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 tokenize
  25 import traceback
  26
  27 from .compat import (
  28     compat_basestring,
  29     compat_cookiejar,
  30     compat_expanduser,
  31     compat_get_terminal_size,
  32     compat_http_client,
  33     compat_kwargs,
  34     compat_os_name,
  35     compat_str,
  36     compat_tokenize_tokenize,
  37     compat_urllib_error,
  38     compat_urllib_request,
  39     compat_urllib_request_DataHandler,
  40 )
  41 from .utils import (
  42     age_restricted,
  43     args_to_str,
  44     ContentTooShortError,
  45     date_from_str,
  46     DateRange,
  47     DEFAULT_OUTTMPL,
  48     determine_ext,
  49     determine_protocol,
  50     DownloadError,
  51     encode_compat_str,
  52     encodeFilename,
  53     error_to_compat_str,
  54     ExtractorError,
  55     format_bytes,
  56     formatSeconds,
  57     locked_file,
  58     make_HTTPS_handler,
  59     MaxDownloadsReached,
  60     PagedList,
  61     parse_filesize,
  62     PerRequestProxyHandler,
  63     platform_name,
  64     PostProcessingError,
  65     preferredencoding,
  66     prepend_extension,
  67     render_table,
  68     replace_extension,
  69     SameFileError,
  70     sanitize_filename,
  71     sanitize_path,
  72     sanitize_url,
  73     sanitized_Request,
  74     std_headers,
  75     subtitles_filename,
  76     UnavailableVideoError,
  77     url_basename,
  78     version_tuple,
  79     write_json_file,
  80     write_string,
  81     YoutubeDLCookieProcessor,
  82     YoutubeDLHandler,
  83 )
  84 from .cache import Cache
  85 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
  86 from .downloader import get_suitable_downloader
  87 from .downloader.rtmp import rtmpdump_version
  88 from .postprocessor import (
  89     FFmpegFixupM3u8PP,
  90     FFmpegFixupM4aPP,
  91     FFmpegFixupStretchedPP,
  92     FFmpegMergerPP,
  93     FFmpegPostProcessor,
  94     get_postprocessor,
  95 )
  96 from .version import __version__
  97
  98 if compat_os_name == 'nt':
  99     import ctypes
 100
 101
 102 class YoutubeDL(object):
 103     """YoutubeDL class.
 104
 105     YoutubeDL objects are the ones responsible of downloading the
 106     actual video file and writing it to disk if the user has requested
 107     it, among some other tasks. In most cases there should be one per
 108     program. As, given a video URL, the downloader doesn't know how to
 109     extract all the needed information, task that InfoExtractors do, it
 110     has to pass the URL to one of them.
 111
 112     For this, YoutubeDL objects have a method that allows
 113     InfoExtractors to be registered in a given order. When it is passed
 114     a URL, the YoutubeDL object handles it to the first InfoExtractor it
 115     finds that reports being able to handle it. The InfoExtractor extracts
 116     all the information about the video or videos the URL refers to, and
 117     YoutubeDL process the extracted information, possibly using a File
 118     Downloader to download the video.
 119
 120     YoutubeDL objects accept a lot of parameters. In order not to saturate
 121     the object constructor with arguments, it receives a dictionary of
 122     options instead. These options are available through the params
 123     attribute for the InfoExtractors to use. The YoutubeDL also
 124     registers itself as the downloader in charge for the InfoExtractors
 125     that are added to it, so this is a "mutual registration".
 126
 127     Available options:
 128
 129     username:          Username for authentication purposes.
 130     password:          Password for authentication purposes.
 131     videopassword:     Password for accessing a video.
 132     usenetrc:          Use netrc for authentication instead.
 133     verbose:           Print additional info to stdout.
 134     quiet:             Do not print messages to stdout.
 135     no_warnings:       Do not print out anything for warnings.
 136     forceurl:          Force printing final URL.
 137     forcetitle:        Force printing title.
 138     forceid:           Force printing ID.
 139     forcethumbnail:    Force printing thumbnail URL.
 140     forcedescription:  Force printing description.
 141     forcefilename:     Force printing final filename.
 142     forceduration:     Force printing duration.
 143     forcejson:         Force printing info_dict as JSON.
 144     dump_single_json:  Force printing the info_dict of the whole playlist
 145                        (or video) as a single JSON line.
 146     simulate:          Do not download the video files.
 147     format:            Video format code. See options.py for more information.
 148     outtmpl:           Template for output names.
 149     restrictfilenames: Do not allow "&" and spaces in file names
 150     ignoreerrors:      Do not stop on download errors.
 151     force_generic_extractor: Force downloader to use the generic extractor
 152     nooverwrites:      Prevent overwriting files.
 153     playliststart:     Playlist item to start at.
 154     playlistend:       Playlist item to end at.
 155     playlist_items:    Specific indices of playlist to download.
 156     playlistreverse:   Download playlist items in reverse order.
 157     matchtitle:        Download only matching titles.
 158     rejecttitle:       Reject downloads for matching titles.
 159     logger:            Log messages to a logging.Logger instance.
 160     logtostderr:       Log messages to stderr instead of stdout.
 161     writedescription:  Write the video description to a .description file
 162     writeinfojson:     Write the video description to a .info.json file
 163     writeannotations:  Write the video annotations to a .annotations.xml file
 164     writethumbnail:    Write the thumbnail image to a file
 165     write_all_thumbnails:  Write all thumbnail formats to files
 166     writesubtitles:    Write the video subtitles to a file
 167     writeautomaticsub: Write the automatically generated subtitles to a file
 168     allsubtitles:      Downloads all the subtitles of the video
 169                        (requires writesubtitles or writeautomaticsub)
 170     listsubtitles:     Lists all available subtitles for the video
 171     subtitlesformat:   The format code for subtitles
 172     subtitleslangs:    List of languages of the subtitles to download
 173     keepvideo:         Keep the video file after post-processing
 174     daterange:         A DateRange object, download only if the upload_date is in the range.
 175     skip_download:     Skip the actual download of the video file
 176     cachedir:          Location of the cache files in the filesystem.
 177                        False to disable filesystem cache.
 178     noplaylist:        Download single video instead of a playlist if in doubt.
 179     age_limit:         An integer representing the user's age in years.
 180                        Unsuitable videos for the given age are skipped.
 181     min_views:         An integer representing the minimum view count the video
 182                        must have in order to not be skipped.
 183                        Videos without view count information are always
 184                        downloaded. None for no limit.
 185     max_views:         An integer representing the maximum view count.
 186                        Videos that are more popular than that are not
 187                        downloaded.
 188                        Videos without view count information are always
 189                        downloaded. None for no limit.
 190     download_archive:  File name of a file where all downloads are recorded.
 191                        Videos already present in the file are not downloaded
 192                        again.
 193     cookiefile:        File name where cookies should be read from and dumped to.
 194     nocheckcertificate:Do not verify SSL certificates
 195     prefer_insecure:   Use HTTP instead of HTTPS to retrieve information.
 196                        At the moment, this is only supported by YouTube.
 197     proxy:             URL of the proxy server to use
 198     cn_verification_proxy:  URL of the proxy to use for IP address verification
 199                        on Chinese sites. (Experimental)
 200     socket_timeout:    Time to wait for unresponsive hosts, in seconds
 201     bidi_workaround:   Work around buggy terminals without bidirectional text
 202                        support, using fridibi
 203     debug_printtraffic:Print out sent and received HTTP traffic
 204     include_ads:       Download ads as well
 205     default_search:    Prepend this string if an input url is not valid.
 206                        'auto' for elaborate guessing
 207     encoding:          Use this encoding instead of the system-specified.
 208     extract_flat:      Do not resolve URLs, return the immediate result.
 209                        Pass in 'in_playlist' to only show this behavior for
 210                        playlist items.
 211     postprocessors:    A list of dictionaries, each with an entry
 212                        * key:  The name of the postprocessor. See
 213                                youtube_dl/postprocessor/__init__.py for a list.
 214                        as well as any further keyword arguments for the
 215                        postprocessor.
 216     progress_hooks:    A list of functions that get called on download
 217                        progress, with a dictionary with the entries
 218                        * status: One of "downloading", "error", or "finished".
 219                                  Check this first and ignore unknown values.
 220
 221                        If status is one of "downloading", or "finished", the
 222                        following properties may also be present:
 223                        * filename: The final filename (always present)
 224                        * tmpfilename: The filename we're currently writing to
 225                        * downloaded_bytes: Bytes on disk
 226                        * total_bytes: Size of the whole file, None if unknown
 227                        * total_bytes_estimate: Guess of the eventual file size,
 228                                                None if unavailable.
 229                        * elapsed: The number of seconds since download started.
 230                        * eta: The estimated time in seconds, None if unknown
 231                        * speed: The download speed in bytes/second, None if
 232                                 unknown
 233                        * fragment_index: The counter of the currently
 234                                          downloaded video fragment.
 235                        * fragment_count: The number of fragments (= individual
 236                                          files that will be merged)
 237
 238                        Progress hooks are guaranteed to be called at least once
 239                        (with status "finished") if the download is successful.
 240     merge_output_format: Extension to use when merging formats.
 241     fixup:             Automatically correct known faults of the file.
 242                        One of:
 243                        - "never": do nothing
 244                        - "warn": only emit a warning
 245                        - "detect_or_warn": check whether we can do anything
 246                                            about it, warn otherwise (default)
 247     source_address:    (Experimental) Client-side IP address to bind to.
 248     call_home:         Boolean, true iff we are allowed to contact the
 249                        youtube-dl servers for debugging.
 250     sleep_interval:    Number of seconds to sleep before each download.
 251     listformats:       Print an overview of available video formats and exit.
