Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/YoutubeDL.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3
   4 from __future__ import absolute_import
   5
   6 import collections
   7 import errno
   8 import io
   9 import json
  10 import os
  11 import platform
  12 import re
  13 import shutil
  14 import subprocess
  15 import socket
  16 import sys
  17 import time
  18 import traceback
  19
  20 if os.name == 'nt':
  21     import ctypes
  22
  23 from .utils import (
  24     compat_cookiejar,
  25     compat_http_client,
  26     compat_str,
  27     compat_urllib_error,
  28     compat_urllib_request,
  29     ContentTooShortError,
  30     date_from_str,
  31     DateRange,
  32     determine_ext,
  33     DownloadError,
  34     encodeFilename,
  35     ExtractorError,
  36     format_bytes,
  37     formatSeconds,
  38     get_term_width,
  39     locked_file,
  40     make_HTTPS_handler,
  41     MaxDownloadsReached,
  42     PostProcessingError,
  43     platform_name,
  44     preferredencoding,
  45     SameFileError,
  46     sanitize_filename,
  47     subtitles_filename,
  48     takewhile_inclusive,
  49     UnavailableVideoError,
  50     url_basename,
  51     write_json_file,
  52     write_string,
  53     YoutubeDLHandler,
  54 )
  55 from .extractor import get_info_extractor, gen_extractors
  56 from .FileDownloader import FileDownloader
  57 from .version import __version__
  58
  59
  60 class YoutubeDL(object):
  61     """YoutubeDL class.
  62
  63     YoutubeDL objects are the ones responsible of downloading the
  64     actual video file and writing it to disk if the user has requested
  65     it, among some other tasks. In most cases there should be one per
  66     program. As, given a video URL, the downloader doesn't know how to
  67     extract all the needed information, task that InfoExtractors do, it
  68     has to pass the URL to one of them.
  69
  70     For this, YoutubeDL objects have a method that allows
  71     InfoExtractors to be registered in a given order. When it is passed
  72     a URL, the YoutubeDL object handles it to the first InfoExtractor it
  73     finds that reports being able to handle it. The InfoExtractor extracts
  74     all the information about the video or videos the URL refers to, and
  75     YoutubeDL process the extracted information, possibly using a File
  76     Downloader to download the video.
  77
  78     YoutubeDL objects accept a lot of parameters. In order not to saturate
  79     the object constructor with arguments, it receives a dictionary of
  80     options instead. These options are available through the params
  81     attribute for the InfoExtractors to use. The YoutubeDL also
  82     registers itself as the downloader in charge for the InfoExtractors
  83     that are added to it, so this is a "mutual registration".
  84
  85     Available options:
  86
  87     username:          Username for authentication purposes.
  88     password:          Password for authentication purposes.
  89     videopassword:     Password for acces a video.
  90     usenetrc:          Use netrc for authentication instead.
  91     verbose:           Print additional info to stdout.
  92     quiet:             Do not print messages to stdout.
  93     forceurl:          Force printing final URL.
  94     forcetitle:        Force printing title.
  95     forceid:           Force printing ID.
  96     forcethumbnail:    Force printing thumbnail URL.
  97     forcedescription:  Force printing description.
  98     forcefilename:     Force printing final filename.
  99     forceduration:     Force printing duration.
 100     forcejson:         Force printing info_dict as JSON.
 101     simulate:          Do not download the video files.
 102     format:            Video format code.
 103     format_limit:      Highest quality format to try.
 104     outtmpl:           Template for output names.
 105     restrictfilenames: Do not allow "&" and spaces in file names
 106     ignoreerrors:      Do not stop on download errors.
 107     nooverwrites:      Prevent overwriting files.
 108     playliststart:     Playlist item to start at.
 109     playlistend:       Playlist item to end at.
 110     matchtitle:        Download only matching titles.
 111     rejecttitle:       Reject downloads for matching titles.
 112     logger:            Log messages to a logging.Logger instance.
 113     logtostderr:       Log messages to stderr instead of stdout.
 114     writedescription:  Write the video description to a .description file
 115     writeinfojson:     Write the video description to a .info.json file
 116     writeannotations:  Write the video annotations to a .annotations.xml file
 117     writethumbnail:    Write the thumbnail image to a file
 118     writesubtitles:    Write the video subtitles to a file
 119     writeautomaticsub: Write the automatic subtitles to a file
 120     allsubtitles:      Downloads all the subtitles of the video
 121                        (requires writesubtitles or writeautomaticsub)
 122     listsubtitles:     Lists all available subtitles for the video
 123     subtitlesformat:   Subtitle format [srt/sbv/vtt] (default=srt)
 124     subtitleslangs:    List of languages of the subtitles to download
 125     keepvideo:         Keep the video file after post-processing
 126     daterange:         A DateRange object, download only if the upload_date is in the range.
 127     skip_download:     Skip the actual download of the video file
 128     cachedir:          Location of the cache files in the filesystem.
 129                        None to disable filesystem cache.
 130     noplaylist:        Download single video instead of a playlist if in doubt.
 131     age_limit:         An integer representing the user's age in years.
 132                        Unsuitable videos for the given age are skipped.
 133     min_views:         An integer representing the minimum view count the video
 134                        must have in order to not be skipped.
 135                        Videos without view count information are always
 136                        downloaded. None for no limit.
 137     max_views:         An integer representing the maximum view count.
 138                        Videos that are more popular than that are not
 139                        downloaded.
