]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/YoutubeDL.py
Prepare to release.
[youtubedl] / 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 copy
9 import datetime
10 import errno
11 import fileinput
12 import io
13 import itertools
14 import json
15 import locale
16 import operator
17 import os
18 import platform
19 import re
20 import shutil
21 import subprocess
22 import socket
23 import sys
24 import time
25 import tokenize
26 import traceback
27 import random
28
29 from string import ascii_letters
30
31 from .compat import (
32 compat_basestring,
33 compat_cookiejar,
34 compat_get_terminal_size,
35 compat_http_client,
36 compat_kwargs,
37 compat_numeric_types,
38 compat_os_name,
39 compat_str,
40 compat_tokenize_tokenize,
41 compat_urllib_error,
42 compat_urllib_request,
43 compat_urllib_request_DataHandler,
44 )
45 from .utils import (
46 age_restricted,
47 args_to_str,
48 ContentTooShortError,
49 date_from_str,
50 DateRange,
51 DEFAULT_OUTTMPL,
52 determine_ext,
53 determine_protocol,
54 DownloadError,
55 encode_compat_str,
56 encodeFilename,
57 error_to_compat_str,
58 expand_path,
59 ExtractorError,
60 format_bytes,
61 formatSeconds,
62 GeoRestrictedError,
63 int_or_none,
64 ISO3166Utils,
65 locked_file,
66 make_HTTPS_handler,
67 MaxDownloadsReached,
68 orderedSet,
69 PagedList,
70 parse_filesize,
71 PerRequestProxyHandler,
72 platform_name,
73 PostProcessingError,
74 preferredencoding,
75 prepend_extension,
76 register_socks_protocols,
77 render_table,
78 replace_extension,
79 SameFileError,
80 sanitize_filename,
81 sanitize_path,
82 sanitize_url,
83 sanitized_Request,
84 std_headers,
85 subtitles_filename,
86 UnavailableVideoError,
87 url_basename,
88 version_tuple,
89 write_json_file,
90 write_string,
91 YoutubeDLCookieProcessor,
92 YoutubeDLHandler,
93 )
94 from .cache import Cache
95 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
96 from .extractor.openload import PhantomJSwrapper
97 from .downloader import get_suitable_downloader
98 from .downloader.rtmp import rtmpdump_version
99 from .postprocessor import (
100 FFmpegFixupM3u8PP,
101 FFmpegFixupM4aPP,
102 FFmpegFixupStretchedPP,
103 FFmpegMergerPP,
104 FFmpegPostProcessor,
105 get_postprocessor,
106 )
107 from .version import __version__
108
109 if compat_os_name == 'nt':
110 import ctypes
111
112
113 class YoutubeDL(object):
114 """YoutubeDL class.
115
116 YoutubeDL objects are the ones responsible of downloading the
117 actual video file and writing it to disk if the user has requested
118 it, among some other tasks. In most cases there should be one per
119 program. As, given a video URL, the downloader doesn't know how to
120 extract all the needed information, task that InfoExtractors do, it
121 has to pass the URL to one of them.
122
123 For this, YoutubeDL objects have a method that allows
124 InfoExtractors to be registered in a given order. When it is passed
125 a URL, the YoutubeDL object handles it to the first InfoExtractor it
126 finds that reports being able to handle it. The InfoExtractor extracts
127 all the information about the video or videos the URL refers to, and
128 YoutubeDL process the extracted information, possibly using a File
129 Downloader to download the video.
130
131 YoutubeDL objects accept a lot of parameters. In order not to saturate
132 the object constructor with arguments, it receives a dictionary of
133 options instead. These options are available through the params
134 attribute for the InfoExtractors to use. The YoutubeDL also
135 registers itself as the downloader in charge for the InfoExtractors
136 that are added to it, so this is a "mutual registration".
137
138 Available options:
139
140 username: Username for authentication purposes.
141 password: Password for authentication purposes.
142 videopassword: Password for accessing a video.
143 ap_mso: Adobe Pass multiple-system operator identifier.
144 ap_username: Multiple-system operator account username.
145 ap_password: Multiple-system operator account password.
146 usenetrc: Use netrc for authentication instead.
147 verbose: Print additional info to stdout.