 252     list_thumbnails:   Print a table of all thumbnails and exit.
 253     match_filter:      A function that gets called with the info_dict of
 254                        every video.
 255                        If it returns a message, the video is ignored.
 256                        If it returns None, the video is downloaded.
 257                        match_filter_func in utils.py is one example for this.
 258     no_color:          Do not emit color codes in output.
 259
 260     The following options determine which downloader is picked:
 261     external_downloader: Executable of the external downloader to call.
 262                        None or unset for standard (built-in) downloader.
 263     hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
 264                        if True, otherwise use ffmpeg/avconv if False, otherwise
 265                        use downloader suggested by extractor if None.
 266
 267     The following parameters are not used by YoutubeDL itself, they are used by
 268     the downloader (see youtube_dl/downloader/common.py):
 269     nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
 270     noresizebuffer, retries, continuedl, noprogress, consoletitle,
 271     xattr_set_filesize, external_downloader_args, hls_use_mpegts.
 272
 273     The following options are used by the post processors:
 274     prefer_ffmpeg:     If True, use ffmpeg instead of avconv if both are available,
 275                        otherwise prefer avconv.
 276     postprocessor_args: A list of additional command-line arguments for the
 277                         postprocessor.
 278     """
 279
 280     params = None
 281     _ies = []
 282     _pps = []
 283     _download_retcode = None
 284     _num_downloads = None
 285     _screen_file = None
 286
 287     def __init__(self, params=None, auto_init=True):
 288         """Create a FileDownloader object with the given options."""
 289         if params is None:
 290             params = {}
 291         self._ies = []
 292         self._ies_instances = {}
 293         self._pps = []
 294         self._progress_hooks = []
 295         self._download_retcode = 0
 296         self._num_downloads = 0
 297         self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
 298         self._err_file = sys.stderr
 299         self.params = {
 300             # Default parameters
 301             'nocheckcertificate': False,
 302         }
 303         self.params.update(params)
 304         self.cache = Cache(self)
 305
 306         if params.get('bidi_workaround', False):
 307             try:
 308                 import pty
 309                 master, slave = pty.openpty()
 310                 width = compat_get_terminal_size().columns
 311                 if width is None:
 312                     width_args = []
 313                 else:
 314                     width_args = ['-w', str(width)]
 315                 sp_kwargs = dict(
 316                     stdin=subprocess.PIPE,
 317                     stdout=slave,
 318                     stderr=self._err_file)
 319                 try:
 320                     self._output_process = subprocess.Popen(
 321                         ['bidiv'] + width_args, **sp_kwargs
 322                     )
 323                 except OSError:
 324                     self._output_process = subprocess.Popen(
 325                         ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
 326                 self._output_channel = os.fdopen(master, 'rb')
 327             except OSError as ose:
 328                 if ose.errno == 2:
 329                     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 $PATH.')
 330                 else:
 331                     raise
 332
 333         if (sys.version_info >= (3,) and sys.platform != 'win32' and
 334                 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
 335                 not params.get('restrictfilenames', False)):
 336             # On Python 3, the Unicode filesystem API will throw errors (#1474)
 337             self.report_warning(
 338                 'Assuming --restrict-filenames since file system encoding '
 339                 'cannot encode all characters. '
 340                 'Set the LC_ALL environment variable to fix this.')
 341             self.params['restrictfilenames'] = True
 342
 343         if isinstance(params.get('outtmpl'), bytes):
 344             self.report_warning(
 345                 'Parameter outtmpl is bytes, but should be a unicode string. '
 346                 'Put  from __future__ import unicode_literals  at the top of your code file or consider switching to Python 3.x.')
 347
 348         self._setup_opener()
 349
 350         if auto_init:
 351             self.print_debug_header()
 352             self.add_default_info_extractors()
 353
 354         for pp_def_raw in self.params.get('postprocessors', []):
 355             pp_class = get_postprocessor(pp_def_raw['key'])
 356             pp_def = dict(pp_def_raw)
 357             del pp_def['key']
 358             pp = pp_class(self, **compat_kwargs(pp_def))
 359             self.add_post_processor(pp)
 360
 361         for ph in self.params.get('progress_hooks', []):
 362             self.add_progress_hook(ph)
 363
 364     def warn_if_short_id(self, argv):
 365         # short YouTube ID starting with dash?
 366         idxs = [
 367             i for i, a in enumerate(argv)
 368             if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
 369         if idxs:
 370             correct_argv = (
 371                 ['youtube-dl'] +
 372                 [a for i, a in enumerate(argv) if i not in idxs] +
 373                 ['--'] + [argv[i] for i in idxs]
 374             )
 375             self.report_warning(
 376                 'Long argument string detected. '
 377                 'Use -- to separate parameters and URLs, like this:\n%s\n' %
 378                 args_to_str(correct_argv))
 379
 380     def add_info_extractor(self, ie):
 381         """Add an InfoExtractor object to the end of the list."""
 382         self._ies.append(ie)
 383         if not isinstance(ie, type):
 384             self._ies_instances[ie.ie_key()] = ie
 385             ie.set_downloader(self)
 386
 387     def get_info_extractor(self, ie_key):
 388         """
 389         Get an instance of an IE with name ie_key, it will try to get one from
 390         the _ies list, if there's no instance it will create a new one and add
 391         it to the extractor list.
 392         """
 393         ie = self._ies_instances.get(ie_key)
 394         if ie is None:
 395             ie = get_info_extractor(ie_key)()
 396             self.add_info_extractor(ie)
 397         return ie
 398
 399     def add_default_info_extractors(self):
 400         """
 401         Add the InfoExtractors returned by gen_extractors to the end of the list
 402         """
 403         for ie in gen_extractor_classes():
 404             self.add_info_extractor(ie)
 405
 406     def add_post_processor(self, pp):
 407         """Add a PostProcessor object to the end of the chain."""
 408         self._pps.append(pp)
 409         pp.set_downloader(self)
 410
 411     def add_progress_hook(self, ph):
 412         """Add the progress hook (currently only for the file downloader)"""
 413         self._progress_hooks.append(ph)
 414
 415     def _bidi_workaround(self, message):
 416         if not hasattr(self, '_output_channel'):
 417             return message
 418
 419         assert hasattr(self, '_output_process')
 420         assert isinstance(message, compat_str)
 421         line_count = message.count('\n') + 1
 422         self._output_process.stdin.write((message + '\n').encode('utf-8'))
 423         self._output_process.stdin.flush()
 424         res = ''.join(self._output_channel.readline().decode('utf-8')
 425                       for _ in range(line_count))
 426         return res[:-len('\n')]
 427
 428     def to_screen(self, message, skip_eol=False):
 429         """Print message to stdout if not in quiet mode."""
 430         return self.to_stdout(message, skip_eol, check_quiet=True)
 431
 432     def _write_string(self, s, out=None):
 433         write_string(s, out=out, encoding=self.params.get('encoding'))
 434
 435     def to_stdout(self, message, skip_eol=False, check_quiet=False):
 436         """Print message to stdout if not in quiet mode."""
 437         if self.params.get('logger'):
 438             self.params['logger'].debug(message)
 439         elif not check_quiet or not self.params.get('quiet', False):
 440             message = self._bidi_workaround(message)
 441             terminator = ['\n', ''][skip_eol]
 442             output = message + terminator
 443
 444             self._write_string(output, self._screen_file)
 445
 446     def to_stderr(self, message):
 447         """Print message to stderr."""
 448         assert isinstance(message, compat_str)
 449         if self.params.get('logger'):
 450             self.params['logger'].error(message)
 451         else:
 452             message = self._bidi_workaround(message)
 453             output = message + '\n'
 454             self._write_string(output, self._err_file)
 455
 456     def to_console_title(self, message):
 457         if not self.params.get('consoletitle', False):
 458             return
 459         if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
 460             # c_wchar_p() might not be necessary if `message` is
 461             # already of type unicode()
 462             ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
 463         elif 'TERM' in os.environ:
 464             self._write_string('\033]0;%s\007' % message, self._screen_file)
 465
 466     def save_console_title(self):
 467         if not self.params.get('consoletitle', False):
 468             return
 469         if 'TERM' in os.environ:
 470             # Save the title on stack
 471             self._write_string('\033[22;0t', self._screen_file)
 472
 473     def restore_console_title(self):
 474         if not self.params.get('consoletitle', False):
 475             return
 476         if 'TERM' in os.environ:
 477             # Restore the title from stack
 478             self._write_string('\033[23;0t', self._screen_file)
 479
 480     def __enter__(self):
 481         self.save_console_title()
 482         return self
 483
 484     def __exit__(self, *args):
 485         self.restore_console_title()
 486
 487         if self.params.get('cookiefile') is not None:
 488             self.cookiejar.save()
 489
 490     def trouble(self, message=None, tb=None):
 491         """Determine action to take when a download problem appears.
 492
 493         Depending on if the downloader has been configured to ignore
 494         download errors or not, this method may throw an exception or
 495         not when errors are found, after printing the message.
 496
 497         tb, if given, is additional traceback information.