 140                        Videos without view count information are always
 141                        downloaded. None for no limit.
 142     download_archive:  File name of a file where all downloads are recorded.
 143                        Videos already present in the file are not downloaded
 144                        again.
 145     cookiefile:        File name where cookies should be read from and dumped to.
 146     nocheckcertificate:Do not verify SSL certificates
 147     proxy:             URL of the proxy server to use
 148     socket_timeout:    Time to wait for unresponsive hosts, in seconds
 149     bidi_workaround:   Work around buggy terminals without bidirectional text
 150                        support, using fridibi
 151
 152     The following parameters are not used by YoutubeDL itself, they are used by
 153     the FileDownloader:
 154     nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
 155     noresizebuffer, retries, continuedl, noprogress, consoletitle
 156     """
 157
 158     params = None
 159     _ies = []
 160     _pps = []
 161     _download_retcode = None
 162     _num_downloads = None
 163     _screen_file = None
 164
 165     def __init__(self, params=None):
 166         """Create a FileDownloader object with the given options."""
 167         self._ies = []
 168         self._ies_instances = {}
 169         self._pps = []
 170         self._progress_hooks = []
 171         self._download_retcode = 0
 172         self._num_downloads = 0
 173         self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
 174         self._err_file = sys.stderr
 175         self.params = {} if params is None else params
 176
 177         if params.get('bidi_workaround', False):
 178             try:
 179                 import pty
 180                 master, slave = pty.openpty()
 181                 width = get_term_width()
 182                 if width is None:
 183                     width_args = []
 184                 else:
 185                     width_args = ['-w', str(width)]
 186                 self._fribidi = subprocess.Popen(
 187                     ['fribidi', '-c', 'UTF-8'] + width_args,
 188                     stdin=subprocess.PIPE,
 189                     stdout=slave,
 190                     stderr=self._err_file)
 191                 self._fribidi_channel = os.fdopen(master, 'rb')
 192             except OSError as ose:
 193                 if ose.errno == 2:
 194                     self.report_warning(u'Could not find fribidi executable, ignoring --bidi-workaround . Make sure that  fribidi  is an executable file in one of the directories in your $PATH.')
 195                 else:
 196                     raise
 197
 198         if (sys.version_info >= (3,) and sys.platform != 'win32' and
 199                 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
 200                 and not params['restrictfilenames']):
 201             # On Python 3, the Unicode filesystem API will throw errors (#1474)
 202             self.report_warning(
 203                 u'Assuming --restrict-filenames since file system encoding '
 204                 u'cannot encode all charactes. '
 205                 u'Set the LC_ALL environment variable to fix this.')
 206             self.params['restrictfilenames'] = True
 207
 208         self.fd = FileDownloader(self, self.params)
 209
 210         if '%(stitle)s' in self.params.get('outtmpl', ''):
 211             self.report_warning(u'%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
 212
 213         self._setup_opener()
 214
 215     def add_info_extractor(self, ie):
 216         """Add an InfoExtractor object to the end of the list."""
 217         self._ies.append(ie)
 218         self._ies_instances[ie.ie_key()] = ie
 219         ie.set_downloader(self)
 220
 221     def get_info_extractor(self, ie_key):
 222         """
 223         Get an instance of an IE with name ie_key, it will try to get one from
 224         the _ies list, if there's no instance it will create a new one and add
 225         it to the extractor list.
 226         """
 227         ie = self._ies_instances.get(ie_key)
 228         if ie is None:
 229             ie = get_info_extractor(ie_key)()
 230             self.add_info_extractor(ie)
 231         return ie
 232
 233     def add_default_info_extractors(self):
 234         """
 235         Add the InfoExtractors returned by gen_extractors to the end of the list
 236         """
 237         for ie in gen_extractors():
 238             self.add_info_extractor(ie)
 239
 240     def add_post_processor(self, pp):
 241         """Add a PostProcessor object to the end of the chain."""
 242         self._pps.append(pp)
 243         pp.set_downloader(self)
 244
 245     def _bidi_workaround(self, message):
 246         if not hasattr(self, '_fribidi_channel'):
 247             return message
 248
 249         assert type(message) == type(u'')
 250         line_count = message.count(u'\n') + 1
 251         self._fribidi.stdin.write((message + u'\n').encode('utf-8'))
 252         self._fribidi.stdin.flush()
 253         res = u''.join(self._fribidi_channel.readline().decode('utf-8')
 254                        for _ in range(line_count))
 255         return res[:-len(u'\n')]
 256
 257     def to_screen(self, message, skip_eol=False):
 258         """Print message to stdout if not in quiet mode."""
 259         return self.to_stdout(message, skip_eol, check_quiet=True)
 260
 261     def to_stdout(self, message, skip_eol=False, check_quiet=False):
 262         """Print message to stdout if not in quiet mode."""
 263         if self.params.get('logger'):
 264             self.params['logger'].debug(message)
 265         elif not check_quiet or not self.params.get('quiet', False):
 266             message = self._bidi_workaround(message)
 267             terminator = [u'\n', u''][skip_eol]
 268             output = message + terminator
 269
 270             write_string(output, self._screen_file)
 271
 272     def to_stderr(self, message):
 273         """Print message to stderr."""