148 quiet: Do not print messages to stdout.
149 no_warnings: Do not print out anything for warnings.
150 forceurl: Force printing final URL.
151 forcetitle: Force printing title.
152 forceid: Force printing ID.
153 forcethumbnail: Force printing thumbnail URL.
154 forcedescription: Force printing description.
155 forcefilename: Force printing final filename.
156 forceduration: Force printing duration.
157 forcejson: Force printing info_dict as JSON.
158 dump_single_json: Force printing the info_dict of the whole playlist
159 (or video) as a single JSON line.
160 simulate: Do not download the video files.
161 format: Video format code. See options.py for more information.
162 outtmpl: Template for output names.
163 restrictfilenames: Do not allow "&" and spaces in file names
164 ignoreerrors: Do not stop on download errors.
165 force_generic_extractor: Force downloader to use the generic extractor
166 nooverwrites: Prevent overwriting files.
167 playliststart: Playlist item to start at.
168 playlistend: Playlist item to end at.
169 playlist_items: Specific indices of playlist to download.
170 playlistreverse: Download playlist items in reverse order.
171 playlistrandom: Download playlist items in random order.
172 matchtitle: Download only matching titles.
173 rejecttitle: Reject downloads for matching titles.
174 logger: Log messages to a logging.Logger instance.
175 logtostderr: Log messages to stderr instead of stdout.
176 writedescription: Write the video description to a .description file
177 writeinfojson: Write the video description to a .info.json file
178 writeannotations: Write the video annotations to a .annotations.xml file
179 writethumbnail: Write the thumbnail image to a file
180 write_all_thumbnails: Write all thumbnail formats to files
181 writesubtitles: Write the video subtitles to a file
182 writeautomaticsub: Write the automatically generated subtitles to a file
183 allsubtitles: Downloads all the subtitles of the video
184 (requires writesubtitles or writeautomaticsub)
185 listsubtitles: Lists all available subtitles for the video
186 subtitlesformat: The format code for subtitles
187 subtitleslangs: List of languages of the subtitles to download
188 keepvideo: Keep the video file after post-processing
189 daterange: A DateRange object, download only if the upload_date is in the range.
190 skip_download: Skip the actual download of the video file
191 cachedir: Location of the cache files in the filesystem.
192 False to disable filesystem cache.
193 noplaylist: Download single video instead of a playlist if in doubt.
194 age_limit: An integer representing the user's age in years.
195 Unsuitable videos for the given age are skipped.
196 min_views: An integer representing the minimum view count the video
197 must have in order to not be skipped.
198 Videos without view count information are always
199 downloaded. None for no limit.
200 max_views: An integer representing the maximum view count.
201 Videos that are more popular than that are not
202 downloaded.
203 Videos without view count information are always
204 downloaded. None for no limit.
205 download_archive: File name of a file where all downloads are recorded.
206 Videos already present in the file are not downloaded
207 again.
208 cookiefile: File name where cookies should be read from and dumped to.
209 nocheckcertificate:Do not verify SSL certificates
210 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
211 At the moment, this is only supported by YouTube.
212 proxy: URL of the proxy server to use
213 geo_verification_proxy: URL of the proxy to use for IP address verification
214 on geo-restricted sites.
215 socket_timeout: Time to wait for unresponsive hosts, in seconds
216 bidi_workaround: Work around buggy terminals without bidirectional text
217 support, using fridibi
218 debug_printtraffic:Print out sent and received HTTP traffic
219 include_ads: Download ads as well
220 default_search: Prepend this string if an input url is not valid.
221 'auto' for elaborate guessing
222 encoding: Use this encoding instead of the system-specified.
223 extract_flat: Do not resolve URLs, return the immediate result.
224 Pass in 'in_playlist' to only show this behavior for
225 playlist items.
226 postprocessors: A list of dictionaries, each with an entry
227 * key: The name of the postprocessor. See
228 youtube_dl/postprocessor/__init__.py for a list.
229 as well as any further keyword arguments for the
230 postprocessor.
231 progress_hooks: A list of functions that get called on download
232 progress, with a dictionary with the entries
233 * status: One of "downloading", "error", or "finished".
