 800                 extra = {
 801                     'n_entries': n_entries,
 802                     'playlist': playlist,
 803                     'playlist_id': ie_result.get('id'),
 804                     'playlist_title': ie_result.get('title'),
 805                     'playlist_index': i + playliststart,
 806                     'extractor': ie_result['extractor'],
 807                     'webpage_url': ie_result['webpage_url'],
 808                     'webpage_url_basename': url_basename(ie_result['webpage_url']),
 809                     'extractor_key': ie_result['extractor_key'],
 810                 }
 811
 812                 reason = self._match_entry(entry, incomplete=True)
 813                 if reason is not None:
 814                     self.to_screen('[download] ' + reason)
 815                     continue
 816
 817                 entry_result = self.process_ie_result(entry,
 818                                                       download=download,
 819                                                       extra_info=extra)
 820                 playlist_results.append(entry_result)
 821             ie_result['entries'] = playlist_results
 822             return ie_result
 823         elif result_type == 'compat_list':
 824             self.report_warning(
 825                 'Extractor %s returned a compat_list result. '
 826                 'It needs to be updated.' % ie_result.get('extractor'))
 827
 828             def _fixup(r):
 829                 self.add_extra_info(
 830                     r,
 831                     {
 832                         'extractor': ie_result['extractor'],
 833                         'webpage_url': ie_result['webpage_url'],
 834                         'webpage_url_basename': url_basename(ie_result['webpage_url']),
 835                         'extractor_key': ie_result['extractor_key'],
 836                     }
 837                 )
 838                 return r
 839             ie_result['entries'] = [
 840                 self.process_ie_result(_fixup(r), download, extra_info)
 841                 for r in ie_result['entries']
 842             ]
 843             return ie_result
 844         else:
 845             raise Exception('Invalid result type: %s' % result_type)
 846
 847     def _apply_format_filter(self, format_spec, available_formats):
 848         " Returns a tuple of the remaining format_spec and filtered formats "
 849
 850         OPERATORS = {
 851             '<': operator.lt,
 852             '<=': operator.le,
 853             '>': operator.gt,
 854             '>=': operator.ge,
 855             '=': operator.eq,
 856             '!=': operator.ne,
 857         }
 858         operator_rex = re.compile(r'''(?x)\s*\[
 859             (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
 860             \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
 861             (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
 862             \]$
 863             ''' % '|'.join(map(re.escape, OPERATORS.keys())))
 864         m = operator_rex.search(format_spec)
 865         if m:
 866             try:
 867                 comparison_value = int(m.group('value'))
 868             except ValueError:
 869                 comparison_value = parse_filesize(m.group('value'))
 870                 if comparison_value is None:
 871                     comparison_value = parse_filesize(m.group('value') + 'B')
 872                 if comparison_value is None:
 873                     raise ValueError(
 874                         'Invalid value %r in format specification %r' % (
 875                             m.group('value'), format_spec))
 876             op = OPERATORS[m.group('op')]
 877
 878         if not m:
 879             STR_OPERATORS = {
 880                 '=': operator.eq,
 881                 '!=': operator.ne,
 882             }
 883             str_operator_rex = re.compile(r'''(?x)\s*\[
 884                 \s*(?P<key>ext|acodec|vcodec|container|protocol)
 885                 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
 886                 \s*(?P<value>[a-zA-Z0-9_-]+)
 887                 \s*\]$
 888                 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
 889             m = str_operator_rex.search(format_spec)
 890             if m:
 891                 comparison_value = m.group('value')
 892                 op = STR_OPERATORS[m.group('op')]
 893
 894         if not m:
 895             raise ValueError('Invalid format specification %r' % format_spec)
 896
 897         def _filter(f):
 898             actual_value = f.get(m.group('key'))
 899             if actual_value is None:
 900                 return m.group('none_inclusive')
 901             return op(actual_value, comparison_value)
 902         new_formats = [f for f in available_formats if _filter(f)]
 903
 904         new_format_spec = format_spec[:-len(m.group(0))]
 905         if not new_format_spec:
 906             new_format_spec = 'best'
 907
 908         return (new_format_spec, new_formats)
 909
 910     def select_format(self, format_spec, available_formats):
 911         while format_spec.endswith(']'):
 912             format_spec, available_formats = self._apply_format_filter(
 913                 format_spec, available_formats)
 914         if not available_formats:
 915             return None
 916
 917         if format_spec == 'best' or format_spec is None:
 918             return available_formats[-1]
 919         elif format_spec == 'worst':
 920             return available_formats[0]
 921         elif format_spec == 'bestaudio':
 922             audio_formats = [
 923                 f for f in available_formats
 924                 if f.get('vcodec') == 'none']
 925             if audio_formats:
 926                 return audio_formats[-1]
 927         elif format_spec == 'worstaudio':
 928             audio_formats = [
 929                 f for f in available_formats
 930                 if f.get('vcodec') == 'none']
 931             if audio_formats:
 932                 return audio_formats[0]
 933         elif format_spec == 'bestvideo':
 934             video_formats = [
 935                 f for f in available_formats
 936                 if f.get('acodec') == 'none']
 937             if video_formats:
 938                 return video_formats[-1]
 939         elif format_spec == 'worstvideo':
 940             video_formats = [
 941                 f for f in available_formats
 942                 if f.get('acodec') == 'none']
 943             if video_formats:
 944                 return video_formats[0]
 945         else:
 946             extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
 947             if format_spec in extensions:
 948                 filter_f = lambda f: f['ext'] == format_spec
 949             else:
 950                 filter_f = lambda f: f['format_id'] == format_spec
 951             matches = list(filter(filter_f, available_formats))
 952             if matches:
 953                 return matches[-1]
 954         return None
 955
 956     def _calc_headers(self, info_dict):
 957         res = std_headers.copy()
 958
 959         add_headers = info_dict.get('http_headers')
 960         if add_headers:
 961             res.update(add_headers)
 962
 963         cookies = self._calc_cookies(info_dict)
 964         if cookies:
 965             res['Cookie'] = cookies
 966
 967         return res
 968
 969     def _calc_cookies(self, info_dict):
 970         pr = compat_urllib_request.Request(info_dict['url'])
 971         self.cookiejar.add_cookie_header(pr)
 972         return pr.get_header('Cookie')
 973
 974     def process_video_result(self, info_dict, download=True):
 975         assert info_dict.get('_type', 'video') == 'video'
 976
 977         if 'id' not in info_dict:
 978             raise ExtractorError('Missing "id" field in extractor result')
 979         if 'title' not in info_dict:
 980             raise ExtractorError('Missing "title" field in extractor result')
 981
 982         if 'playlist' not in info_dict:
 983             # It isn't part of a playlist
 984             info_dict['playlist'] = None
 985             info_dict['playlist_index'] = None
 986
 987         thumbnails = info_dict.get('thumbnails')
 988         if thumbnails is None:
 989             thumbnail = info_dict.get('thumbnail')
 990             if thumbnail:
 991                 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
 992         if thumbnails:
 993             thumbnails.sort(key=lambda t: (
 994                 t.get('preference'), t.get('width'), t.get('height'),
 995                 t.get('id'), t.get('url')))
 996             for i, t in enumerate(thumbnails):
 997                 if 'width' in t and 'height' in t:
 998                     t['resolution'] = '%dx%d' % (t['width'], t['height'])
 999                 if t.get('id') is None:
1000                     t['id'] = '%d' % i
1001
1002         if thumbnails and 'thumbnail' not in info_dict:
1003             info_dict['thumbnail'] = thumbnails[-1]['url']
1004
1005         if 'display_id' not in info_dict and 'id' in info_dict:
1006             info_dict['display_id'] = info_dict['id']
1007
1008         if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1009             # Working around negative timestamps in Windows
1010             # (see http://bugs.python.org/issue1646728)
1011             if info_dict['timestamp'] < 0 and os.name == 'nt':
1012                 info_dict['timestamp'] = 0
1013             upload_date = datetime.datetime.utcfromtimestamp(
1014                 info_dict['timestamp'])
1015             info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1016
1017         if self.params.get('listsubtitles', False):
1018             if 'automatic_captions' in info_dict:
1019                 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1020             self.list_subtitles(info_dict['id'], info_dict.get('subtitles'), 'subtitles')
1021             return
1022         info_dict['requested_subtitles'] = self.process_subtitles(
1023             info_dict['id'], info_dict.get('subtitles'),
1024             info_dict.get('automatic_captions'))
1025
1026         # This extractors handle format selection themselves
1027         if info_dict['extractor'] in ['Youku']:
1028             if download:
1029                 self.process_info(info_dict)
1030             return info_dict
1031
1032         # We now pick which formats have to be downloaded
1033         if info_dict.get('formats') is None:
1034             # There's only one format available
1035             formats = [info_dict]
1036         else:
1037             formats = info_dict['formats']
1038
1039         if not formats:
1040             raise ExtractorError('No video formats found!')