 274         assert type(message) == type(u'')
 275         if self.params.get('logger'):
 276             self.params['logger'].error(message)
 277         else:
 278             message = self._bidi_workaround(message)
 279             output = message + u'\n'
 280             write_string(output, self._err_file)
 281
 282     def to_console_title(self, message):
 283         if not self.params.get('consoletitle', False):
 284             return
 285         if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
 286             # c_wchar_p() might not be necessary if `message` is
 287             # already of type unicode()
 288             ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
 289         elif 'TERM' in os.environ:
 290             write_string(u'\033]0;%s\007' % message, self._screen_file)
 291
 292     def save_console_title(self):
 293         if not self.params.get('consoletitle', False):
 294             return
 295         if 'TERM' in os.environ:
 296             # Save the title on stack
 297             write_string(u'\033[22;0t', self._screen_file)
 298
 299     def restore_console_title(self):
 300         if not self.params.get('consoletitle', False):
 301             return
 302         if 'TERM' in os.environ:
 303             # Restore the title from stack
 304             write_string(u'\033[23;0t', self._screen_file)
 305
 306     def __enter__(self):
 307         self.save_console_title()
 308         return self
 309
 310     def __exit__(self, *args):
 311         self.restore_console_title()
 312
 313         if self.params.get('cookiefile') is not None:
 314             self.cookiejar.save()
 315
 316     def trouble(self, message=None, tb=None):
 317         """Determine action to take when a download problem appears.
 318
 319         Depending on if the downloader has been configured to ignore
 320         download errors or not, this method may throw an exception or
 321         not when errors are found, after printing the message.
 322
 323         tb, if given, is additional traceback information.
 324         """
 325         if message is not None:
 326             self.to_stderr(message)
 327         if self.params.get('verbose'):
 328             if tb is None:
 329                 if sys.exc_info()[0]:  # if .trouble has been called from an except block
 330                     tb = u''
 331                     if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
 332                         tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
 333                     tb += compat_str(traceback.format_exc())
 334                 else:
 335                     tb_data = traceback.format_list(traceback.extract_stack())
 336                     tb = u''.join(tb_data)
 337             self.to_stderr(tb)
 338         if not self.params.get('ignoreerrors', False):
 339             if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
 340                 exc_info = sys.exc_info()[1].exc_info
 341             else:
 342                 exc_info = sys.exc_info()
 343             raise DownloadError(message, exc_info)
 344         self._download_retcode = 1
 345
 346     def report_warning(self, message):
 347         '''
 348         Print the message to stderr, it will be prefixed with 'WARNING:'
 349         If stderr is a tty file the 'WARNING:' will be colored
 350         '''
 351         if self._err_file.isatty() and os.name != 'nt':
 352             _msg_header = u'\033[0;33mWARNING:\033[0m'
 353         else:
 354             _msg_header = u'WARNING:'
 355         warning_message = u'%s %s' % (_msg_header, message)
 356         self.to_stderr(warning_message)
 357
 358     def report_error(self, message, tb=None):
 359         '''
 360         Do the same as trouble, but prefixes the message with 'ERROR:', colored
 361         in red if stderr is a tty file.
 362         '''
 363         if self._err_file.isatty() and os.name != 'nt':
 364             _msg_header = u'\033[0;31mERROR:\033[0m'
 365         else:
 366             _msg_header = u'ERROR:'
 367         error_message = u'%s %s' % (_msg_header, message)
 368         self.trouble(error_message, tb)
 369
 370     def report_file_already_downloaded(self, file_name):
 371         """Report file has already been fully downloaded."""
 372         try:
 373             self.to_screen(u'[download] %s has already been downloaded' % file_name)
 374         except UnicodeEncodeError:
 375             self.to_screen(u'[download] The file has already been downloaded')
 376
 377     def increment_downloads(self):
 378         """Increment the ordinal that assigns a number to each file."""
 379         self._num_downloads += 1
 380
 381     def prepare_filename(self, info_dict):
 382         """Generate the output filename."""
 383         try:
 384             template_dict = dict(info_dict)
 385
 386             template_dict['epoch'] = int(time.time())
 387             autonumber_size = self.params.get('autonumber_size')
 388             if autonumber_size is None:
 389                 autonumber_size = 5
 390             autonumber_templ = u'%0' + str(autonumber_size) + u'd'
 391             template_dict['autonumber'] = autonumber_templ % self._num_downloads
 392             if template_dict.get('playlist_index') is not None:
 393                 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
 394
 395             sanitize = lambda k, v: sanitize_filename(
 396                 compat_str(v),
 397                 restricted=self.params.get('restrictfilenames'),
 398                 is_id=(k == u'id'))
 399             template_dict = dict((k, sanitize(k, v))
 400                                  for k, v in template_dict.items()
 401                                  if v is not None)
 402             template_dict = collections.defaultdict(lambda: u'NA', template_dict)
 403
 404             tmpl = os.path.expanduser(self.params['outtmpl'])
 405             filename = tmpl % template_dict
 406             return filename
 407         except ValueError as err:
 408             self.report_error(u'Error in output template: ' + str(err) + u' (encoding: ' + repr(preferredencoding()) + ')')
 409             return None
 410
 411     def _match_entry(self, info_dict):
 412         """ Returns None iff the file should be downloaded """
 413
 414         video_title = info_dict.get('title', info_dict.get('id', u'video'))
 415         if 'title' in info_dict:
 416             # This can happen when we're just evaluating the playlist
 417             title = info_dict['title']
 418             matchtitle = self.params.get('matchtitle', False)
 419             if matchtitle:
 420                 if not re.search(matchtitle, title, re.IGNORECASE):
 421                     return u'"' + title + '" title did not match pattern "' + matchtitle + '"'
 422             rejecttitle = self.params.get('rejecttitle', False)
 423             if rejecttitle:
 424                 if re.search(rejecttitle, title, re.IGNORECASE):
 425                     return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
 426         date = info_dict.get('upload_date', None)
 427         if date is not None:
 428             dateRange = self.params.get('daterange', DateRange())
 429             if date not in dateRange:
 430                 return u'%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
 431         view_count = info_dict.get('view_count', None)
 432         if view_count is not None:
 433             min_views = self.params.get('min_views')
 434             if min_views is not None and view_count < min_views:
 435                 return u'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
 436             max_views = self.params.get('max_views')
 437             if max_views is not None and view_count > max_views:
 438                 return u'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
 439         age_limit = self.params.get('age_limit')
 440         if age_limit is not None:
 441             if age_limit < info_dict.get('age_limit', 0):
 442                 return u'Skipping "' + title + '" because it is age restricted'
 443         if self.in_download_archive(info_dict):
 444             return u'%s has already been recorded in archive' % video_title
 445         return None
 446
 447     @staticmethod
 448     def add_extra_info(info_dict, extra_info):
 449         '''Set the keys from extra_info in info dict if they are missing'''
 450         for key, value in extra_info.items():
 451             info_dict.setdefault(key, value)
 452
 453     def extract_info(self, url, download=True, ie_key=None, extra_info={},
 454                      process=True):
 455         '''