234 Check this first and ignore unknown values.
235
236 If status is one of "downloading", or "finished", the
237 following properties may also be present:
238 * filename: The final filename (always present)
239 * tmpfilename: The filename we're currently writing to
240 * downloaded_bytes: Bytes on disk
241 * total_bytes: Size of the whole file, None if unknown
242 * total_bytes_estimate: Guess of the eventual file size,
243 None if unavailable.
244 * elapsed: The number of seconds since download started.
245 * eta: The estimated time in seconds, None if unknown
246 * speed: The download speed in bytes/second, None if
247 unknown
248 * fragment_index: The counter of the currently
249 downloaded video fragment.
250 * fragment_count: The number of fragments (= individual
251 files that will be merged)
252
253 Progress hooks are guaranteed to be called at least once
254 (with status "finished") if the download is successful.
255 merge_output_format: Extension to use when merging formats.
256 fixup: Automatically correct known faults of the file.
257 One of:
258 - "never": do nothing
259 - "warn": only emit a warning
260 - "detect_or_warn": check whether we can do anything
261 about it, warn otherwise (default)
262 source_address: Client-side IP address to bind to.
263 call_home: Boolean, true iff we are allowed to contact the
264 youtube-dl servers for debugging.
265 sleep_interval: Number of seconds to sleep before each download when
266 used alone or a lower bound of a range for randomized
267 sleep before each download (minimum possible number
268 of seconds to sleep) when used along with
269 max_sleep_interval.
270 max_sleep_interval:Upper bound of a range for randomized sleep before each
271 download (maximum possible number of seconds to sleep).
272 Must only be used along with sleep_interval.
273 Actual sleep time will be a random float from range
274 [sleep_interval; max_sleep_interval].
275 listformats: Print an overview of available video formats and exit.
276 list_thumbnails: Print a table of all thumbnails and exit.
277 match_filter: A function that gets called with the info_dict of
278 every video.
279 If it returns a message, the video is ignored.
280 If it returns None, the video is downloaded.
281 match_filter_func in utils.py is one example for this.
282 no_color: Do not emit color codes in output.
283 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
284 HTTP header
285 geo_bypass_country:
286 Two-letter ISO 3166-2 country code that will be used for
287 explicit geographic restriction bypassing via faking
288 X-Forwarded-For HTTP header
289 geo_bypass_ip_block:
290 IP range in CIDR notation that will be used similarly to
291 geo_bypass_country
292
293 The following options determine which downloader is picked:
294 external_downloader: Executable of the external downloader to call.
295 None or unset for standard (built-in) downloader.
296 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
297 if True, otherwise use ffmpeg/avconv if False, otherwise
298 use downloader suggested by extractor if None.
299
300 The following parameters are not used by YoutubeDL itself, they are used by
301 the downloader (see youtube_dl/downloader/common.py):
302 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
303 noresizebuffer, retries, continuedl, noprogress, consoletitle,
304 xattr_set_filesize, external_downloader_args, hls_use_mpegts,
305 http_chunk_size.
306
307 The following options are used by the post processors:
308 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
309 otherwise prefer avconv.
310 postprocessor_args: A list of additional command-line arguments for the
311 postprocessor.
312
313 The following options are used by the Youtube extractor:
314 youtube_include_dash_manifest: If True (default), DASH manifests and related
315 data will be downloaded and processed by extractor.
316 You can reduce network I/O by disabling it if you don't
317 care about DASH.
318 """
319
320 _NUMERIC_FIELDS = set((
321 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
322 'timestamp', 'upload_year', 'upload_month', 'upload_day',
323 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
324 'average_rating', 'comment_count', 'age_limit',
325 'start_time', 'end_time',
326 'chapter_number', 'season_number', 'episode_number',
327 'track_number', 'disc_number', 'release_year',
328 'playlist_index',
329 ))
330
331 params = None
332 _ies = []
333 _pps = []
334 _download_retcode = None
335 _num_downloads = None
336 _screen_file = None
337
338 def __init__(self, params=None, auto_init=True):
339 """Create a FileDownloader object with the given options."""