 498         """
 499         if message is not None:
 500             self.to_stderr(message)
 501         if self.params.get('verbose'):
 502             if tb is None:
 503                 if sys.exc_info()[0]:  # if .trouble has been called from an except block
 504                     tb = ''
 505                     if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
 506                         tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
 507                     tb += encode_compat_str(traceback.format_exc())
 508                 else:
 509                     tb_data = traceback.format_list(traceback.extract_stack())
 510                     tb = ''.join(tb_data)
 511             self.to_stderr(tb)
 512         if not self.params.get('ignoreerrors', False):
 513             if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
 514                 exc_info = sys.exc_info()[1].exc_info
 515             else:
 516                 exc_info = sys.exc_info()
 517             raise DownloadError(message, exc_info)
 518         self._download_retcode = 1
 519
 520     def report_warning(self, message):
 521         '''
 522         Print the message to stderr, it will be prefixed with 'WARNING:'
 523         If stderr is a tty file the 'WARNING:' will be colored
 524         '''
 525         if self.params.get('logger') is not None:
 526             self.params['logger'].warning(message)
 527         else:
 528             if self.params.get('no_warnings'):
 529                 return
 530             if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
 531                 _msg_header = '\033[0;33mWARNING:\033[0m'
 532             else:
 533                 _msg_header = 'WARNING:'
 534             warning_message = '%s %s' % (_msg_header, message)
 535             self.to_stderr(warning_message)
 536
 537     def report_error(self, message, tb=None):
 538         '''
 539         Do the same as trouble, but prefixes the message with 'ERROR:', colored
 540         in red if stderr is a tty file.
 541         '''
 542         if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
 543             _msg_header = '\033[0;31mERROR:\033[0m'
 544         else:
 545             _msg_header = 'ERROR:'
 546         error_message = '%s %s' % (_msg_header, message)
 547         self.trouble(error_message, tb)
 548
 549     def report_file_already_downloaded(self, file_name):
 550         """Report file has already been fully downloaded."""
 551         try:
 552             self.to_screen('[download] %s has already been downloaded' % file_name)
 553         except UnicodeEncodeError:
 554             self.to_screen('[download] The file has already been downloaded')
 555
 556     def prepare_filename(self, info_dict):
 557         """Generate the output filename."""
 558         try:
 559             template_dict = dict(info_dict)
 560
 561             template_dict['epoch'] = int(time.time())
 562             autonumber_size = self.params.get('autonumber_size')
 563             if autonumber_size is None:
 564                 autonumber_size = 5
 565             autonumber_templ = '%0' + str(autonumber_size) + 'd'
 566             template_dict['autonumber'] = autonumber_templ % self._num_downloads
 567             if template_dict.get('playlist_index') is not None:
 568                 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
 569             if template_dict.get('resolution') is None:
 570                 if template_dict.get('width') and template_dict.get('height'):
 571                     template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
 572                 elif template_dict.get('height'):
 573                     template_dict['resolution'] = '%sp' % template_dict['height']
 574                 elif template_dict.get('width'):
 575                     template_dict['resolution'] = '%dx?' % template_dict['width']
 576
 577             sanitize = lambda k, v: sanitize_filename(
 578                 compat_str(v),
 579                 restricted=self.params.get('restrictfilenames'),
 580                 is_id=(k == 'id'))
 581             template_dict = dict((k, sanitize(k, v))
 582                                  for k, v in template_dict.items()
 583                                  if v is not None and not isinstance(v, (list, tuple, dict)))
 584             template_dict = collections.defaultdict(lambda: 'NA', template_dict)
 585
 586             outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
 587             tmpl = compat_expanduser(outtmpl)
 588             filename = tmpl % template_dict
 589             # Temporary fix for #4787
 590             # 'Treat' all problem characters by passing filename through preferredencoding
 591             # to workaround encoding issues with subprocess on python2 @ Windows
 592             if sys.version_info < (3, 0) and sys.platform == 'win32':
 593                 filename = encodeFilename(filename, True).decode(preferredencoding())
 594             return sanitize_path(filename)
 595         except ValueError as err:
 596             self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
 597             return None
 598
 599     def _match_entry(self, info_dict, incomplete):
 600         """ Returns None iff the file should be downloaded """
 601
 602         video_title = info_dict.get('title', info_dict.get('id', 'video'))
 603         if 'title' in info_dict:
 604             # This can happen when we're just evaluating the playlist
 605             title = info_dict['title']
 606             matchtitle = self.params.get('matchtitle', False)
 607             if matchtitle:
 608                 if not re.search(matchtitle, title, re.IGNORECASE):
 609                     return '"' + title + '" title did not match pattern "' + matchtitle + '"'
 610             rejecttitle = self.params.get('rejecttitle', False)
 611             if rejecttitle:
 612                 if re.search(rejecttitle, title, re.IGNORECASE):
 613                     return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
 614         date = info_dict.get('upload_date')
 615         if date is not None:
 616             dateRange = self.params.get('daterange', DateRange())
 617             if date not in dateRange:
 618                 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
 619         view_count = info_dict.get('view_count')
 620         if view_count is not None:
 621             min_views = self.params.get('min_views')
 622             if min_views is not None and view_count < min_views:
 623                 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
 624             max_views = self.params.get('max_views')
 625             if max_views is not None and view_count > max_views:
 626                 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
 627         if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
 628             return 'Skipping "%s" because it is age restricted' % video_title
 629         if self.in_download_archive(info_dict):
 630             return '%s has already been recorded in archive' % video_title
 631
 632         if not incomplete:
 633             match_filter = self.params.get('match_filter')
 634             if match_filter is not None:
 635                 ret = match_filter(info_dict)
 636                 if ret is not None:
 637                     return ret
 638
 639         return None
 640
 641     @staticmethod
 642     def add_extra_info(info_dict, extra_info):
 643         '''Set the keys from extra_info in info dict if they are missing'''
 644         for key, value in extra_info.items():
 645             info_dict.setdefault(key, value)
 646
 647     def extract_info(self, url, download=True, ie_key=None, extra_info={},
 648                      process=True, force_generic_extractor=False):
 649         '''
 650         Returns a list with a dictionary for each video we find.
 651         If 'download', also downloads the videos.
 652         extra_info is a dict containing the extra values to add to each result
 653         '''
 654
 655         if not ie_key and force_generic_extractor:
 656             ie_key = 'Generic'
 657
 658         if ie_key:
 659             ies = [self.get_info_extractor(ie_key)]
 660         else:
 661             ies = self._ies
 662
 663         for ie in ies:
 664             if not ie.suitable(url):
 665                 continue
 666
 667             ie = self.get_info_extractor(ie.ie_key())
 668             if not ie.working():
 669                 self.report_warning('The program functionality for this site has been marked as broken, '
 670                                     'and will probably not work.')
 671
 672             try:
 673                 ie_result = ie.extract(url)
 674                 if ie_result is None:  # Finished already (backwards compatibility; listformats and friends should be moved here)
 675                     break
 676                 if isinstance(ie_result, list):
 677                     # Backwards compatibility: old IE result format
 678                     ie_result = {
 679                         '_type': 'compat_list',
 680                         'entries': ie_result,
 681                     }
 682                 self.add_default_extra_info(ie_result, ie, url)
 683                 if process:
 684                     return self.process_ie_result(ie_result, download, extra_info)
 685                 else:
 686                     return ie_result
 687             except ExtractorError as e:  # An error we somewhat expected
 688                 self.report_error(compat_str(e), e.format_traceback())
 689                 break
 690             except MaxDownloadsReached:
 691                 raise
 692             except Exception as e:
 693                 if self.params.get('ignoreerrors', False):
 694                     self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
 695                     break
 696                 else:
 697                     raise
 698         else:
 699             self.report_error('no suitable InfoExtractor for URL %s' % url)
 700
 701     def add_default_extra_info(self, ie_result, ie, url):
 702         self.add_extra_info(ie_result, {
 703             'extractor': ie.IE_NAME,
 704             'webpage_url': url,
 705             'webpage_url_basename': url_basename(url),
 706             'extractor_key': ie.ie_key(),
 707         })
 708
 709     def process_ie_result(self, ie_result, download=True, extra_info={}):
 710         """
 711         Take the result of the ie(may be modified) and resolve all unresolved
 712         references (URLs, playlist items).