 456         Returns a list with a dictionary for each video we find.
 457         If 'download', also downloads the videos.
 458         extra_info is a dict containing the extra values to add to each result
 459          '''
 460
 461         if ie_key:
 462             ies = [self.get_info_extractor(ie_key)]
 463         else:
 464             ies = self._ies
 465
 466         for ie in ies:
 467             if not ie.suitable(url):
 468                 continue
 469
 470             if not ie.working():
 471                 self.report_warning(u'The program functionality for this site has been marked as broken, '
 472                                     u'and will probably not work.')
 473
 474             try:
 475                 ie_result = ie.extract(url)
 476                 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
 477                     break
 478                 if isinstance(ie_result, list):
 479                     # Backwards compatibility: old IE result format
 480                     ie_result = {
 481                         '_type': 'compat_list',
 482                         'entries': ie_result,
 483                     }
 484                 self.add_extra_info(ie_result,
 485                     {
 486                         'extractor': ie.IE_NAME,
 487                         'webpage_url': url,
 488                         'webpage_url_basename': url_basename(url),
 489                         'extractor_key': ie.ie_key(),
 490                     })
 491                 if process:
 492                     return self.process_ie_result(ie_result, download, extra_info)
 493                 else:
 494                     return ie_result
 495             except ExtractorError as de: # An error we somewhat expected
 496                 self.report_error(compat_str(de), de.format_traceback())
 497                 break
 498             except Exception as e:
 499                 if self.params.get('ignoreerrors', False):
 500                     self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
 501                     break
 502                 else:
 503                     raise
 504         else:
 505             self.report_error(u'no suitable InfoExtractor: %s' % url)
 506
 507     def process_ie_result(self, ie_result, download=True, extra_info={}):
 508         """
 509         Take the result of the ie(may be modified) and resolve all unresolved
 510         references (URLs, playlist items).
 511
 512         It will also download the videos if 'download'.
 513         Returns the resolved ie_result.
 514         """
 515
 516         result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system
 517         if result_type == 'video':
 518             self.add_extra_info(ie_result, extra_info)
 519             return self.process_video_result(ie_result, download=download)
 520         elif result_type == 'url':
 521             # We have to add extra_info to the results because it may be
 522             # contained in a playlist
 523             return self.extract_info(ie_result['url'],
 524                                      download,
 525                                      ie_key=ie_result.get('ie_key'),
 526                                      extra_info=extra_info)
 527         elif result_type == 'url_transparent':
 528             # Use the information from the embedding page
 529             info = self.extract_info(
 530                 ie_result['url'], ie_key=ie_result.get('ie_key'),
 531                 extra_info=extra_info, download=False, process=False)
 532
 533             def make_result(embedded_info):
 534                 new_result = ie_result.copy()
 535                 for f in ('_type', 'url', 'ext', 'player_url', 'formats',
 536                           'entries', 'urlhandle', 'ie_key', 'duration',
 537                           'subtitles', 'annotations', 'format',
 538                           'thumbnail', 'thumbnails'):
 539                     if f in new_result:
 540                         del new_result[f]
 541                     if f in embedded_info:
 542                         new_result[f] = embedded_info[f]
 543                 return new_result
 544             new_result = make_result(info)
 545
 546             assert new_result.get('_type') != 'url_transparent'
 547             if new_result.get('_type') == 'compat_list':
 548                 new_result['entries'] = [
 549                     make_result(e) for e in new_result['entries']]
 550
 551             return self.process_ie_result(
 552                 new_result, download=download, extra_info=extra_info)
 553         elif result_type == 'playlist':
 554             # We process each entry in the playlist
 555             playlist = ie_result.get('title', None) or ie_result.get('id', None)
 556             self.to_screen(u'[download] Downloading playlist: %s' % playlist)
 557
 558             playlist_results = []
 559
 560             n_all_entries = len(ie_result['entries'])
 561             playliststart = self.params.get('playliststart', 1) - 1
 562             playlistend = self.params.get('playlistend', None)
 563             # For backwards compatibility, interpret -1 as whole list
 564             if playlistend == -1:
 565                 playlistend = None
 566
 567             entries = ie_result['entries'][playliststart:playlistend]
 568             n_entries = len(entries)
 569
 570             self.to_screen(
 571                 u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
 572                 (ie_result['extractor'], playlist, n_all_entries, n_entries))
 573
 574             for i, entry in enumerate(entries, 1):
 575                 self.to_screen(u'[download] Downloading video #%s of %s' % (i, n_entries))
 576                 extra = {
 577                     'playlist': playlist,
 578                     'playlist_index': i + playliststart,
 579                     'extractor': ie_result['extractor'],
 580                     'webpage_url': ie_result['webpage_url'],
 581                     'webpage_url_basename': url_basename(ie_result['webpage_url']),
 582                     'extractor_key': ie_result['extractor_key'],
 583                 }