1041
1042         # We check that all the formats have the format and format_id fields
1043         for i, format in enumerate(formats):
1044             if 'url' not in format:
1045                 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1046
1047             if format.get('format_id') is None:
1048                 format['format_id'] = compat_str(i)
1049             if format.get('format') is None:
1050                 format['format'] = '{id} - {res}{note}'.format(
1051                     id=format['format_id'],
1052                     res=self.format_resolution(format),
1053                     note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1054                 )
1055             # Automatically determine file extension if missing
1056             if 'ext' not in format:
1057                 format['ext'] = determine_ext(format['url']).lower()
1058             # Add HTTP headers, so that external programs can use them from the
1059             # json output
1060             full_format_info = info_dict.copy()
1061             full_format_info.update(format)
1062             format['http_headers'] = self._calc_headers(full_format_info)
1063
1064         format_limit = self.params.get('format_limit', None)
1065         if format_limit:
1066             formats = list(takewhile_inclusive(
1067                 lambda f: f['format_id'] != format_limit, formats
1068             ))
1069
1070         # TODO Central sorting goes here
1071
1072         if formats[0] is not info_dict:
1073             # only set the 'formats' fields if the original info_dict list them
1074             # otherwise we end up with a circular reference, the first (and unique)
1075             # element in the 'formats' field in info_dict is info_dict itself,
1076             # wich can't be exported to json
1077             info_dict['formats'] = formats
1078         if self.params.get('listformats'):
1079             self.list_formats(info_dict)
1080             return
1081         if self.params.get('list_thumbnails'):
1082             self.list_thumbnails(info_dict)
1083             return
1084
1085         req_format = self.params.get('format')
1086         if req_format is None:
1087             req_format = 'best'
1088         formats_to_download = []
1089         if req_format == 'all':
1090             formats_to_download = formats
1091         else:
1092             for rfstr in req_format.split(','):
1093                 # We can accept formats requested in the format: 34/5/best, we pick
1094                 # the first that is available, starting from left
1095                 req_formats = rfstr.split('/')
1096                 for rf in req_formats:
1097                     if re.match(r'.+?\+.+?', rf) is not None:
1098                         # Two formats have been requested like '137+139'
1099                         format_1, format_2 = rf.split('+')
1100                         formats_info = (self.select_format(format_1, formats),
1101                                         self.select_format(format_2, formats))
1102                         if all(formats_info):
1103                             # The first format must contain the video and the
1104                             # second the audio
1105                             if formats_info[0].get('vcodec') == 'none':
1106                                 self.report_error('The first format must '
1107                                                   'contain the video, try using '
1108                                                   '"-f %s+%s"' % (format_2, format_1))
1109                                 return
1110                             output_ext = (
1111                                 formats_info[0]['ext']
1112                                 if self.params.get('merge_output_format') is None
1113                                 else self.params['merge_output_format'])
1114                             selected_format = {
1115                                 'requested_formats': formats_info,
1116                                 'format': '%s+%s' % (formats_info[0].get('format'),
1117                                                      formats_info[1].get('format')),
1118                                 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1119                                                         formats_info[1].get('format_id')),
1120                                 'width': formats_info[0].get('width'),
1121                                 'height': formats_info[0].get('height'),
1122                                 'resolution': formats_info[0].get('resolution'),
1123                                 'fps': formats_info[0].get('fps'),
1124                                 'vcodec': formats_info[0].get('vcodec'),
1125                                 'vbr': formats_info[0].get('vbr'),
1126                                 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1127                                 'acodec': formats_info[1].get('acodec'),
1128                                 'abr': formats_info[1].get('abr'),
1129                                 'ext': output_ext,
1130                             }
1131                         else:
1132                             selected_format = None
1133                     else:
1134                         selected_format = self.select_format(rf, formats)
1135                     if selected_format is not None:
1136                         formats_to_download.append(selected_format)
1137                         break
1138         if not formats_to_download:
1139             raise ExtractorError('requested format not available',
1140                                  expected=True)
1141
1142         if download:
1143             if len(formats_to_download) > 1:
1144                 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1145             for format in formats_to_download:
1146                 new_info = dict(info_dict)
1147                 new_info.update(format)
1148                 self.process_info(new_info)
1149         # We update the info dict with the best quality format (backwards compatibility)
1150         info_dict.update(formats_to_download[-1])
1151         return info_dict
1152
1153     def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1154         """Select the requested subtitles and their format"""
1155         available_subs = {}
1156         if normal_subtitles and self.params.get('writesubtitles'):
1157             available_subs.update(normal_subtitles)
1158         if automatic_captions and self.params.get('writeautomaticsub'):
1159             for lang, cap_info in automatic_captions.items():
1160                 if lang not in available_subs:
1161                     available_subs[lang] = cap_info
1162
1163         if (not self.params.get('writesubtitles') and not
1164                 self.params.get('writeautomaticsub') or not
1165                 available_subs):
1166             return None
1167
1168         if self.params.get('allsubtitles', False):
1169             requested_langs = available_subs.keys()
1170         else:
1171             if self.params.get('subtitleslangs', False):
1172                 requested_langs = self.params.get('subtitleslangs')
1173             elif 'en' in available_subs:
1174                 requested_langs = ['en']
1175             else:
1176                 requested_langs = [list(available_subs.keys())[0]]
1177
1178         formats_query = self.params.get('subtitlesformat', 'best')
1179         formats_preference = formats_query.split('/') if formats_query else []
1180         subs = {}
1181         for lang in requested_langs:
1182             formats = available_subs.get(lang)
1183             if formats is None:
1184                 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1185                 continue
1186             for ext in formats_preference:
1187                 if ext == 'best':
1188                     f = formats[-1]
1189                     break
1190                 matches = list(filter(lambda f: f['ext'] == ext, formats))
1191                 if matches:
1192                     f = matches[-1]
1193                     break
1194             else:
1195                 f = formats[-1]
1196                 self.report_warning(
1197                     'No subtitle format found matching "%s" for language %s, '
1198                     'using %s' % (formats_query, lang, f['ext']))
1199             subs[lang] = f
1200         return subs
1201
1202     def process_info(self, info_dict):
1203         """Process a single resolved IE result."""