 713
 714         It will also download the videos if 'download'.
 715         Returns the resolved ie_result.
 716         """
 717         result_type = ie_result.get('_type', 'video')
 718
 719         if result_type in ('url', 'url_transparent'):
 720             extract_flat = self.params.get('extract_flat', False)
 721             if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
 722                     extract_flat is True):
 723                 if self.params.get('forcejson', False):
 724                     self.to_stdout(json.dumps(ie_result))
 725                 return ie_result
 726
 727         if result_type == 'video':
 728             self.add_extra_info(ie_result, extra_info)
 729             return self.process_video_result(ie_result, download=download)
 730         elif result_type == 'url':
 731             # We have to add extra_info to the results because it may be
 732             # contained in a playlist
 733             return self.extract_info(ie_result['url'],
 734                                      download,
 735                                      ie_key=ie_result.get('ie_key'),
 736                                      extra_info=extra_info)
 737         elif result_type == 'url_transparent':
 738             # Use the information from the embedding page
 739             info = self.extract_info(
 740                 ie_result['url'], ie_key=ie_result.get('ie_key'),
 741                 extra_info=extra_info, download=False, process=False)
 742
 743             force_properties = dict(
 744                 (k, v) for k, v in ie_result.items() if v is not None)
 745             for f in ('_type', 'url', 'ie_key'):
 746                 if f in force_properties:
 747                     del force_properties[f]
 748             new_result = info.copy()
 749             new_result.update(force_properties)
 750
 751             assert new_result.get('_type') != 'url_transparent'
 752
 753             return self.process_ie_result(
 754                 new_result, download=download, extra_info=extra_info)
 755         elif result_type == 'playlist' or result_type == 'multi_video':
 756             # We process each entry in the playlist
 757             playlist = ie_result.get('title') or ie_result.get('id')
 758             self.to_screen('[download] Downloading playlist: %s' % playlist)
 759
 760             playlist_results = []
 761
 762             playliststart = self.params.get('playliststart', 1) - 1
 763             playlistend = self.params.get('playlistend')
 764             # For backwards compatibility, interpret -1 as whole list
 765             if playlistend == -1:
 766                 playlistend = None
 767
 768             playlistitems_str = self.params.get('playlist_items')
 769             playlistitems = None
 770             if playlistitems_str is not None:
 771                 def iter_playlistitems(format):
 772                     for string_segment in format.split(','):
 773                         if '-' in string_segment:
 774                             start, end = string_segment.split('-')
 775                             for item in range(int(start), int(end) + 1):
 776                                 yield int(item)
 777                         else:
 778                             yield int(string_segment)
 779                 playlistitems = iter_playlistitems(playlistitems_str)
 780
 781             ie_entries = ie_result['entries']
 782             if isinstance(ie_entries, list):
 783                 n_all_entries = len(ie_entries)
 784                 if playlistitems:
 785                     entries = [
 786                         ie_entries[i - 1] for i in playlistitems
 787                         if -n_all_entries <= i - 1 < n_all_entries]
 788                 else:
 789                     entries = ie_entries[playliststart:playlistend]
 790                 n_entries = len(entries)
 791                 self.to_screen(
 792                     '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
 793                     (ie_result['extractor'], playlist, n_all_entries, n_entries))
 794             elif isinstance(ie_entries, PagedList):
 795                 if playlistitems:
 796                     entries = []
 797                     for item in playlistitems:
 798                         entries.extend(ie_entries.getslice(
 799                             item - 1, item
 800                         ))
 801                 else:
 802                     entries = ie_entries.getslice(
 803                         playliststart, playlistend)
 804                 n_entries = len(entries)
 805                 self.to_screen(
 806                     '[%s] playlist %s: Downloading %d videos' %
 807                     (ie_result['extractor'], playlist, n_entries))
 808             else:  # iterable
 809                 if playlistitems:
 810                     entry_list = list(ie_entries)
 811                     entries = [entry_list[i - 1] for i in playlistitems]
 812                 else:
 813                     entries = list(itertools.islice(
 814                         ie_entries, playliststart, playlistend))
 815                 n_entries = len(entries)
 816                 self.to_screen(
 817                     '[%s] playlist %s: Downloading %d videos' %
 818                     (ie_result['extractor'], playlist, n_entries))
 819
 820             if self.params.get('playlistreverse', False):
 821                 entries = entries[::-1]
 822
 823             for i, entry in enumerate(entries, 1):
 824                 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
 825                 extra = {
 826                     'n_entries': n_entries,
 827                     'playlist': playlist,
 828                     'playlist_id': ie_result.get('id'),
 829                     'playlist_title': ie_result.get('title'),
 830                     'playlist_index': i + playliststart,
 831                     'extractor': ie_result['extractor'],
 832                     'webpage_url': ie_result['webpage_url'],
 833                     'webpage_url_basename': url_basename(ie_result['webpage_url']),
 834                     'extractor_key': ie_result['extractor_key'],
 835                 }
 836
 837                 reason = self._match_entry(entry, incomplete=True)
 838                 if reason is not None:
 839                     self.to_screen('[download] ' + reason)
 840                     continue
 841
 842                 entry_result = self.process_ie_result(entry,
 843                                                       download=download,
 844                                                       extra_info=extra)
 845                 playlist_results.append(entry_result)
 846             ie_result['entries'] = playlist_results
 847             self.to_screen('[download] Finished downloading playlist: %s' % playlist)
 848             return ie_result
 849         elif result_type == 'compat_list':
 850             self.report_warning(
 851                 'Extractor %s returned a compat_list result. '
 852                 'It needs to be updated.' % ie_result.get('extractor'))
 853
 854             def _fixup(r):
 855                 self.add_extra_info(
 856                     r,
 857                     {
 858                         'extractor': ie_result['extractor'],
 859                         'webpage_url': ie_result['webpage_url'],
 860                         'webpage_url_basename': url_basename(ie_result['webpage_url']),
 861                         'extractor_key': ie_result['extractor_key'],
 862                     }
 863                 )
 864                 return r
 865             ie_result['entries'] = [
 866                 self.process_ie_result(_fixup(r), download, extra_info)
 867                 for r in ie_result['entries']
 868             ]
 869             return ie_result
 870         else:
 871             raise Exception('Invalid result type: %s' % result_type)
 872
 873     def _build_format_filter(self, filter_spec):
 874         " Returns a function to filter the formats according to the filter_spec "
 875
 876         OPERATORS = {
 877             '<': operator.lt,
 878             '<=': operator.le,
 879             '>': operator.gt,
 880             '>=': operator.ge,
 881             '=': operator.eq,
 882             '!=': operator.ne,
 883         }
 884         operator_rex = re.compile(r'''(?x)\s*
 885             (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
 886             \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
 887             (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
 888             $
 889             ''' % '|'.join(map(re.escape, OPERATORS.keys())))
 890         m = operator_rex.search(filter_spec)
 891         if m:
 892             try:
 893                 comparison_value = int(m.group('value'))
 894             except ValueError:
 895                 comparison_value = parse_filesize(m.group('value'))
 896                 if comparison_value is None:
 897                     comparison_value = parse_filesize(m.group('value') + 'B')
 898                 if comparison_value is None:
 899                     raise ValueError(
 900                         'Invalid value %r in format specification %r' % (
 901                             m.group('value'), filter_spec))
 902             op = OPERATORS[m.group('op')]
 903
 904         if not m:
 905             STR_OPERATORS = {
 906                 '=': operator.eq,
 907                 '!=': operator.ne,
 908                 '^=': lambda attr, value: attr.startswith(value),
 909                 '$=': lambda attr, value: attr.endswith(value),
 910                 '*=': lambda attr, value: value in attr,
 911             }
 912             str_operator_rex = re.compile(r'''(?x)
 913                 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
 914                 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
 915                 \s*(?P<value>[a-zA-Z0-9._-]+)
 916                 \s*$
 917                 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
 918             m = str_operator_rex.search(filter_spec)
 919             if m:
 920                 comparison_value = m.group('value')
 921                 op = STR_OPERATORS[m.group('op')]
 922
 923         if not m:
 924             raise ValueError('Invalid filter specification %r' % filter_spec)
 925
 926         def _filter(f):
 927             actual_value = f.get(m.group('key'))
 928             if actual_value is None:
 929                 return m.group('none_inclusive')
 930             return op(actual_value, comparison_value)
 931         return _filter
 932
 933     def build_format_selector(self, format_spec):
 934         def syntax_error(note, start):
 935             message = (
 936                 'Invalid format specification: '
 937                 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
 938             return SyntaxError(message)
 939
 940         PICKFIRST = 'PICKFIRST'
 941         MERGE = 'MERGE'
 942         SINGLE = 'SINGLE'
 943         GROUP = 'GROUP'
 944         FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
 945
 946         def _parse_filter(tokens):
 947             filter_parts = []
 948             for type, string, start, _, _ in tokens:
 949                 if type == tokenize.OP and string == ']':
 950                     return ''.join(filter_parts)
 951                 else:
 952                     filter_parts.append(string)
 953
 954         def _remove_unused_ops(tokens):
 955             # Remove operators that we don't use and join them with the surrounding strings
 956             # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
 957             ALLOWED_OPS = ('/', '+', ',', '(', ')')
 958             last_string, last_start, last_end, last_line = None, None, None, None
 959             for type, string, start, end, line in tokens:
 960                 if type == tokenize.OP and string == '[':
 961                     if last_string:
 962                         yield tokenize.NAME, last_string, last_start, last_end, last_line
 963                         last_string = None
 964                     yield type, string, start, end, line
 965                     # everything inside brackets will be handled by _parse_filter
 966                     for type, string, start, end, line in tokens:
 967                         yield type, string, start, end, line
 968                         if type == tokenize.OP and string == ']':
 969                             break
 970                 elif type == tokenize.OP and string in ALLOWED_OPS:
 971                     if last_string:
 972                         yield tokenize.NAME, last_string, last_start, last_end, last_line
 973                         last_string = None
 974                     yield type, string, start, end, line
 975                 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
 976                     if not last_string:
 977                         last_string = string
 978                         last_start = start
 979                         last_end = end
 980                     else:
 981                         last_string += string
 982             if last_string:
 983                 yield tokenize.NAME, last_string, last_start, last_end, last_line
 984
 985         def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
 986             selectors = []
 987             current_selector = None
 988             for type, string, start, _, _ in tokens:
 989                 # ENCODING is only defined in python 3.x
 990                 if type == getattr(tokenize, 'ENCODING', None):
 991                     continue
 992                 elif type in [tokenize.NAME, tokenize.NUMBER]:
 993                     current_selector = FormatSelector(SINGLE, string, [])
 994                 elif type == tokenize.OP:
 995                     if string == ')':
 996                         if not inside_group:
 997                             # ')' will be handled by the parentheses group
 998                             tokens.restore_last_token()
 999                         break
1000                     elif inside_merge and string in ['/', ',']:
1001                         tokens.restore_last_token()
1002                         break
1003                     elif inside_choice and string == ',':
1004                         tokens.restore_last_token()
1005                         break
1006                     elif string == ',':
1007                         if not current_selector:
1008                             raise syntax_error('"," must follow a format selector', start)