340 if params is None:
341 params = {}
342 self._ies = []
343 self._ies_instances = {}
344 self._pps = []
345 self._progress_hooks = []
346 self._download_retcode = 0
347 self._num_downloads = 0
348 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
349 self._err_file = sys.stderr
350 self.params = {
351 # Default parameters
352 'nocheckcertificate': False,
353 }
354 self.params.update(params)
355 self.cache = Cache(self)
356
357 def check_deprecated(param, option, suggestion):
358 if self.params.get(param) is not None:
359 self.report_warning(
360 '%s is deprecated. Use %s instead.' % (option, suggestion))
361 return True
362 return False
363
364 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
365 if self.params.get('geo_verification_proxy') is None:
366 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
367
368 check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
369 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
370 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
371
372 if params.get('bidi_workaround', False):
373 try:
374 import pty
375 master, slave = pty.openpty()
376 width = compat_get_terminal_size().columns
377 if width is None:
378 width_args = []
379 else:
380 width_args = ['-w', str(width)]
381 sp_kwargs = dict(
382 stdin=subprocess.PIPE,
383 stdout=slave,
384 stderr=self._err_file)
385 try:
386 self._output_process = subprocess.Popen(
387 ['bidiv'] + width_args, **sp_kwargs
388 )
389 except OSError:
390 self._output_process = subprocess.Popen(
391 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
392 self._output_channel = os.fdopen(master, 'rb')
393 except OSError as ose:
394 if ose.errno == errno.ENOENT:
395 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 $PATH.')
396 else:
397 raise
398
399 if (sys.platform != 'win32' and
400 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
401 not params.get('restrictfilenames', False)):
402 # Unicode filesystem API will throw errors (#1474, #13027)
403 self.report_warning(
404 'Assuming --restrict-filenames since file system encoding '
405 'cannot encode all characters. '
406 'Set the LC_ALL environment variable to fix this.')
407 self.params['restrictfilenames'] = True
408
409 if isinstance(params.get('outtmpl'), bytes):
410 self.report_warning(
411 'Parameter outtmpl is bytes, but should be a unicode string. '
412 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
413
414 self._setup_opener()
415
416 if auto_init:
417 self.print_debug_header()
418 self.add_default_info_extractors()
419
420 for pp_def_raw in self.params.get('postprocessors', []):
421 pp_class = get_postprocessor(pp_def_raw['key'])
422 pp_def = dict(pp_def_raw)
423 del pp_def['key']
424 pp = pp_class(self, **compat_kwargs(pp_def))
425 self.add_post_processor(pp)
426
427 for ph in self.params.get('progress_hooks', []):
428 self.add_progress_hook(ph)
429
430 register_socks_protocols()
431
432 def warn_if_short_id(self, argv):
433 # short YouTube ID starting with dash?
434 idxs = [
435 i for i, a in enumerate(argv)
436 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
437 if idxs:
438 correct_argv = (
439 ['youtube-dl'] +
440 [a for i, a in enumerate(argv) if i not in idxs] +
441 ['--'] + [argv[i] for i in idxs]
442 )
443 self.report_warning(
444 'Long argument string detected. '
445 'Use -- to separate parameters and URLs, like this:\n%s\n' %
446 args_to_str(correct_argv))
447
448 def add_info_extractor(self, ie):
449 """Add an InfoExtractor object to the end of the list."""
450 self._ies.append(ie)
451 if not isinstance(ie, type):
452 self._ies_instances[ie.ie_key()] = ie
453 ie.set_downloader(self)
454
455 def get_info_extractor(self, ie_key):
456 """
457 Get an instance of an IE with name ie_key, it will try to get one from
458 the _ies list, if there's no instance it will create a new one and add
459 it to the extractor list.
460 """
461 ie = self._ies_instances.get(ie_key)
462 if ie is None:
463 ie = get_info_extractor(ie_key)()
464 self.add_info_extractor(ie)
465 return ie
466
467 def add_default_info_extractors(self):
468 """
469 Add the InfoExtractors returned by gen_extractors to the end of the list
470 """
471 for ie in gen_extractor_classes():
472 self.add_info_extractor(ie)
473
474 def add_post_processor(self, pp):
475 """Add a PostProcessor object to the end of the chain."""