1204
1205         assert info_dict.get('_type', 'video') == 'video'
1206
1207         max_downloads = self.params.get('max_downloads')
1208         if max_downloads is not None:
1209             if self._num_downloads >= int(max_downloads):
1210                 raise MaxDownloadsReached()
1211
1212         info_dict['fulltitle'] = info_dict['title']
1213         if len(info_dict['title']) > 200:
1214             info_dict['title'] = info_dict['title'][:197] + '...'
1215
1216         # Keep for backwards compatibility
1217         info_dict['stitle'] = info_dict['title']
1218
1219         if 'format' not in info_dict:
1220             info_dict['format'] = info_dict['ext']
1221
1222         reason = self._match_entry(info_dict, incomplete=False)
1223         if reason is not None:
1224             self.to_screen('[download] ' + reason)
1225             return
1226
1227         self._num_downloads += 1
1228
1229         info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1230
1231         # Forced printings
1232         if self.params.get('forcetitle', False):
1233             self.to_stdout(info_dict['fulltitle'])
1234         if self.params.get('forceid', False):
1235             self.to_stdout(info_dict['id'])
1236         if self.params.get('forceurl', False):
1237             if info_dict.get('requested_formats') is not None:
1238                 for f in info_dict['requested_formats']:
1239                     self.to_stdout(f['url'] + f.get('play_path', ''))
1240             else:
1241                 # For RTMP URLs, also include the playpath
1242                 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1243         if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1244             self.to_stdout(info_dict['thumbnail'])
1245         if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1246             self.to_stdout(info_dict['description'])
1247         if self.params.get('forcefilename', False) and filename is not None:
1248             self.to_stdout(filename)
1249         if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1250             self.to_stdout(formatSeconds(info_dict['duration']))
1251         if self.params.get('forceformat', False):
1252             self.to_stdout(info_dict['format'])
1253         if self.params.get('forcejson', False):
1254             self.to_stdout(json.dumps(info_dict))
1255
1256         # Do nothing else if in simulate mode
1257         if self.params.get('simulate', False):
1258             return
1259
1260         if filename is None:
1261             return
1262
1263         try:
1264             dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1265             if dn and not os.path.exists(dn):
1266                 os.makedirs(dn)
1267         except (OSError, IOError) as err:
1268             self.report_error('unable to create directory ' + compat_str(err))
1269             return
1270
1271         if self.params.get('writedescription', False):
1272             descfn = filename + '.description'
1273             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1274                 self.to_screen('[info] Video description is already present')
1275             elif info_dict.get('description') is None:
1276                 self.report_warning('There\'s no description to write.')
1277             else:
1278                 try:
1279                     self.to_screen('[info] Writing video description to: ' + descfn)
1280                     with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1281                         descfile.write(info_dict['description'])
1282                 except (OSError, IOError):
1283                     self.report_error('Cannot write description file ' + descfn)
1284                     return
1285
1286         if self.params.get('writeannotations', False):
1287             annofn = filename + '.annotations.xml'
1288             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1289                 self.to_screen('[info] Video annotations are already present')
1290             else:
1291                 try:
1292                     self.to_screen('[info] Writing video annotations to: ' + annofn)
1293                     with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1294                         annofile.write(info_dict['annotations'])
1295                 except (KeyError, TypeError):
1296                     self.report_warning('There are no annotations to write.')