1009                         selectors.append(current_selector)
1010                         current_selector = None
1011                     elif string == '/':
1012                         if not current_selector:
1013                             raise syntax_error('"/" must follow a format selector', start)
1014                         first_choice = current_selector
1015                         second_choice = _parse_format_selection(tokens, inside_choice=True)
1016                         current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1017                     elif string == '[':
1018                         if not current_selector:
1019                             current_selector = FormatSelector(SINGLE, 'best', [])
1020                         format_filter = _parse_filter(tokens)
1021                         current_selector.filters.append(format_filter)
1022                     elif string == '(':
1023                         if current_selector:
1024                             raise syntax_error('Unexpected "("', start)
1025                         group = _parse_format_selection(tokens, inside_group=True)
1026                         current_selector = FormatSelector(GROUP, group, [])
1027                     elif string == '+':
1028                         video_selector = current_selector
1029                         audio_selector = _parse_format_selection(tokens, inside_merge=True)
1030                         if not video_selector or not audio_selector:
1031                             raise syntax_error('"+" must be between two format selectors', start)
1032                         current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1033                     else:
1034                         raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1035                 elif type == tokenize.ENDMARKER:
1036                     break
1037             if current_selector:
1038                 selectors.append(current_selector)
1039             return selectors
1040
1041         def _build_selector_function(selector):
1042             if isinstance(selector, list):
1043                 fs = [_build_selector_function(s) for s in selector]
1044
1045                 def selector_function(formats):
1046                     for f in fs:
1047                         for format in f(formats):
1048                             yield format
1049                 return selector_function
1050             elif selector.type == GROUP:
1051                 selector_function = _build_selector_function(selector.selector)
1052             elif selector.type == PICKFIRST:
1053                 fs = [_build_selector_function(s) for s in selector.selector]
1054
1055                 def selector_function(formats):
1056                     for f in fs:
1057                         picked_formats = list(f(formats))
1058                         if picked_formats:
1059                             return picked_formats
1060                     return []
1061             elif selector.type == SINGLE:
1062                 format_spec = selector.selector
1063
1064                 def selector_function(formats):
1065                     formats = list(formats)
1066                     if not formats:
1067                         return
1068                     if format_spec == 'all':
1069                         for f in formats:
1070                             yield f
1071                     elif format_spec in ['best', 'worst', None]:
1072                         format_idx = 0 if format_spec == 'worst' else -1
1073                         audiovideo_formats = [
1074                             f for f in formats
1075                             if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1076                         if audiovideo_formats:
1077                             yield audiovideo_formats[format_idx]
1078                         # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1079                         elif (all(f.get('acodec') != 'none' for f in formats) or
1080                               all(f.get('vcodec') != 'none' for f in formats)):
1081                             yield formats[format_idx]
1082                     elif format_spec == 'bestaudio':
1083                         audio_formats = [
1084                             f for f in formats
1085                             if f.get('vcodec') == 'none']
1086                         if audio_formats:
1087                             yield audio_formats[-1]
1088                     elif format_spec == 'worstaudio':
1089                         audio_formats = [
1090                             f for f in formats
1091                             if f.get('vcodec') == 'none']
1092                         if audio_formats:
1093                             yield audio_formats[0]
1094                     elif format_spec == 'bestvideo':
1095                         video_formats = [
1096                             f for f in formats
1097                             if f.get('acodec') == 'none']
1098                         if video_formats:
1099                             yield video_formats[-1]
1100                     elif format_spec == 'worstvideo':
1101                         video_formats = [
1102                             f for f in formats
1103                             if f.get('acodec') == 'none']
1104                         if video_formats:
1105                             yield video_formats[0]
1106                     else:
1107                         extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1108                         if format_spec in extensions:
1109                             filter_f = lambda f: f['ext'] == format_spec
1110                         else:
1111                             filter_f = lambda f: f['format_id'] == format_spec
1112                         matches = list(filter(filter_f, formats))
1113                         if matches:
1114                             yield matches[-1]
1115             elif selector.type == MERGE:
1116                 def _merge(formats_info):
1117                     format_1, format_2 = [f['format_id'] for f in formats_info]
1118                     # The first format must contain the video and the
1119                     # second the audio
1120                     if formats_info[0].get('vcodec') == 'none':
1121                         self.report_error('The first format must '
1122                                           'contain the video, try using '
1123                                           '"-f %s+%s"' % (format_2, format_1))
1124                         return
1125                     # Formats must be opposite (video+audio)
1126                     if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1127                         self.report_error(
1128                             'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1129                             % (format_1, format_2))
1130                         return
1131                     output_ext = (
1132                         formats_info[0]['ext']
1133                         if self.params.get('merge_output_format') is None
1134                         else self.params['merge_output_format'])
1135                     return {
1136                         'requested_formats': formats_info,
1137                         'format': '%s+%s' % (formats_info[0].get('format'),
1138                                              formats_info[1].get('format')),
1139                         'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1140                                                 formats_info[1].get('format_id')),
1141                         'width': formats_info[0].get('width'),
1142                         'height': formats_info[0].get('height'),
1143                         'resolution': formats_info[0].get('resolution'),
1144                         'fps': formats_info[0].get('fps'),
1145                         'vcodec': formats_info[0].get('vcodec'),
1146                         'vbr': formats_info[0].get('vbr'),
1147                         'stretched_ratio': formats_info[0].get('stretched_ratio'),
1148                         'acodec': formats_info[1].get('acodec'),
1149                         'abr': formats_info[1].get('abr'),
1150                         'ext': output_ext,
1151                     }
1152                 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1153
1154                 def selector_function(formats):
1155                     formats = list(formats)
1156                     for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1157                         yield _merge(pair)
1158
1159             filters = [self._build_format_filter(f) for f in selector.filters]
1160
1161             def final_selector(formats):
1162                 for _filter in filters:
1163                     formats = list(filter(_filter, formats))
1164                 return selector_function(formats)
1165             return final_selector
1166
1167         stream = io.BytesIO(format_spec.encode('utf-8'))
1168         try:
1169             tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1170         except tokenize.TokenError:
1171             raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1172
1173         class TokenIterator(object):
1174             def __init__(self, tokens):
1175                 self.tokens = tokens
1176                 self.counter = 0
1177
1178             def __iter__(self):
1179                 return self
1180
1181             def __next__(self):
1182                 if self.counter >= len(self.tokens):
1183                     raise StopIteration()
1184                 value = self.tokens[self.counter]
1185                 self.counter += 1
1186                 return value
1187
1188             next = __next__
1189
1190             def restore_last_token(self):
1191                 self.counter -= 1
1192
1193         parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1194         return _build_selector_function(parsed_selector)
1195
1196     def _calc_headers(self, info_dict):
1197         res = std_headers.copy()
1198
1199         add_headers = info_dict.get('http_headers')
1200         if add_headers:
1201             res.update(add_headers)
1202
1203         cookies = self._calc_cookies(info_dict)
1204         if cookies:
1205             res['Cookie'] = cookies
1206
1207         return res
1208
1209     def _calc_cookies(self, info_dict):
1210         pr = sanitized_Request(info_dict['url'])
1211         self.cookiejar.add_cookie_header(pr)
1212         return pr.get_header('Cookie')
1213
1214     def process_video_result(self, info_dict, download=True):
1215         assert info_dict.get('_type', 'video') == 'video'
1216
1217         if 'id' not in info_dict:
1218             raise ExtractorError('Missing "id" field in extractor result')
1219         if 'title' not in info_dict:
1220             raise ExtractorError('Missing "title" field in extractor result')
1221
1222         if 'playlist' not in info_dict:
1223             # It isn't part of a playlist
1224             info_dict['playlist'] = None
1225             info_dict['playlist_index'] = None
1226
1227         thumbnails = info_dict.get('thumbnails')
1228         if thumbnails is None:
1229             thumbnail = info_dict.get('thumbnail')
1230             if thumbnail:
1231                 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1232         if thumbnails:
1233             thumbnails.sort(key=lambda t: (
1234                 t.get('preference'), t.get('width'), t.get('height'),
1235                 t.get('id'), t.get('url')))
1236             for i, t in enumerate(thumbnails):
1237                 t['url'] = sanitize_url(t['url'])
1238                 if t.get('width') and t.get('height'):
1239                     t['resolution'] = '%dx%d' % (t['width'], t['height'])
1240                 if t.get('id') is None:
1241                     t['id'] = '%d' % i
1242
1243         if self.params.get('list_thumbnails'):
1244             self.list_thumbnails(info_dict)
1245             return
1246
1247         thumbnail = info_dict.get('thumbnail')
1248         if thumbnail:
1249             info_dict['thumbnail'] = sanitize_url(thumbnail)
1250         elif thumbnails:
1251             info_dict['thumbnail'] = thumbnails[-1]['url']
1252
1253         if 'display_id' not in info_dict and 'id' in info_dict:
1254             info_dict['display_id'] = info_dict['id']
1255
1256         if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1257             # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1258             # see http://bugs.python.org/issue1646728)
1259             try:
1260                 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1261                 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1262             except (ValueError, OverflowError, OSError):
1263                 pass
1264
1265         # Auto generate title fields corresponding to the *_number fields when missing
1266         # in order to always have clean titles. This is very common for TV series.
1267         for field in ('chapter', 'season', 'episode'):
1268             if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1269                 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1270
1271         subtitles = info_dict.get('subtitles')
1272         if subtitles:
1273             for _, subtitle in subtitles.items():
1274                 for subtitle_format in subtitle:
1275                     if subtitle_format.get('url'):
1276                         subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1277                     if 'ext' not in subtitle_format:
1278                         subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1279
1280         if self.params.get('listsubtitles', False):
1281             if 'automatic_captions' in info_dict:
1282                 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1283             self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1284             return
1285         info_dict['requested_subtitles'] = self.process_subtitles(
1286             info_dict['id'], subtitles,
1287             info_dict.get('automatic_captions'))
1288
1289         # We now pick which formats have to be downloaded
1290         if info_dict.get('formats') is None:
1291             # There's only one format available
1292             formats = [info_dict]
1293         else:
1294             formats = info_dict['formats']
1295
1296         if not formats:
1297             raise ExtractorError('No video formats found!')