476 self._pps.append(pp)
477 pp.set_downloader(self)
478
479 def add_progress_hook(self, ph):
480 """Add the progress hook (currently only for the file downloader)"""
481 self._progress_hooks.append(ph)
482
483 def _bidi_workaround(self, message):
484 if not hasattr(self, '_output_channel'):
485 return message
486
487 assert hasattr(self, '_output_process')
488 assert isinstance(message, compat_str)
489 line_count = message.count('\n') + 1
490 self._output_process.stdin.write((message + '\n').encode('utf-8'))
491 self._output_process.stdin.flush()
492 res = ''.join(self._output_channel.readline().decode('utf-8')
493 for _ in range(line_count))
494 return res[:-len('\n')]
495
496 def to_screen(self, message, skip_eol=False):
497 """Print message to stdout if not in quiet mode."""
498 return self.to_stdout(message, skip_eol, check_quiet=True)
499
500 def _write_string(self, s, out=None):
501 write_string(s, out=out, encoding=self.params.get('encoding'))
502
503 def to_stdout(self, message, skip_eol=False, check_quiet=False):
504 """Print message to stdout if not in quiet mode."""
505 if self.params.get('logger'):
506 self.params['logger'].debug(message)
507 elif not check_quiet or not self.params.get('quiet', False):
508 message = self._bidi_workaround(message)
509 terminator = ['\n', ''][skip_eol]
510 output = message + terminator
511
512 self._write_string(output, self._screen_file)
513
514 def to_stderr(self, message):
515 """Print message to stderr."""
516 assert isinstance(message, compat_str)
517 if self.params.get('logger'):
518 self.params['logger'].error(message)
519 else:
520 message = self._bidi_workaround(message)
521 output = message + '\n'
522 self._write_string(output, self._err_file)
523
524 def to_console_title(self, message):
525 if not self.params.get('consoletitle', False):
526 return
527 if compat_os_name == 'nt':
528 if ctypes.windll.kernel32.GetConsoleWindow():
529 # c_wchar_p() might not be necessary if `message` is
530 # already of type unicode()
531 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
532 elif 'TERM' in os.environ:
533 self._write_string('\033]0;%s\007' % message, self._screen_file)
534
535 def save_console_title(self):
536 if not self.params.get('consoletitle', False):
537 return
538 if self.params.get('simulate', False):
539 return
540 if compat_os_name != 'nt' and 'TERM' in os.environ:
541 # Save the title on stack
542 self._write_string('\033[22;0t', self._screen_file)
543
544 def restore_console_title(self):
545 if not self.params.get('consoletitle', False):
546 return
547 if self.params.get('simulate', False):
548 return
549 if compat_os_name != 'nt' and 'TERM' in os.environ:
550 # Restore the title from stack
551 self._write_string('\033[23;0t', self._screen_file)
552
553 def __enter__(self):
554 self.save_console_title()
555 return self
556
557 def __exit__(self, *args):
558 self.restore_console_title()
559
560 if self.params.get('cookiefile') is not None:
561 self.cookiejar.save()
562
563 def trouble(self, message=None, tb=None):
564 """Determine action to take when a download problem appears.
565
566 Depending on if the downloader has been configured to ignore
567 download errors or not, this method may throw an exception or
568 not when errors are found, after printing the message.
569
570 tb, if given, is additional traceback information.
571 """
572 if message is not None:
573 self.to_stderr(message)
574 if self.params.get('verbose'):
575 if tb is None:
576 if sys.exc_info()[0]: # if .trouble has been called from an except block
577 tb = ''
578 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
579 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
580 tb += encode_compat_str(traceback.format_exc())
581 else:
582 tb_data = traceback.format_list(traceback.extract_stack())
583 tb = ''.join(tb_data)
584 self.to_stderr(tb)
585 if not self.params.get('ignoreerrors', False):
586 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
587 exc_info = sys.exc_info()[1].exc_info
588 else:
589 exc_info = sys.exc_info()
590 raise DownloadError(message, exc_info)
591 self._download_retcode = 1
592
593 def report_warning(self, message):
594 '''
595 Print the message to stderr, it will be prefixed with 'WARNING:'
596 If stderr is a tty file the 'WARNING:' will be colored
597 '''
598 if self.params.get('logger') is not None:
599 self.params['logger'].warning(message)
600 else:
601 if self.params.get('no_warnings'):
602 return
603 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
604 _msg_header = '\033[0;33mWARNING:\033[0m'
605 else:
606 _msg_header = 'WARNING:'
607 warning_message = '%s %s' % (_msg_header, message)
608 self.to_stderr(warning_message)
609
610 def report_error(self, message, tb=None):
611 '''
612 Do the same as trouble, but prefixes the message with 'ERROR:', colored
613 in red if stderr is a tty file.