 584
 585                 reason = self._match_entry(entry)
 586                 if reason is not None:
 587                     self.to_screen(u'[download] ' + reason)
 588                     continue
 589
 590                 entry_result = self.process_ie_result(entry,
 591                                                       download=download,
 592                                                       extra_info=extra)
 593                 playlist_results.append(entry_result)
 594             ie_result['entries'] = playlist_results
 595             return ie_result
 596         elif result_type == 'compat_list':
 597             def _fixup(r):
 598                 self.add_extra_info(r,
 599                     {
 600                         'extractor': ie_result['extractor'],
 601                         'webpage_url': ie_result['webpage_url'],
 602                         'webpage_url_basename': url_basename(ie_result['webpage_url']),
 603                         'extractor_key': ie_result['extractor_key'],
 604                     })
 605                 return r
 606             ie_result['entries'] = [
 607                 self.process_ie_result(_fixup(r), download, extra_info)
 608                 for r in ie_result['entries']
 609             ]
 610             return ie_result
 611         else:
 612             raise Exception('Invalid result type: %s' % result_type)
 613
 614     def select_format(self, format_spec, available_formats):
 615         if format_spec == 'best' or format_spec is None:
 616             return available_formats[-1]
 617         elif format_spec == 'worst':
 618             return available_formats[0]
 619         else:
 620             extensions = [u'mp4', u'flv', u'webm', u'3gp']
 621             if format_spec in extensions:
 622                 filter_f = lambda f: f['ext'] == format_spec
 623             else:
 624                 filter_f = lambda f: f['format_id'] == format_spec
 625             matches = list(filter(filter_f, available_formats))
 626             if matches:
 627                 return matches[-1]
 628         return None
 629
 630     def process_video_result(self, info_dict, download=True):
 631         assert info_dict.get('_type', 'video') == 'video'
 632
 633         if 'playlist' not in info_dict:
 634             # It isn't part of a playlist
 635             info_dict['playlist'] = None
 636             info_dict['playlist_index'] = None
 637
 638         # This extractors handle format selection themselves
 639         if info_dict['extractor'] in [u'youtube', u'Youku']:
 640             if download:
 641                 self.process_info(info_dict)
 642             return info_dict
 643
 644         # We now pick which formats have to be downloaded
 645         if info_dict.get('formats') is None:
 646             # There's only one format available
 647             formats = [info_dict]
 648         else:
 649             formats = info_dict['formats']
 650
 651         # We check that all the formats have the format and format_id fields
 652         for (i, format) in enumerate(formats):
 653             if format.get('format_id') is None:
 654                 format['format_id'] = compat_str(i)
 655             if format.get('format') is None:
 656                 format['format'] = u'{id} - {res}{note}'.format(
 657                     id=format['format_id'],
 658                     res=self.format_resolution(format),
 659                     note=u' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
 660                 )
 661             # Automatically determine file extension if missing
 662             if 'ext' not in format:
 663                 format['ext'] = determine_ext(format['url'])
 664
 665         if self.params.get('listformats', None):
 666             self.list_formats(info_dict)
 667             return
 668
 669         format_limit = self.params.get('format_limit', None)
 670         if format_limit:
 671             formats = list(takewhile_inclusive(
 672                 lambda f: f['format_id'] != format_limit, formats
 673             ))
 674         if self.params.get('prefer_free_formats'):
 675             def _free_formats_key(f):
 676                 try:
 677                     ext_ord = [u'flv', u'mp4', u'webm'].index(f['ext'])
 678                 except ValueError:
 679                     ext_ord = -1
 680                 # We only compare the extension if they have the same height and width
 681                 return (f.get('height'), f.get('width'), ext_ord)
 682             formats = sorted(formats, key=_free_formats_key)
 683
 684         req_format = self.params.get('format', 'best')
 685         if req_format is None:
 686             req_format = 'best'
 687         formats_to_download = []
 688         # The -1 is for supporting YoutubeIE
 689         if req_format in ('-1', 'all'):
 690             formats_to_download = formats
 691         else:
 692             # We can accept formats requestd in the format: 34/5/best, we pick
 693             # the first that is available, starting from left
 694             req_formats = req_format.split('/')
 695             for rf in req_formats:
 696                 selected_format = self.select_format(rf, formats)
 697                 if selected_format is not None:
 698                     formats_to_download = [selected_format]
 699                     break
 700         if not formats_to_download:
 701             raise ExtractorError(u'requested format not available',
 702                                  expected=True)
 703
 704         if download:
 705             if len(formats_to_download) > 1:
 706                 self.to_screen(u'[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
 707             for format in formats_to_download:
 708                 new_info = dict(info_dict)
 709                 new_info.update(format)
 710                 self.process_info(new_info)
 711         # We update the info dict with the best quality format (backwards compatibility)
 712         info_dict.update(formats_to_download[-1])
 713         return info_dict
 714
 715     def process_info(self, info_dict):
 716         """Process a single resolved IE result."""