1298
1299         formats_dict = {}
1300
1301         # We check that all the formats have the format and format_id fields
1302         for i, format in enumerate(formats):
1303             if 'url' not in format:
1304                 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1305
1306             format['url'] = sanitize_url(format['url'])
1307
1308             if format.get('format_id') is None:
1309                 format['format_id'] = compat_str(i)
1310             else:
1311                 # Sanitize format_id from characters used in format selector expression
1312                 format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
1313             format_id = format['format_id']
1314             if format_id not in formats_dict:
1315                 formats_dict[format_id] = []
1316             formats_dict[format_id].append(format)
1317
1318         # Make sure all formats have unique format_id
1319         for format_id, ambiguous_formats in formats_dict.items():
1320             if len(ambiguous_formats) > 1:
1321                 for i, format in enumerate(ambiguous_formats):
1322                     format['format_id'] = '%s-%d' % (format_id, i)
1323
1324         for i, format in enumerate(formats):
1325             if format.get('format') is None:
1326                 format['format'] = '{id} - {res}{note}'.format(
1327                     id=format['format_id'],
1328                     res=self.format_resolution(format),
1329                     note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1330                 )
1331             # Automatically determine file extension if missing
1332             if 'ext' not in format:
1333                 format['ext'] = determine_ext(format['url']).lower()
1334             # Automatically determine protocol if missing (useful for format
1335             # selection purposes)
1336             if 'protocol' not in format:
1337                 format['protocol'] = determine_protocol(format)
1338             # Add HTTP headers, so that external programs can use them from the
1339             # json output
1340             full_format_info = info_dict.copy()
1341             full_format_info.update(format)
1342             format['http_headers'] = self._calc_headers(full_format_info)
1343
1344         # TODO Central sorting goes here
1345
1346         if formats[0] is not info_dict:
1347             # only set the 'formats' fields if the original info_dict list them
1348             # otherwise we end up with a circular reference, the first (and unique)
1349             # element in the 'formats' field in info_dict is info_dict itself,
1350             # which can't be exported to json
1351             info_dict['formats'] = formats
1352         if self.params.get('listformats'):
1353             self.list_formats(info_dict)
1354             return
1355
1356         req_format = self.params.get('format')
1357         if req_format is None:
1358             req_format_list = []
1359             if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1360                     not info_dict.get('is_live')):
1361                 merger = FFmpegMergerPP(self)
1362                 if merger.available and merger.can_merge():
1363                     req_format_list.append('bestvideo+bestaudio')
1364             req_format_list.append('best')
1365             req_format = '/'.join(req_format_list)
1366         format_selector = self.build_format_selector(req_format)
1367         formats_to_download = list(format_selector(formats))
1368         if not formats_to_download:
1369             raise ExtractorError('requested format not available',
1370                                  expected=True)
1371
1372         if download:
1373             if len(formats_to_download) > 1:
1374                 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1375             for format in formats_to_download:
1376                 new_info = dict(info_dict)
1377                 new_info.update(format)
1378                 self.process_info(new_info)
1379         # We update the info dict with the best quality format (backwards compatibility)
1380         info_dict.update(formats_to_download[-1])
1381         return info_dict
1382
1383     def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1384         """Select the requested subtitles and their format"""
1385         available_subs = {}
1386         if normal_subtitles and self.params.get('writesubtitles'):
1387             available_subs.update(normal_subtitles)
1388         if automatic_captions and self.params.get('writeautomaticsub'):
1389             for lang, cap_info in automatic_captions.items():
1390                 if lang not in available_subs:
1391                     available_subs[lang] = cap_info
1392
1393         if (not self.params.get('writesubtitles') and not
1394                 self.params.get('writeautomaticsub') or not
1395                 available_subs):
1396             return None
1397
1398         if self.params.get('allsubtitles', False):
1399             requested_langs = available_subs.keys()
1400         else:
1401             if self.params.get('subtitleslangs', False):
1402                 requested_langs = self.params.get('subtitleslangs')
1403             elif 'en' in available_subs:
1404                 requested_langs = ['en']
1405             else:
1406                 requested_langs = [list(available_subs.keys())[0]]
1407
1408         formats_query = self.params.get('subtitlesformat', 'best')
1409         formats_preference = formats_query.split('/') if formats_query else []
1410         subs = {}
1411         for lang in requested_langs:
1412             formats = available_subs.get(lang)
1413             if formats is None:
1414                 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1415                 continue
1416             for ext in formats_preference:
1417                 if ext == 'best':
1418                     f = formats[-1]
1419                     break
1420                 matches = list(filter(lambda f: f['ext'] == ext, formats))
1421                 if matches:
1422                     f = matches[-1]
1423                     break
1424             else:
1425                 f = formats[-1]
1426                 self.report_warning(
1427                     'No subtitle format found matching "%s" for language %s, '
1428                     'using %s' % (formats_query, lang, f['ext']))
1429             subs[lang] = f
1430         return subs
1431
1432     def process_info(self, info_dict):
1433         """Process a single resolved IE result."""
1434
1435         assert info_dict.get('_type', 'video') == 'video'
1436
1437         max_downloads = self.params.get('max_downloads')
1438         if max_downloads is not None:
1439             if self._num_downloads >= int(max_downloads):
1440                 raise MaxDownloadsReached()
1441
1442         info_dict['fulltitle'] = info_dict['title']
1443         if len(info_dict['title']) > 200:
1444             info_dict['title'] = info_dict['title'][:197] + '...'
1445
1446         if 'format' not in info_dict:
1447             info_dict['format'] = info_dict['ext']
1448
1449         reason = self._match_entry(info_dict, incomplete=False)
1450         if reason is not None:
1451             self.to_screen('[download] ' + reason)
1452             return
1453
1454         self._num_downloads += 1
1455
1456         info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1457
1458         # Forced printings
1459         if self.params.get('forcetitle', False):
1460             self.to_stdout(info_dict['fulltitle'])
1461         if self.params.get('forceid', False):
1462             self.to_stdout(info_dict['id'])
1463         if self.params.get('forceurl', False):
1464             if info_dict.get('requested_formats') is not None:
1465                 for f in info_dict['requested_formats']:
1466                     self.to_stdout(f['url'] + f.get('play_path', ''))
1467             else:
1468                 # For RTMP URLs, also include the playpath
1469                 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1470         if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1471             self.to_stdout(info_dict['thumbnail'])
1472         if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1473             self.to_stdout(info_dict['description'])
1474         if self.params.get('forcefilename', False) and filename is not None:
1475             self.to_stdout(filename)
1476         if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1477             self.to_stdout(formatSeconds(info_dict['duration']))
1478         if self.params.get('forceformat', False):
1479             self.to_stdout(info_dict['format'])
1480         if self.params.get('forcejson', False):
1481             self.to_stdout(json.dumps(info_dict))
1482
1483         # Do nothing else if in simulate mode
1484         if self.params.get('simulate', False):
1485             return
1486
1487         if filename is None:
1488             return
1489
1490         try:
1491             dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1492             if dn and not os.path.exists(dn):
1493                 os.makedirs(dn)
1494         except (OSError, IOError) as err:
1495             self.report_error('unable to create directory ' + error_to_compat_str(err))
1496             return
1497
1498         if self.params.get('writedescription', False):
1499             descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1500             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1501                 self.to_screen('[info] Video description is already present')
1502             elif info_dict.get('description') is None:
1503                 self.report_warning('There\'s no description to write.')
1504             else:
1505                 try:
1506                     self.to_screen('[info] Writing video description to: ' + descfn)
1507                     with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1508                         descfile.write(info_dict['description'])
1509                 except (OSError, IOError):
1510                     self.report_error('Cannot write description file ' + descfn)
1511                     return
1512
1513         if self.params.get('writeannotations', False):
1514             annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1515             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1516                 self.to_screen('[info] Video annotations are already present')
1517             else:
1518                 try:
1519                     self.to_screen('[info] Writing video annotations to: ' + annofn)
1520                     with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1521                         annofile.write(info_dict['annotations'])
1522                 except (KeyError, TypeError):
1523                     self.report_warning('There are no annotations to write.')
1524                 except (OSError, IOError):
1525                     self.report_error('Cannot write annotations file: ' + annofn)
1526                     return
1527
1528         subtitles_are_requested = any([self.params.get('writesubtitles', False),
1529                                        self.params.get('writeautomaticsub')])
1530
1531         if subtitles_are_requested and info_dict.get('requested_subtitles'):
1532             # subtitles download errors are already managed as troubles in relevant IE
1533             # that way it will silently go on when used with unsupporting IE
1534             subtitles = info_dict['requested_subtitles']
1535             ie = self.get_info_extractor(info_dict['extractor_key'])
1536             for sub_lang, sub_info in subtitles.items():
1537                 sub_format = sub_info['ext']
1538                 if sub_info.get('data') is not None:
1539                     sub_data = sub_info['data']
1540                 else:
1541                     try:
1542                         sub_data = ie._download_webpage(
1543                             sub_info['url'], info_dict['id'], note=False)
1544                     except ExtractorError as err:
1545                         self.report_warning('Unable to download subtitle for "%s": %s' %
1546                                             (sub_lang, error_to_compat_str(err.cause)))
1547                         continue
1548                 try:
1549                     sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1550                     if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1551                         self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1552                     else:
1553                         self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1554                         with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1555                             subfile.write(sub_data)
1556                 except (OSError, IOError):
1557                     self.report_error('Cannot write subtitles file ' + sub_filename)
1558                     return
1559
1560         if self.params.get('writeinfojson', False):
1561             infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1562             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1563                 self.to_screen('[info] Video description metadata is already present')
1564             else:
1565                 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1566                 try:
1567                     write_json_file(self.filter_requested_info(info_dict), infofn)
1568                 except (OSError, IOError):
1569                     self.report_error('Cannot write metadata to JSON file ' + infofn)
1570                     return
1571
1572         self._write_thumbnails(info_dict, filename)
1573
1574         if not self.params.get('skip_download', False):
1575             try:
1576                 def dl(name, info):
1577                     fd = get_suitable_downloader(info, self.params)(self, self.params)
1578                     for ph in self._progress_hooks:
1579                         fd.add_progress_hook(ph)
1580                     if self.params.get('verbose'):
1581                         self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1582                     return fd.download(name, info)
1583
1584                 if info_dict.get('requested_formats') is not None:
1585                     downloaded = []
1586                     success = True
1587                     merger = FFmpegMergerPP(self)
1588                     if not merger.available:
1589                         postprocessors = []
1590                         self.report_warning('You have requested multiple '
1591                                             'formats but ffmpeg or avconv are not installed.'