614 '''
615 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
616 _msg_header = '\033[0;31mERROR:\033[0m'
617 else:
618 _msg_header = 'ERROR:'
619 error_message = '%s %s' % (_msg_header, message)
620 self.trouble(error_message, tb)
621
622 def report_file_already_downloaded(self, file_name):
623 """Report file has already been fully downloaded."""
624 try:
625 self.to_screen('[download] %s has already been downloaded' % file_name)
626 except UnicodeEncodeError:
627 self.to_screen('[download] The file has already been downloaded')
628
629 def prepare_filename(self, info_dict):
630 """Generate the output filename."""
631 try:
632 template_dict = dict(info_dict)
633
634 template_dict['epoch'] = int(time.time())
635 autonumber_size = self.params.get('autonumber_size')
636 if autonumber_size is None:
637 autonumber_size = 5
638 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
639 if template_dict.get('resolution') is None:
640 if template_dict.get('width') and template_dict.get('height'):
641 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
642 elif template_dict.get('height'):
643 template_dict['resolution'] = '%sp' % template_dict['height']
644 elif template_dict.get('width'):
645 template_dict['resolution'] = '%dx?' % template_dict['width']
646
647 sanitize = lambda k, v: sanitize_filename(
648 compat_str(v),
649 restricted=self.params.get('restrictfilenames'),
650 is_id=(k == 'id' or k.endswith('_id')))
651 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
652 for k, v in template_dict.items()
653 if v is not None and not isinstance(v, (list, tuple, dict)))
654 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
655
656 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
657
658 # For fields playlist_index and autonumber convert all occurrences
659 # of %(field)s to %(field)0Nd for backward compatibility
660 field_size_compat_map = {
661 'playlist_index': len(str(template_dict['n_entries'])),
662 'autonumber': autonumber_size,
663 }
664 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
665 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
666 if mobj:
667 outtmpl = re.sub(
668 FIELD_SIZE_COMPAT_RE,
669 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
670 outtmpl)
671
672 # Missing numeric fields used together with integer presentation types
673 # in format specification will break the argument substitution since
674 # string 'NA' is returned for missing fields. We will patch output
675 # template for missing fields to meet string presentation type.
676 for numeric_field in self._NUMERIC_FIELDS:
677 if numeric_field not in template_dict:
678 # As of [1] format syntax is:
679 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
680 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
681 FORMAT_RE = r'''(?x)
682 (?<!%)
683 %
684 \({0}\) # mapping key
685 (?:[#0\-+ ]+)? # conversion flags (optional)
686 (?:\d+)? # minimum field width (optional)
687 (?:\.\d+)? # precision (optional)
688 [hlL]? # length modifier (optional)
689 [diouxXeEfFgGcrs%] # conversion type
690 '''
691 outtmpl = re.sub(
692 FORMAT_RE.format(numeric_field),
693 r'%({0})s'.format(numeric_field), outtmpl)
694
695 # expand_path translates '%%' into '%' and '$$' into '$'
696 # correspondingly that is not what we want since we need to keep
697 # '%%' intact for template dict substitution step. Working around
698 # with boundary-alike separator hack.
699 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
700 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
701
702 # outtmpl should be expand_path'ed before template dict substitution
703 # because meta fields may contain env variables we don't want to
704 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
705 # title "Hello $PATH", we don't want `$PATH` to be expanded.