 717
 718         assert info_dict.get('_type', 'video') == 'video'
 719         #We increment the download the download count here to match the previous behaviour.
 720         self.increment_downloads()
 721
 722         info_dict['fulltitle'] = info_dict['title']
 723         if len(info_dict['title']) > 200:
 724             info_dict['title'] = info_dict['title'][:197] + u'...'
 725
 726         # Keep for backwards compatibility
 727         info_dict['stitle'] = info_dict['title']
 728
 729         if not 'format' in info_dict:
 730             info_dict['format'] = info_dict['ext']
 731
 732         reason = self._match_entry(info_dict)
 733         if reason is not None:
 734             self.to_screen(u'[download] ' + reason)
 735             return
 736
 737         max_downloads = self.params.get('max_downloads')
 738         if max_downloads is not None:
 739             if self._num_downloads > int(max_downloads):
 740                 raise MaxDownloadsReached()
 741
 742         filename = self.prepare_filename(info_dict)
 743
 744         # Forced printings
 745         if self.params.get('forcetitle', False):
 746             self.to_stdout(info_dict['fulltitle'])
 747         if self.params.get('forceid', False):
 748             self.to_stdout(info_dict['id'])
 749         if self.params.get('forceurl', False):
 750             # For RTMP URLs, also include the playpath
 751             self.to_stdout(info_dict['url'] + info_dict.get('play_path', u''))
 752         if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
 753             self.to_stdout(info_dict['thumbnail'])
 754         if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
 755             self.to_stdout(info_dict['description'])
 756         if self.params.get('forcefilename', False) and filename is not None:
 757             self.to_stdout(filename)
 758         if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
 759             self.to_stdout(formatSeconds(info_dict['duration']))
 760         if self.params.get('forceformat', False):
 761             self.to_stdout(info_dict['format'])
 762         if self.params.get('forcejson', False):
 763             info_dict['_filename'] = filename
 764             self.to_stdout(json.dumps(info_dict))
 765
 766         # Do nothing else if in simulate mode
 767         if self.params.get('simulate', False):
 768             return
 769
 770         if filename is None:
 771             return
 772
 773         try:
 774             dn = os.path.dirname(encodeFilename(filename))
 775             if dn != '' and not os.path.exists(dn):
 776                 os.makedirs(dn)
 777         except (OSError, IOError) as err:
 778             self.report_error(u'unable to create directory ' + compat_str(err))
 779             return
 780
 781         if self.params.get('writedescription', False):
 782             descfn = filename + u'.description'
 783             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
 784                 self.to_screen(u'[info] Video description is already present')
 785             else:
 786                 try:
 787                     self.to_screen(u'[info] Writing video description to: ' + descfn)
 788                     with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
 789                         descfile.write(info_dict['description'])
 790                 except (KeyError, TypeError):
 791                     self.report_warning(u'There\'s no description to write.')
 792                 except (OSError, IOError):
 793                     self.report_error(u'Cannot write description file ' + descfn)
 794                     return
 795
 796         if self.params.get('writeannotations', False):
 797             annofn = filename + u'.annotations.xml'
 798             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
 799                 self.to_screen(u'[info] Video annotations are already present')
 800             else:
 801                 try:
 802                     self.to_screen(u'[info] Writing video annotations to: ' + annofn)
 803                     with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
 804                         annofile.write(info_dict['annotations'])
 805                 except (KeyError, TypeError):
 806                     self.report_warning(u'There are no annotations to write.')
 807                 except (OSError, IOError):
 808                     self.report_error(u'Cannot write annotations file: ' + annofn)
 809                     return
 810
 811         subtitles_are_requested = any([self.params.get('writesubtitles', False),
 812                                        self.params.get('writeautomaticsub')])
 813
 814         if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
 815             # subtitles download errors are already managed as troubles in relevant IE
 816             # that way it will silently go on when used with unsupporting IE
 817             subtitles = info_dict['subtitles']
 818             sub_format = self.params.get('subtitlesformat', 'srt')
 819             for sub_lang in subtitles.keys():
 820                 sub = subtitles[sub_lang]
 821                 if sub is None:
 822                     continue
 823                 try:
 824                     sub_filename = subtitles_filename(filename, sub_lang, sub_format)
 825                     if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
 826                         self.to_screen(u'[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
 827                     else:
 828                         self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
 829                         with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
 830                                 subfile.write(sub)
 831                 except (OSError, IOError):
 832                     self.report_error(u'Cannot write subtitles file ' + descfn)
 833                     return
 834
 835         if self.params.get('writeinfojson', False):
 836             infofn = os.path.splitext(filename)[0] + u'.info.json'
 837             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
 838                 self.to_screen(u'[info] Video description metadata is already present')
 839             else:
 840                 self.to_screen(u'[info] Writing video description metadata as JSON to: ' + infofn)
 841                 try:
 842                     json_info_dict = dict((k, v) for k, v in info_dict.items() if not k in ['urlhandle'])
 843                     write_json_file(json_info_dict, encodeFilename(infofn))