1592                                             ' The formats won\'t be merged.')
1593                     else:
1594                         postprocessors = [merger]
1595
1596                     def compatible_formats(formats):
1597                         video, audio = formats
1598                         # Check extension
1599                         video_ext, audio_ext = audio.get('ext'), video.get('ext')
1600                         if video_ext and audio_ext:
1601                             COMPATIBLE_EXTS = (
1602                                 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1603                                 ('webm')
1604                             )
1605                             for exts in COMPATIBLE_EXTS:
1606                                 if video_ext in exts and audio_ext in exts:
1607                                     return True
1608                         # TODO: Check acodec/vcodec
1609                         return False
1610
1611                     filename_real_ext = os.path.splitext(filename)[1][1:]
1612                     filename_wo_ext = (
1613                         os.path.splitext(filename)[0]
1614                         if filename_real_ext == info_dict['ext']
1615                         else filename)
1616                     requested_formats = info_dict['requested_formats']
1617                     if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1618                         info_dict['ext'] = 'mkv'
1619                         self.report_warning(
1620                             'Requested formats are incompatible for merge and will be merged into mkv.')
1621                     # Ensure filename always has a correct extension for successful merge
1622                     filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1623                     if os.path.exists(encodeFilename(filename)):
1624                         self.to_screen(
1625                             '[download] %s has already been downloaded and '
1626                             'merged' % filename)
1627                     else:
1628                         for f in requested_formats:
1629                             new_info = dict(info_dict)
1630                             new_info.update(f)
1631                             fname = self.prepare_filename(new_info)
1632                             fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1633                             downloaded.append(fname)
1634                             partial_success = dl(fname, new_info)
1635                             success = success and partial_success
1636                         info_dict['__postprocessors'] = postprocessors
1637                         info_dict['__files_to_merge'] = downloaded
1638                 else:
1639                     # Just a single file
1640                     success = dl(filename, info_dict)
1641             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1642                 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1643                 return
1644             except (OSError, IOError) as err:
1645                 raise UnavailableVideoError(err)
1646             except (ContentTooShortError, ) as err:
1647                 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1648                 return
1649
1650             if success and filename != '-':
1651                 # Fixup content
1652                 fixup_policy = self.params.get('fixup')
1653                 if fixup_policy is None:
1654                     fixup_policy = 'detect_or_warn'
1655
1656                 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1657
1658                 stretched_ratio = info_dict.get('stretched_ratio')
1659                 if stretched_ratio is not None and stretched_ratio != 1:
1660                     if fixup_policy == 'warn':
1661                         self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1662                             info_dict['id'], stretched_ratio))
1663                     elif fixup_policy == 'detect_or_warn':
1664                         stretched_pp = FFmpegFixupStretchedPP(self)
1665                         if stretched_pp.available:
1666                             info_dict.setdefault('__postprocessors', [])
1667                             info_dict['__postprocessors'].append(stretched_pp)
1668                         else:
1669                             self.report_warning(
1670                                 '%s: Non-uniform pixel ratio (%s). %s'
1671                                 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1672                     else:
1673                         assert fixup_policy in ('ignore', 'never')
1674
1675                 if (info_dict.get('requested_formats') is None and
1676                         info_dict.get('container') == 'm4a_dash'):
1677                     if fixup_policy == 'warn':
1678                         self.report_warning(
1679                             '%s: writing DASH m4a. '
1680                             'Only some players support this container.'
1681                             % info_dict['id'])
1682                     elif fixup_policy == 'detect_or_warn':
1683                         fixup_pp = FFmpegFixupM4aPP(self)
1684                         if fixup_pp.available:
1685                             info_dict.setdefault('__postprocessors', [])
1686                             info_dict['__postprocessors'].append(fixup_pp)
1687                         else:
1688                             self.report_warning(
1689                                 '%s: writing DASH m4a. '
1690                                 'Only some players support this container. %s'
1691                                 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1692                     else:
1693                         assert fixup_policy in ('ignore', 'never')
1694
1695                 if (info_dict.get('protocol') == 'm3u8_native' or
1696                         info_dict.get('protocol') == 'm3u8' and
1697                         self.params.get('hls_prefer_native')):
1698                     if fixup_policy == 'warn':
1699                         self.report_warning('%s: malformated aac bitstream.' % (
1700                             info_dict['id']))
1701                     elif fixup_policy == 'detect_or_warn':
1702                         fixup_pp = FFmpegFixupM3u8PP(self)
1703                         if fixup_pp.available:
1704                             info_dict.setdefault('__postprocessors', [])
1705                             info_dict['__postprocessors'].append(fixup_pp)
1706                         else:
1707                             self.report_warning(
1708                                 '%s: malformated aac bitstream. %s'
1709                                 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1710                     else:
1711                         assert fixup_policy in ('ignore', 'never')
1712
1713                 try:
1714                     self.post_process(filename, info_dict)
1715                 except (PostProcessingError) as err:
1716                     self.report_error('postprocessing: %s' % str(err))
1717                     return
1718                 self.record_download_archive(info_dict)
1719
1720     def download(self, url_list):
1721         """Download a given list of URLs."""
1722         outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1723         if (len(url_list) > 1 and
1724                 '%' not in outtmpl and
1725                 self.params.get('max_downloads') != 1):
1726             raise SameFileError(outtmpl)
1727
1728         for url in url_list:
1729             try:
1730                 # It also downloads the videos
1731                 res = self.extract_info(
1732                     url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1733             except UnavailableVideoError:
1734                 self.report_error('unable to download video')
1735             except MaxDownloadsReached:
1736                 self.to_screen('[info] Maximum number of downloaded files reached.')
1737                 raise
1738             else:
1739                 if self.params.get('dump_single_json', False):
1740                     self.to_stdout(json.dumps(res))
1741
1742         return self._download_retcode
1743
1744     def download_with_info_file(self, info_filename):
1745         with contextlib.closing(fileinput.FileInput(
1746                 [info_filename], mode='r',
1747                 openhook=fileinput.hook_encoded('utf-8'))) as f:
1748             # FileInput doesn't have a read method, we can't call json.load
1749             info = self.filter_requested_info(json.loads('\n'.join(f)))
1750         try:
1751             self.process_ie_result(info, download=True)
1752         except DownloadError:
1753             webpage_url = info.get('webpage_url')
1754             if webpage_url is not None:
1755                 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1756                 return self.download([webpage_url])
1757             else:
1758                 raise
1759         return self._download_retcode
1760
1761     @staticmethod
1762     def filter_requested_info(info_dict):
1763         return dict(
1764             (k, v) for k, v in info_dict.items()
1765             if k not in ['requested_formats', 'requested_subtitles'])
1766
1767     def post_process(self, filename, ie_info):
1768         """Run all the postprocessors on the given file."""