1297                 except (OSError, IOError):
1298                     self.report_error('Cannot write annotations file: ' + annofn)
1299                     return
1300
1301         subtitles_are_requested = any([self.params.get('writesubtitles', False),
1302                                        self.params.get('writeautomaticsub')])
1303
1304         if subtitles_are_requested and info_dict.get('requested_subtitles'):
1305             # subtitles download errors are already managed as troubles in relevant IE
1306             # that way it will silently go on when used with unsupporting IE
1307             subtitles = info_dict['requested_subtitles']
1308             ie = self.get_info_extractor(info_dict['extractor_key'])
1309             for sub_lang, sub_info in subtitles.items():
1310                 sub_format = sub_info['ext']
1311                 if sub_info.get('data') is not None:
1312                     sub_data = sub_info['data']
1313                 else:
1314                     try:
1315                         sub_data = ie._download_webpage(
1316                             sub_info['url'], info_dict['id'], note=False)
1317                     except ExtractorError as err:
1318                         self.report_warning('Unable to download subtitle for "%s": %s' %
1319                                             (sub_lang, compat_str(err.cause)))
1320                         continue
1321                 try:
1322                     sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1323                     if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1324                         self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1325                     else:
1326                         self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1327                         with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1328                             subfile.write(sub_data)
1329                 except (OSError, IOError):
1330                     self.report_error('Cannot write subtitles file ' + sub_filename)
1331                     return
1332
1333         if self.params.get('writeinfojson', False):
1334             infofn = os.path.splitext(filename)[0] + '.info.json'
1335             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1336                 self.to_screen('[info] Video description metadata is already present')
1337             else:
1338                 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1339                 try:
1340                     write_json_file(info_dict, infofn)
1341                 except (OSError, IOError):
1342                     self.report_error('Cannot write metadata to JSON file ' + infofn)
1343                     return
1344
1345         self._write_thumbnails(info_dict, filename)
1346
1347         if not self.params.get('skip_download', False):
1348             try:
1349                 def dl(name, info):
1350                     fd = get_suitable_downloader(info, self.params)(self, self.params)
1351                     for ph in self._progress_hooks:
1352                         fd.add_progress_hook(ph)
1353                     if self.params.get('verbose'):
1354                         self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1355                     return fd.download(name, info)
1356
1357                 if info_dict.get('requested_formats') is not None:
1358                     downloaded = []
1359                     success = True
1360                     merger = FFmpegMergerPP(self, not self.params.get('keepvideo'))
1361                     if not merger.available:
1362                         postprocessors = []
1363                         self.report_warning('You have requested multiple '
1364                                             'formats but ffmpeg or avconv are not installed.'
1365                                             ' The formats won\'t be merged')
1366                     else:
1367                         postprocessors = [merger]
1368                     for f in info_dict['requested_formats']:
1369                         new_info = dict(info_dict)
1370                         new_info.update(f)
1371                         fname = self.prepare_filename(new_info)
1372                         fname = prepend_extension(fname, 'f%s' % f['format_id'])
1373                         downloaded.append(fname)
1374                         partial_success = dl(fname, new_info)
1375                         success = success and partial_success
1376                     info_dict['__postprocessors'] = postprocessors
1377                     info_dict['__files_to_merge'] = downloaded
1378                 else:
1379                     # Just a single file
1380                     success = dl(filename, info_dict)
1381             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1382                 self.report_error('unable to download video data: %s' % str(err))
1383                 return
1384             except (OSError, IOError) as err:
1385                 raise UnavailableVideoError(err)
1386             except (ContentTooShortError, ) as err:
1387                 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1388                 return
1389
1390             if success:
1391                 # Fixup content
1392                 fixup_policy = self.params.get('fixup')
1393                 if fixup_policy is None:
1394                     fixup_policy = 'detect_or_warn'
1395
1396                 stretched_ratio = info_dict.get('stretched_ratio')
1397                 if stretched_ratio is not None and stretched_ratio != 1:
1398                     if fixup_policy == 'warn':
1399                         self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1400                             info_dict['id'], stretched_ratio))
1401                     elif fixup_policy == 'detect_or_warn':
1402                         stretched_pp = FFmpegFixupStretchedPP(self)
1403                         if stretched_pp.available:
1404                             info_dict.setdefault('__postprocessors', [])
1405                             info_dict['__postprocessors'].append(stretched_pp)
1406                         else:
1407                             self.report_warning(
1408                                 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1409                                     info_dict['id'], stretched_ratio))
1410                     else:
1411                         assert fixup_policy in ('ignore', 'never')
1412
1413                 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1414                     if fixup_policy == 'warn':
1415                         self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1416                             info_dict['id']))
1417                     elif fixup_policy == 'detect_or_warn':
1418                         fixup_pp = FFmpegFixupM4aPP(self)
1419                         if fixup_pp.available:
1420                             info_dict.setdefault('__postprocessors', [])
1421                             info_dict['__postprocessors'].append(fixup_pp)
1422                         else:
1423                             self.report_warning(
1424                                 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1425                                     info_dict['id']))
1426                     else:
1427                         assert fixup_policy in ('ignore', 'never')
1428
1429                 try:
1430                     self.post_process(filename, info_dict)
1431                 except (PostProcessingError) as err:
1432                     self.report_error('postprocessing: %s' % str(err))
1433                     return
1434                 self.record_download_archive(info_dict)
1435
1436     def download(self, url_list):
1437         """Download a given list of URLs."""