 844                 except (OSError, IOError):
 845                     self.report_error(u'Cannot write metadata to JSON file ' + infofn)
 846                     return
 847
 848         if self.params.get('writethumbnail', False):
 849             if info_dict.get('thumbnail') is not None:
 850                 thumb_format = determine_ext(info_dict['thumbnail'], u'jpg')
 851                 thumb_filename = os.path.splitext(filename)[0] + u'.' + thumb_format
 852                 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
 853                     self.to_screen(u'[%s] %s: Thumbnail is already present' %
 854                                    (info_dict['extractor'], info_dict['id']))
 855                 else:
 856                     self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
 857                                    (info_dict['extractor'], info_dict['id']))
 858                     try:
 859                         uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
 860                         with open(thumb_filename, 'wb') as thumbf:
 861                             shutil.copyfileobj(uf, thumbf)
 862                         self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
 863                             (info_dict['extractor'], info_dict['id'], thumb_filename))
 864                     except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 865                         self.report_warning(u'Unable to download thumbnail "%s": %s' %
 866                             (info_dict['thumbnail'], compat_str(err)))
 867
 868         if not self.params.get('skip_download', False):
 869             if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
 870                 success = True
 871             else:
 872                 try:
 873                     success = self.fd._do_download(filename, info_dict)
 874                 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 875                     self.report_error(u'unable to download video data: %s' % str(err))
 876                     return
 877                 except (OSError, IOError) as err:
 878                     raise UnavailableVideoError(err)
 879                 except (ContentTooShortError, ) as err:
 880                     self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
 881                     return
 882
 883             if success:
 884                 try:
 885                     self.post_process(filename, info_dict)
 886                 except (PostProcessingError) as err:
 887                     self.report_error(u'postprocessing: %s' % str(err))
 888                     return
 889
 890         self.record_download_archive(info_dict)
 891
 892     def download(self, url_list):
 893         """Download a given list of URLs."""
 894         if (len(url_list) > 1 and
 895                 '%' not in self.params['outtmpl']
 896                 and self.params.get('max_downloads') != 1):
 897             raise SameFileError(self.params['outtmpl'])
 898
 899         for url in url_list:
 900             try:
 901                 #It also downloads the videos
 902                 self.extract_info(url)
 903             except UnavailableVideoError:
 904                 self.report_error(u'unable to download video')
 905             except MaxDownloadsReached:
 906                 self.to_screen(u'[info] Maximum number of downloaded files reached.')
 907                 raise
 908
 909         return self._download_retcode
 910
 911     def download_with_info_file(self, info_filename):
 912         with io.open(info_filename, 'r', encoding='utf-8') as f:
 913             info = json.load(f)
 914         try:
 915             self.process_ie_result(info, download=True)
 916         except DownloadError:
 917             webpage_url = info.get('webpage_url')
 918             if webpage_url is not None:
 919                 self.report_warning(u'The info failed to download, trying with "%s"' % webpage_url)
 920                 return self.download([webpage_url])
 921             else:
 922                 raise
 923         return self._download_retcode
 924
 925     def post_process(self, filename, ie_info):
 926         """Run all the postprocessors on the given file."""
 927         info = dict(ie_info)
 928         info['filepath'] = filename
 929         keep_video = None
 930         for pp in self._pps:
 931             try:
 932                 keep_video_wish, new_info = pp.run(info)
 933                 if keep_video_wish is not None:
 934                     if keep_video_wish:
 935                         keep_video = keep_video_wish
 936                     elif keep_video is None:
 937                         # No clear decision yet, let IE decide
 938                         keep_video = keep_video_wish
 939             except PostProcessingError as e:
 940                 self.report_error(e.msg)
 941         if keep_video is False and not self.params.get('keepvideo', False):
 942             try:
 943                 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
 944                 os.remove(encodeFilename(filename))
 945             except (IOError, OSError):
 946                 self.report_warning(u'Unable to remove downloaded video file')
 947
 948     def _make_archive_id(self, info_dict):
 949         # Future-proof against any change in case
 950         # and backwards compatibility with prior versions
 951         extractor = info_dict.get('extractor_key')
 952         if extractor is None:
 953             if 'id' in info_dict:
 954                 extractor = info_dict.get('ie_key')  # key in a playlist
 955         if extractor is None:
 956             return None  # Incomplete video information
 957         return extractor.lower() + u' ' + info_dict['id']
 958
 959     def in_download_archive(self, info_dict):
 960         fn = self.params.get('download_archive')
 961         if fn is None:
 962             return False
 963
 964         vid_id = self._make_archive_id(info_dict)
 965         if vid_id is None:
 966             return False  # Incomplete video information
 967
 968         try:
 969             with locked_file(fn, 'r', encoding='utf-8') as archive_file:
 970                 for line in archive_file:
 971                     if line.strip() == vid_id:
 972                         return True
 973         except IOError as ioe:
 974             if ioe.errno != errno.ENOENT:
 975                 raise
 976         return False
 977
 978     def record_download_archive(self, info_dict):