1769         info = dict(ie_info)
1770         info['filepath'] = filename
1771         pps_chain = []
1772         if ie_info.get('__postprocessors') is not None:
1773             pps_chain.extend(ie_info['__postprocessors'])
1774         pps_chain.extend(self._pps)
1775         for pp in pps_chain:
1776             files_to_delete = []
1777             try:
1778                 files_to_delete, info = pp.run(info)
1779             except PostProcessingError as e:
1780                 self.report_error(e.msg)
1781             if files_to_delete and not self.params.get('keepvideo', False):
1782                 for old_filename in files_to_delete:
1783                     self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1784                     try:
1785                         os.remove(encodeFilename(old_filename))
1786                     except (IOError, OSError):
1787                         self.report_warning('Unable to remove downloaded original file')
1788
1789     def _make_archive_id(self, info_dict):
1790         # Future-proof against any change in case
1791         # and backwards compatibility with prior versions
1792         extractor = info_dict.get('extractor_key')
1793         if extractor is None:
1794             if 'id' in info_dict:
1795                 extractor = info_dict.get('ie_key')  # key in a playlist
1796         if extractor is None:
1797             return None  # Incomplete video information
1798         return extractor.lower() + ' ' + info_dict['id']
1799
1800     def in_download_archive(self, info_dict):
1801         fn = self.params.get('download_archive')
1802         if fn is None:
1803             return False
1804
1805         vid_id = self._make_archive_id(info_dict)
1806         if vid_id is None:
1807             return False  # Incomplete video information
1808
1809         try:
1810             with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1811                 for line in archive_file:
1812                     if line.strip() == vid_id:
1813                         return True
1814         except IOError as ioe:
1815             if ioe.errno != errno.ENOENT:
1816                 raise
1817         return False
1818
1819     def record_download_archive(self, info_dict):
1820         fn = self.params.get('download_archive')
1821         if fn is None:
1822             return
1823         vid_id = self._make_archive_id(info_dict)
1824         assert vid_id
1825         with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1826             archive_file.write(vid_id + '\n')
1827
1828     @staticmethod
1829     def format_resolution(format, default='unknown'):
1830         if format.get('vcodec') == 'none':
1831             return 'audio only'
1832         if format.get('resolution') is not None:
1833             return format['resolution']
1834         if format.get('height') is not None:
1835             if format.get('width') is not None:
1836                 res = '%sx%s' % (format['width'], format['height'])
1837             else:
1838                 res = '%sp' % format['height']
1839         elif format.get('width') is not None:
1840             res = '%dx?' % format['width']
1841         else:
1842             res = default
1843         return res
1844
1845     def _format_note(self, fdict):
1846         res = ''
1847         if fdict.get('ext') in ['f4f', 'f4m']:
1848             res += '(unsupported) '
1849         if fdict.get('language'):
1850             if res:
1851                 res += ' '
1852             res += '[%s] ' % fdict['language']
1853         if fdict.get('format_note') is not None:
1854             res += fdict['format_note'] + ' '
1855         if fdict.get('tbr') is not None:
1856             res += '%4dk ' % fdict['tbr']
1857         if fdict.get('container') is not None:
1858             if res:
1859                 res += ', '
1860             res += '%s container' % fdict['container']
1861         if (fdict.get('vcodec') is not None and
1862                 fdict.get('vcodec') != 'none'):
1863             if res:
1864                 res += ', '
1865             res += fdict['vcodec']
1866             if fdict.get('vbr') is not None:
1867                 res += '@'
1868         elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1869             res += 'video@'
1870         if fdict.get('vbr') is not None:
1871             res += '%4dk' % fdict['vbr']
1872         if fdict.get('fps') is not None:
1873             if res:
1874                 res += ', '
1875             res += '%sfps' % fdict['fps']
1876         if fdict.get('acodec') is not None:
1877             if res:
1878                 res += ', '
1879             if fdict['acodec'] == 'none':
1880                 res += 'video only'
1881             else:
1882                 res += '%-5s' % fdict['acodec']
1883         elif fdict.get('abr') is not None:
1884             if res:
1885                 res += ', '
1886             res += 'audio'
1887         if fdict.get('abr') is not None:
1888             res += '@%3dk' % fdict['abr']
1889         if fdict.get('asr') is not None:
1890             res += ' (%5dHz)' % fdict['asr']
1891         if fdict.get('filesize') is not None:
1892             if res:
1893                 res += ', '
1894             res += format_bytes(fdict['filesize'])
1895         elif fdict.get('filesize_approx') is not None:
1896             if res:
1897                 res += ', '
1898             res += '~' + format_bytes(fdict['filesize_approx'])
1899         return res
1900
1901     def list_formats(self, info_dict):
1902         formats = info_dict.get('formats', [info_dict])
1903         table = [
1904             [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1905             for f in formats
1906             if f.get('preference') is None or f['preference'] >= -1000]
1907         if len(formats) > 1:
1908             table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1909
1910         header_line = ['format code', 'extension', 'resolution', 'note']
1911         self.to_screen(
1912             '[info] Available formats for %s:\n%s' %
1913             (info_dict['id'], render_table(header_line, table)))
1914
1915     def list_thumbnails(self, info_dict):
1916         thumbnails = info_dict.get('thumbnails')
1917         if not thumbnails:
1918             self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
1919             return
1920
1921         self.to_screen(
1922             '[info] Thumbnails for %s:' % info_dict['id'])
1923         self.to_screen(render_table(
1924             ['ID', 'width', 'height', 'URL'],
1925             [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1926
1927     def list_subtitles(self, video_id, subtitles, name='subtitles'):
1928         if not subtitles:
1929             self.to_screen('%s has no %s' % (video_id, name))
1930             return
1931         self.to_screen(
1932             'Available %s for %s:' % (name, video_id))
1933         self.to_screen(render_table(
1934             ['Language', 'formats'],
1935             [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1936                 for lang, formats in subtitles.items()]))
1937
1938     def urlopen(self, req):
1939         """ Start an HTTP download """
1940         if isinstance(req, compat_basestring):
1941             req = sanitized_Request(req)
1942         return self._opener.open(req, timeout=self._socket_timeout)
1943
1944     def print_debug_header(self):
1945         if not self.params.get('verbose'):
1946             return
1947
1948         if type('') is not compat_str:
1949             # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1950             self.report_warning(
1951                 'Your Python is broken! Update to a newer and supported version')
1952
1953         stdout_encoding = getattr(
1954             sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1955         encoding_str = (
1956             '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1957                 locale.getpreferredencoding(),
1958                 sys.getfilesystemencoding(),
1959                 stdout_encoding,
1960                 self.get_encoding()))
1961         write_string(encoding_str, encoding=None)
1962
1963         self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1964         if _LAZY_LOADER:
1965             self._write_string('[debug] Lazy loading extractors enabled' + '\n')
1966         try:
1967             sp = subprocess.Popen(
1968                 ['git', 'rev-parse', '--short', 'HEAD'],
1969                 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1970                 cwd=os.path.dirname(os.path.abspath(__file__)))
1971             out, err = sp.communicate()
1972             out = out.decode().strip()
1973             if re.match('[0-9a-f]+', out):
1974                 self._write_string('[debug] Git HEAD: ' + out + '\n')
1975         except Exception:
1976             try:
1977                 sys.exc_clear()
1978             except Exception:
1979                 pass
1980         self._write_string('[debug] Python version %s - %s\n' % (
1981             platform.python_version(), platform_name()))
1982
1983         exe_versions = FFmpegPostProcessor.get_versions(self)
1984         exe_versions['rtmpdump'] = rtmpdump_version()
1985         exe_str = ', '.join(
1986             '%s %s' % (exe, v)
1987             for exe, v in sorted(exe_versions.items())
1988             if v
1989         )
1990         if not exe_str:
1991             exe_str = 'none'
1992         self._write_string('[debug] exe versions: %s\n' % exe_str)
1993
1994         proxy_map = {}
1995         for handler in self._opener.handlers:
1996             if hasattr(handler, 'proxies'):
1997                 proxy_map.update(handler.proxies)
1998         self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1999
2000         if self.params.get('call_home', False):
2001             ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2002             self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2003             latest_version = self.urlopen(
2004                 'https://yt-dl.org/latest/version').read().decode('utf-8')
2005             if version_tuple(latest_version) > version_tuple(__version__):
2006                 self.report_warning(
2007                     'You are using an outdated version (newest version: %s)! '
2008                     'See https://yt-dl.org/update if you need help updating.' %
2009                     latest_version)
2010
2011     def _setup_opener(self):
2012         timeout_val = self.params.get('socket_timeout')
2013         self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2014
2015         opts_cookiefile = self.params.get('cookiefile')
2016         opts_proxy = self.params.get('proxy')
2017
2018         if opts_cookiefile is None:
2019             self.cookiejar = compat_cookiejar.CookieJar()
2020         else:
2021             self.cookiejar = compat_cookiejar.MozillaCookieJar(
2022                 opts_cookiefile)
2023             if os.access(opts_cookiefile, os.R_OK):
2024                 self.cookiejar.load()
2025
2026         cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2027         if opts_proxy is not None:
2028             if opts_proxy == '':
2029                 proxies = {}
2030             else:
2031                 proxies = {'http': opts_proxy, 'https': opts_proxy}
2032         else:
2033             proxies = compat_urllib_request.getproxies()
2034             # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2035             if 'http' in proxies and 'https' not in proxies:
2036                 proxies['https'] = proxies['http']
2037         proxy_handler = PerRequestProxyHandler(proxies)
2038
2039         debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2040         https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2041         ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2042         data_handler = compat_urllib_request_DataHandler()
2043
2044         # When passing our own FileHandler instance, build_opener won't add the
2045         # default FileHandler and allows us to disable the file protocol, which
2046         # can be used for malicious purposes (see
2047         # https://github.com/rg3/youtube-dl/issues/8227)
2048         file_handler = compat_urllib_request.FileHandler()
2049
2050         def file_open(*args, **kwargs):
2051             raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2052         file_handler.file_open = file_open
2053
2054         opener = compat_urllib_request.build_opener(
2055             proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2056
2057         # Delete the default user-agent header, which would otherwise apply in
2058         # cases where our custom HTTP handler doesn't come into play
2059         # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2060         opener.addheaders = []
2061         self._opener = opener
2062
2063     def encode(self, s):
2064         if isinstance(s, bytes):
2065             return s  # Already encoded
2066
2067         try:
2068             return s.encode(self.get_encoding())
2069         except UnicodeEncodeError as err:
2070             err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2071             raise
2072
2073     def get_encoding(self):
2074         encoding = self.params.get('encoding')
2075         if encoding is None:
2076             encoding = preferredencoding()
2077         return encoding
2078
2079     def _write_thumbnails(self, info_dict, filename):
2080         if self.params.get('writethumbnail', False):
2081             thumbnails = info_dict.get('thumbnails')
2082             if thumbnails:
2083                 thumbnails = [thumbnails[-1]]
2084         elif self.params.get('write_all_thumbnails', False):
2085             thumbnails = info_dict.get('thumbnails')
2086         else:
2087             return
2088
2089         if not thumbnails:
2090             # No thumbnails present, so return immediately
2091             return
2092
2093         for t in thumbnails:
2094             thumb_ext = determine_ext(t['url'], 'jpg')
2095             suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2096             thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2097             t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2098
2099             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2100                 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2101                                (info_dict['extractor'], info_dict['id'], thumb_display_id))
2102             else:
2103                 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2104                                (info_dict['extractor'], info_dict['id'], thumb_display_id))
2105                 try:
2106                     uf = self.urlopen(t['url'])
2107                     with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2108                         shutil.copyfileobj(uf, thumbf)
2109                     self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2110                                    (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2111                 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2112                     self.report_warning('Unable to download thumbnail "%s": %s' %
2113                                         (t['url'], error_to_compat_str(err)))