1438         outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1439         if (len(url_list) > 1 and
1440                 '%' not in outtmpl and
1441                 self.params.get('max_downloads') != 1):
1442             raise SameFileError(outtmpl)
1443
1444         for url in url_list:
1445             try:
1446                 # It also downloads the videos
1447                 res = self.extract_info(url)
1448             except UnavailableVideoError:
1449                 self.report_error('unable to download video')
1450             except MaxDownloadsReached:
1451                 self.to_screen('[info] Maximum number of downloaded files reached.')
1452                 raise
1453             else:
1454                 if self.params.get('dump_single_json', False):
1455                     self.to_stdout(json.dumps(res))
1456
1457         return self._download_retcode
1458
1459     def download_with_info_file(self, info_filename):
1460         with contextlib.closing(fileinput.FileInput(
1461                 [info_filename], mode='r',
1462                 openhook=fileinput.hook_encoded('utf-8'))) as f:
1463             # FileInput doesn't have a read method, we can't call json.load
1464             info = json.loads('\n'.join(f))
1465         try:
1466             self.process_ie_result(info, download=True)
1467         except DownloadError:
1468             webpage_url = info.get('webpage_url')
1469             if webpage_url is not None:
1470                 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1471                 return self.download([webpage_url])
1472             else:
1473                 raise
1474         return self._download_retcode
1475
1476     def post_process(self, filename, ie_info):
1477         """Run all the postprocessors on the given file."""
1478         info = dict(ie_info)
1479         info['filepath'] = filename
1480         pps_chain = []
1481         if ie_info.get('__postprocessors') is not None:
1482             pps_chain.extend(ie_info['__postprocessors'])
1483         pps_chain.extend(self._pps)
1484         for pp in pps_chain:
1485             keep_video = None
1486             old_filename = info['filepath']
1487             try:
1488                 keep_video_wish, info = pp.run(info)
1489                 if keep_video_wish is not None:
1490                     if keep_video_wish:
1491                         keep_video = keep_video_wish
1492                     elif keep_video is None:
1493                         # No clear decision yet, let IE decide
1494                         keep_video = keep_video_wish
1495             except PostProcessingError as e:
1496                 self.report_error(e.msg)
1497             if keep_video is False and not self.params.get('keepvideo', False):
1498                 try:
1499                     self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1500                     os.remove(encodeFilename(old_filename))
1501                 except (IOError, OSError):
1502                     self.report_warning('Unable to remove downloaded video file')
1503
1504     def _make_archive_id(self, info_dict):
1505         # Future-proof against any change in case
1506         # and backwards compatibility with prior versions
1507         extractor = info_dict.get('extractor_key')
1508         if extractor is None:
1509             if 'id' in info_dict:
1510                 extractor = info_dict.get('ie_key')  # key in a playlist
1511         if extractor is None:
1512             return None  # Incomplete video information
1513         return extractor.lower() + ' ' + info_dict['id']
1514
1515     def in_download_archive(self, info_dict):
1516         fn = self.params.get('download_archive')
1517         if fn is None:
1518             return False
1519
1520         vid_id = self._make_archive_id(info_dict)
1521         if vid_id is None:
1522             return False  # Incomplete video information
1523
1524         try:
1525             with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1526                 for line in archive_file:
1527                     if line.strip() == vid_id:
1528                         return True
1529         except IOError as ioe:
1530             if ioe.errno != errno.ENOENT:
1531                 raise
1532         return False
1533
1534     def record_download_archive(self, info_dict):
1535         fn = self.params.get('download_archive')
1536         if fn is None:
1537             return
1538         vid_id = self._make_archive_id(info_dict)
1539         assert vid_id
1540         with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1541             archive_file.write(vid_id + '\n')
1542
1543     @staticmethod
1544     def format_resolution(format, default='unknown'):
1545         if format.get('vcodec') == 'none':
1546             return 'audio only'
1547         if format.get('resolution') is not None:
1548             return format['resolution']
1549         if format.get('height') is not None:
1550             if format.get('width') is not None:
1551                 res = '%sx%s' % (format['width'], format['height'])
1552             else:
1553                 res = '%sp' % format['height']
1554         elif format.get('width') is not None:
1555             res = '?x%d' % format['width']
1556         else:
1557             res = default
1558         return res
1559
1560     def _format_note(self, fdict):
1561         res = ''
1562         if fdict.get('ext') in ['f4f', 'f4m']:
1563             res += '(unsupported) '
1564         if fdict.get('format_note') is not None:
1565             res += fdict['format_note'] + ' '
1566         if fdict.get('tbr') is not None:
1567             res += '%4dk ' % fdict['tbr']
1568         if fdict.get('container') is not None:
1569             if res:
1570                 res += ', '
1571             res += '%s container' % fdict['container']
1572         if (fdict.get('vcodec') is not None and