 979         fn = self.params.get('download_archive')
 980         if fn is None:
 981             return
 982         vid_id = self._make_archive_id(info_dict)
 983         assert vid_id
 984         with locked_file(fn, 'a', encoding='utf-8') as archive_file:
 985             archive_file.write(vid_id + u'\n')
 986
 987     @staticmethod
 988     def format_resolution(format, default='unknown'):
 989         if format.get('vcodec') == 'none':
 990             return 'audio only'
 991         if format.get('_resolution') is not None:
 992             return format['_resolution']
 993         if format.get('height') is not None:
 994             if format.get('width') is not None:
 995                 res = u'%sx%s' % (format['width'], format['height'])
 996             else:
 997                 res = u'%sp' % format['height']
 998         else:
 999             res = default
1000         return res
1001
1002     def list_formats(self, info_dict):
1003         def format_note(fdict):
1004             res = u''
1005             if fdict.get('format_note') is not None:
1006                 res += fdict['format_note'] + u' '
1007             if (fdict.get('vcodec') is not None and
1008                     fdict.get('vcodec') != 'none'):
1009                 res += u'%-5s' % fdict['vcodec']
1010             elif fdict.get('vbr') is not None:
1011                 res += u'video'
1012             if fdict.get('vbr') is not None:
1013                 res += u'@%4dk' % fdict['vbr']
1014             if fdict.get('acodec') is not None:
1015                 if res:
1016                     res += u', '
1017                 res += u'%-5s' % fdict['acodec']
1018             elif fdict.get('abr') is not None:
1019                 if res:
1020                     res += u', '
1021                 res += 'audio'
1022             if fdict.get('abr') is not None:
1023                 res += u'@%3dk' % fdict['abr']
1024             if fdict.get('filesize') is not None:
1025                 if res:
1026                     res += u', '
1027                 res += format_bytes(fdict['filesize'])
1028             return res
1029
1030         def line(format, idlen=20):
1031             return ((u'%-' + compat_str(idlen + 1) + u's%-10s%-12s%s') % (
1032                 format['format_id'],
1033                 format['ext'],
1034                 self.format_resolution(format),
1035                 format_note(format),
1036             ))
1037
1038         formats = info_dict.get('formats', [info_dict])
1039         idlen = max(len(u'format code'),
1040                     max(len(f['format_id']) for f in formats))
1041         formats_s = [line(f, idlen) for f in formats]
1042         if len(formats) > 1:
1043             formats_s[0] += (' ' if format_note(formats[0]) else '') + '(worst)'
1044             formats_s[-1] += (' ' if format_note(formats[-1]) else '') + '(best)'
1045
1046         header_line = line({
1047             'format_id': u'format code', 'ext': u'extension',
1048             '_resolution': u'resolution', 'format_note': u'note'}, idlen=idlen)
1049         self.to_screen(u'[info] Available formats for %s:\n%s\n%s' %
1050                        (info_dict['id'], header_line, u"\n".join(formats_s)))
1051
1052     def urlopen(self, req):
1053         """ Start an HTTP download """
1054         return self._opener.open(req)
1055
1056     def print_debug_header(self):
1057         if not self.params.get('verbose'):
1058             return
1059         write_string(u'[debug] youtube-dl version ' + __version__ + u'\n')
1060         try:
1061             sp = subprocess.Popen(
1062                 ['git', 'rev-parse', '--short', 'HEAD'],
1063                 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1064                 cwd=os.path.dirname(os.path.abspath(__file__)))
1065             out, err = sp.communicate()
1066             out = out.decode().strip()
1067             if re.match('[0-9a-f]+', out):
1068                 write_string(u'[debug] Git HEAD: ' + out + u'\n')
1069         except:
1070             try:
1071                 sys.exc_clear()
1072             except:
1073                 pass
1074         write_string(u'[debug] Python version %s - %s' %
1075                      (platform.python_version(), platform_name()) + u'\n')
1076
1077         proxy_map = {}
1078         for handler in self._opener.handlers:
1079             if hasattr(handler, 'proxies'):
1080                 proxy_map.update(handler.proxies)
1081         write_string(u'[debug] Proxy map: ' + compat_str(proxy_map) + u'\n')
1082
1083     def _setup_opener(self):
1084         timeout_val = self.params.get('socket_timeout')
1085         timeout = 600 if timeout_val is None else float(timeout_val)
1086
1087         opts_cookiefile = self.params.get('cookiefile')
1088         opts_proxy = self.params.get('proxy')
1089
1090         if opts_cookiefile is None:
1091             self.cookiejar = compat_cookiejar.CookieJar()
1092         else:
1093             self.cookiejar = compat_cookiejar.MozillaCookieJar(
1094                 opts_cookiefile)
1095             if os.access(opts_cookiefile, os.R_OK):
1096                 self.cookiejar.load()
1097
1098         cookie_processor = compat_urllib_request.HTTPCookieProcessor(
1099             self.cookiejar)
1100         if opts_proxy is not None:
1101             if opts_proxy == '':
1102                 proxies = {}
1103             else:
1104                 proxies = {'http': opts_proxy, 'https': opts_proxy}
1105         else:
1106             proxies = compat_urllib_request.getproxies()
1107             # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1108             if 'http' in proxies and 'https' not in proxies:
1109                 proxies['https'] = proxies['http']
1110         proxy_handler = compat_urllib_request.ProxyHandler(proxies)
1111         https_handler = make_HTTPS_handler(
1112             self.params.get('nocheckcertificate', False))
1113         opener = compat_urllib_request.build_opener(
1114             https_handler, proxy_handler, cookie_processor, YoutubeDLHandler())
1115         # Delete the default user-agent header, which would otherwise apply in
1116         # cases where our custom HTTP handler doesn't come into play
1117         # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
1118         opener.addheaders = []
1119         self._opener = opener
1120
1121         # TODO remove this global modification
1122         compat_urllib_request.install_opener(opener)
1123         socket.setdefaulttimeout(timeout)