706 filename = expand_path(outtmpl).replace(sep, '') % template_dict
707
708 # Temporary fix for #4787
709 # 'Treat' all problem characters by passing filename through preferredencoding
710 # to workaround encoding issues with subprocess on python2 @ Windows
711 if sys.version_info < (3, 0) and sys.platform == 'win32':
712 filename = encodeFilename(filename, True).decode(preferredencoding())
713 return sanitize_path(filename)
714 except ValueError as err:
715 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
716 return None
717
718 def _match_entry(self, info_dict, incomplete):
719 """ Returns None iff the file should be downloaded """
720
721 video_title = info_dict.get('title', info_dict.get('id', 'video'))
722 if 'title' in info_dict:
723 # This can happen when we're just evaluating the playlist
724 title = info_dict['title']
725 matchtitle = self.params.get('matchtitle', False)
726 if matchtitle:
727 if not re.search(matchtitle, title, re.IGNORECASE):
728 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
729 rejecttitle = self.params.get('rejecttitle', False)
730 if rejecttitle:
731 if re.search(rejecttitle, title, re.IGNORECASE):
732 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
733 date = info_dict.get('upload_date')
734 if date is not None:
735 dateRange = self.params.get('daterange', DateRange())
736 if date not in dateRange:
737 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
738 view_count = info_dict.get('view_count')
739 if view_count is not None:
740 min_views = self.params.get('min_views')
741 if min_views is not None and view_count < min_views:
742 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
743 max_views = self.params.get('max_views')
744 if max_views is not None and view_count > max_views:
745 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
746 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
747 return 'Skipping "%s" because it is age restricted' % video_title
748 if self.in_download_archive(info_dict):
749 return '%s has already been recorded in archive' % video_title
750
751 if not incomplete:
752 match_filter = self.params.get('match_filter')
753 if match_filter is not None:
754 ret = match_filter(info_dict)
755 if ret is not None:
756 return ret
757
758 return None
759
760 @staticmethod
761 def add_extra_info(info_dict, extra_info):
762 '''Set the keys from extra_info in info dict if they are missing'''
763 for key, value in extra_info.items():
764 info_dict.setdefault(key, value)
765
766 def extract_info(self, url, download=True, ie_key=None, extra_info={},
767 process=True, force_generic_extractor=False):
768 '''
769 Returns a list with a dictionary for each video we find.
770 If 'download', also downloads the videos.
771 extra_info is a dict containing the extra values to add to each result
772 '''
773
774 if not ie_key and force_generic_extractor:
775 ie_key = 'Generic'
776
777 if ie_key:
778 ies = [self.get_info_extractor(ie_key)]
779 else:
780 ies = self._ies
781
782 for ie in ies:
783 if not ie.suitable(url):
784 continue
785
786 ie = self.get_info_extractor(ie.ie_key())
787 if not ie.working():
788 self.report_warning('The program functionality for this site has been marked as broken, '
789 'and will probably not work.')
790
791 try:
792 ie_result = ie.extract(url)
793 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
794 break
795 if isinstance(ie_result, list):
796 # Backwards compatibility: old IE result format
797 ie_result = {
798 '_type': 'compat_list',
799 'entries': ie_result,
800 }
801 self.add_default_extra_info(ie_result, ie, url)
802 if process:
803 return self.process_ie_result(ie_result, download, extra_info)
804 else:
805 return ie_result
806 except GeoRestrictedError as e:
807 msg = e.msg
808 if e.countries:
809 msg += '\nThis video is available in %s.' % ', '.join(
810 map(ISO3166Utils.short2full, e.countries))
811 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
812 self.report_error(msg)
813 break
814 except ExtractorError as e: # An error we somewhat expected
815 self.report_error(compat_str(e), e.format_traceback())
816 break
817 except MaxDownloadsReached:
818 raise
819 except Exception as e:
820 if self.params.get('ignoreerrors', False):
821 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
822 break
823 else:
824 raise
825 else:
826 self.report_error('no suitable InfoExtractor for URL %s' % url)
827
828 def add_default_extra_info(self, ie_result, ie, url):
829 self.add_extra_info(ie_result, {
830 'extractor': ie.IE_NAME,
831 'webpage_url': url,
832 'webpage_url_basename': url_basename(url),
833 'extractor_key': ie.ie_key(),
834 })
835
836 def process_ie_result(self, ie_result, download=True, extra_info={}):
837 """
838 Take the result of the ie(may be modified) and resolve all unresolved
839 references (URLs, playlist items).