1573                 fdict.get('vcodec') != 'none'):
1574             if res:
1575                 res += ', '
1576             res += fdict['vcodec']
1577             if fdict.get('vbr') is not None:
1578                 res += '@'
1579         elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1580             res += 'video@'
1581         if fdict.get('vbr') is not None:
1582             res += '%4dk' % fdict['vbr']
1583         if fdict.get('fps') is not None:
1584             res += ', %sfps' % fdict['fps']
1585         if fdict.get('acodec') is not None:
1586             if res:
1587                 res += ', '
1588             if fdict['acodec'] == 'none':
1589                 res += 'video only'
1590             else:
1591                 res += '%-5s' % fdict['acodec']
1592         elif fdict.get('abr') is not None:
1593             if res:
1594                 res += ', '
1595             res += 'audio'
1596         if fdict.get('abr') is not None:
1597             res += '@%3dk' % fdict['abr']
1598         if fdict.get('asr') is not None:
1599             res += ' (%5dHz)' % fdict['asr']
1600         if fdict.get('filesize') is not None:
1601             if res:
1602                 res += ', '
1603             res += format_bytes(fdict['filesize'])
1604         elif fdict.get('filesize_approx') is not None:
1605             if res:
1606                 res += ', '
1607             res += '~' + format_bytes(fdict['filesize_approx'])
1608         return res
1609
1610     def list_formats(self, info_dict):
1611         formats = info_dict.get('formats', [info_dict])
1612         table = [
1613             [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1614             for f in formats
1615             if f.get('preference') is None or f['preference'] >= -1000]
1616         if len(formats) > 1:
1617             table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1618
1619         header_line = ['format code', 'extension', 'resolution', 'note']
1620         self.to_screen(
1621             '[info] Available formats for %s:\n%s' %
1622             (info_dict['id'], render_table(header_line, table)))
1623
1624     def list_thumbnails(self, info_dict):
1625         thumbnails = info_dict.get('thumbnails')
1626         if not thumbnails:
1627             tn_url = info_dict.get('thumbnail')
1628             if tn_url:
1629                 thumbnails = [{'id': '0', 'url': tn_url}]
1630             else:
1631                 self.to_screen(
1632                     '[info] No thumbnails present for %s' % info_dict['id'])
1633                 return
1634
1635         self.to_screen(
1636             '[info] Thumbnails for %s:' % info_dict['id'])
1637         self.to_screen(render_table(
1638             ['ID', 'width', 'height', 'URL'],
1639             [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1640
1641     def list_subtitles(self, video_id, subtitles, name='subtitles'):
1642         if not subtitles:
1643             self.to_screen('%s has no %s' % (video_id, name))
1644             return
1645         self.to_screen(
1646             'Available %s for %s:' % (name, video_id))
1647         self.to_screen(render_table(
1648             ['Language', 'formats'],
1649             [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1650                 for lang, formats in subtitles.items()]))
1651
1652     def urlopen(self, req):
1653         """ Start an HTTP download """
1654
1655         # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1656         # always respected by websites, some tend to give out URLs with non percent-encoded
1657         # non-ASCII characters (see telemb.py, ard.py [#3412])
1658         # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1659         # To work around aforementioned issue we will replace request's original URL with
1660         # percent-encoded one
1661         req_is_string = isinstance(req, compat_basestring)
1662         url = req if req_is_string else req.get_full_url()
1663         url_escaped = escape_url(url)
1664
1665         # Substitute URL if any change after escaping
1666         if url != url_escaped:
1667             if req_is_string:
1668                 req = url_escaped
1669             else:
1670                 req = compat_urllib_request.Request(
1671                     url_escaped, data=req.data, headers=req.headers,
1672                     origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
1673
1674         return self._opener.open(req, timeout=self._socket_timeout)
1675
1676     def print_debug_header(self):
1677         if not self.params.get('verbose'):
1678             return
1679
1680         if type('') is not compat_str:
1681             # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1682             self.report_warning(
1683                 'Your Python is broken! Update to a newer and supported version')
1684
1685         stdout_encoding = getattr(
1686             sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1687         encoding_str = (
1688             '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1689                 locale.getpreferredencoding(),
1690                 sys.getfilesystemencoding(),
1691                 stdout_encoding,
1692                 self.get_encoding()))
1693         write_string(encoding_str, encoding=None)
1694
1695         self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1696         try:
1697             sp = subprocess.Popen(
1698                 ['git', 'rev-parse', '--short', 'HEAD'],
1699                 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1700                 cwd=os.path.dirname(os.path.abspath(__file__)))
1701             out, err = sp.communicate()
1702             out = out.decode().strip()
1703             if re.match('[0-9a-f]+', out):
1704                 self._write_string('[debug] Git HEAD: ' + out + '\n')
1705         except:
1706             try:
1707                 sys.exc_clear()
1708             except:
1709                 pass
1710         self._write_string('[debug] Python version %s - %s\n' % (
1711             platform.python_version(), platform_name()))
1712
1713         exe_versions = FFmpegPostProcessor.get_versions(self)
1714         exe_versions['rtmpdump'] = rtmpdump_version()
1715         exe_str = ', '.join(
1716             '%s %s' % (exe, v)
1717             for exe, v in sorted(exe_versions.items())
1718             if v
1719         )
1720         if not exe_str:
1721             exe_str = 'none'
1722         self._write_string('[debug] exe versions: %s\n' % exe_str)
1723
1724         proxy_map = {}
1725         for handler in self._opener.handlers:
1726             if hasattr(handler, 'proxies'):
1727                 proxy_map.update(handler.proxies)
1728         self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1729
1730         if self.params.get('call_home', False):
1731             ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1732             self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1733             latest_version = self.urlopen(
1734                 'https://yt-dl.org/latest/version').read().decode('utf-8')
1735             if version_tuple(latest_version) > version_tuple(__version__):
1736                 self.report_warning(
1737                     'You are using an outdated version (newest version: %s)! '
1738                     'See https://yt-dl.org/update if you need help updating.' %
1739                     latest_version)
1740
1741     def _setup_opener(self):
1742         timeout_val = self.params.get('socket_timeout')
1743         self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1744
1745         opts_cookiefile = self.params.get('cookiefile')
1746         opts_proxy = self.params.get('proxy')
1747
1748         if opts_cookiefile is None:
1749             self.cookiejar = compat_cookiejar.CookieJar()
1750         else:
1751             self.cookiejar = compat_cookiejar.MozillaCookieJar(
1752                 opts_cookiefile)
1753             if os.access(opts_cookiefile, os.R_OK):
1754                 self.cookiejar.load()
1755
1756         cookie_processor = compat_urllib_request.HTTPCookieProcessor(
1757             self.cookiejar)
1758         if opts_proxy is not None:
1759             if opts_proxy == '':
1760                 proxies = {}
1761             else:
1762                 proxies = {'http': opts_proxy, 'https': opts_proxy}
1763         else:
1764             proxies = compat_urllib_request.getproxies()
1765             # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1766             if 'http' in proxies and 'https' not in proxies:
1767                 proxies['https'] = proxies['http']
1768         proxy_handler = PerRequestProxyHandler(proxies)
1769
1770         debuglevel = 1 if self.params.get('debug_printtraffic') else 0
1771         https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
1772         ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
1773         opener = compat_urllib_request.build_opener(
1774             proxy_handler, https_handler, cookie_processor, ydlh)
1775
1776         # Delete the default user-agent header, which would otherwise apply in
1777         # cases where our custom HTTP handler doesn't come into play
1778         # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
1779         opener.addheaders = []
1780         self._opener = opener
1781
1782     def encode(self, s):
1783         if isinstance(s, bytes):
1784             return s  # Already encoded
1785
1786         try:
1787             return s.encode(self.get_encoding())
1788         except UnicodeEncodeError as err:
1789             err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
1790             raise
1791
1792     def get_encoding(self):
1793         encoding = self.params.get('encoding')
1794         if encoding is None:
1795             encoding = preferredencoding()
1796         return encoding
1797
1798     def _write_thumbnails(self, info_dict, filename):
1799         if self.params.get('writethumbnail', False):
1800             thumbnails = info_dict.get('thumbnails')
1801             if thumbnails:
1802                 thumbnails = [thumbnails[-1]]
1803         elif self.params.get('write_all_thumbnails', False):
1804             thumbnails = info_dict.get('thumbnails')
1805         else:
1806             return
1807
1808         if not thumbnails:
1809             # No thumbnails present, so return immediately
1810             return
1811
1812         for t in thumbnails:
1813             thumb_ext = determine_ext(t['url'], 'jpg')
1814             suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
1815             thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
1816             thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
1817
1818             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
1819                 self.to_screen('[%s] %s: Thumbnail %sis already present' %
1820                                (info_dict['extractor'], info_dict['id'], thumb_display_id))
1821             else:
1822                 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
1823                                (info_dict['extractor'], info_dict['id'], thumb_display_id))
1824                 try:
1825                     uf = self.urlopen(t['url'])
1826                     with open(thumb_filename, 'wb') as thumbf:
1827                         shutil.copyfileobj(uf, thumbf)
1828                     self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
1829                                    (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
1830                 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1831                     self.report_warning('Unable to download thumbnail "%s": %s' %
1832                                         (t['url'], compat_str(err)))