840
841 It will also download the videos if 'download'.
842 Returns the resolved ie_result.
843 """
844 result_type = ie_result.get('_type', 'video')
845
846 if result_type in ('url', 'url_transparent'):
847 ie_result['url'] = sanitize_url(ie_result['url'])
848 extract_flat = self.params.get('extract_flat', False)
849 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
850 extract_flat is True):
851 if self.params.get('forcejson', False):
852 self.to_stdout(json.dumps(ie_result))
853 return ie_result
854
855 if result_type == 'video':
856 self.add_extra_info(ie_result, extra_info)
857 return self.process_video_result(ie_result, download=download)
858 elif result_type == 'url':
859 # We have to add extra_info to the results because it may be
860 # contained in a playlist
861 return self.extract_info(ie_result['url'],
862 download,
863 ie_key=ie_result.get('ie_key'),
864 extra_info=extra_info)
865 elif result_type == 'url_transparent':
866 # Use the information from the embedding page
867 info = self.extract_info(
868 ie_result['url'], ie_key=ie_result.get('ie_key'),
869 extra_info=extra_info, download=False, process=False)
870
871 # extract_info may return None when ignoreerrors is enabled and
872 # extraction failed with an error, don't crash and return early
873 # in this case
874 if not info:
875 return info
876
877 force_properties = dict(
878 (k, v) for k, v in ie_result.items() if v is not None)
879 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
880 if f in force_properties:
881 del force_properties[f]
882 new_result = info.copy()
883 new_result.update(force_properties)
884
885 # Extracted info may not be a video result (i.e.
886 # info.get('_type', 'video') != video) but rather an url or
887 # url_transparent. In such cases outer metadata (from ie_result)
888 # should be propagated to inner one (info). For this to happen
889 # _type of info should be overridden with url_transparent. This
890 # fixes issue from https://github.com/rg3/youtube-dl/pull/11163.
891 if new_result.get('_type') == 'url':
892 new_result['_type'] = 'url_transparent'
893
894 return self.process_ie_result(
895 new_result, download=download, extra_info=extra_info)
896 elif result_type in ('playlist', 'multi_video'):
897 # We process each entry in the playlist
898 playlist = ie_result.get('title') or ie_result.get('id')
899 self.to_screen('[download] Downloading playlist: %s' % playlist)
900
901 playlist_results = []
902
903 playliststart = self.params.get('playliststart', 1) - 1
904 playlistend = self.params.get('playlistend')
905 # For backwards compatibility, interpret -1 as whole list
906 if playlistend == -1:
907 playlistend = None
908
909 playlistitems_str = self.params.get('playlist_items')
910 playlistitems = None
911 if playlistitems_str is not None:
912 def iter_playlistitems(format):
913 for string_segment in format.split(','):
914 if '-' in string_segment:
915 start, end = string_segment.split('-')
916 for item in range(int(start), int(end) + 1):
917 yield int(item)
918 else:
919 yield int(string_segment)
920 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
921
922 ie_entries = ie_result['entries']
923
924 def make_playlistitems_entries(list_ie_entries):
925 num_entries = len(list_ie_entries)
926 return [
927 list_ie_entries[i - 1] for i in playlistitems
928 if -num_entries <= i - 1 < num_entries]
929
930 def report_download(num_entries):
931 self.to_screen(
932 '[%s] playlist %s: Downloading %d videos' %
933 (ie_result['extractor'], playlist, num_entries))
934
935 if isinstance(ie_entries, list):
936 n_all_entries = len(ie_entries)
937 if playlistitems:
938 entries = make_playlistitems_entries(ie_entries)
939 else:
940 entries = ie_entries[playliststart:playlistend]
941 n_entries = len(entries)
942 self.to_screen(
943 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
944 (ie_result['extractor'], playlist, n_all_entries, n_entries))
945 elif isinstance(ie_entries, PagedList):
946 if playlistitems:
947 entries = []
948 for item in playlistitems:
949 entries.extend(ie_entries.getslice(
950 item - 1, item
951 ))
952 else:
953 entries = ie_entries.getslice(
954 playliststart, playlistend)
955 n_entries = len(entries)
956 report_download(n_entries)
957 else: # iterable
958 if playlistitems:
959 entries = make_playlistitems_entries(list(itertools.islice(
960 ie_entries, 0, max(playlistitems))))
961 else:
962 entries = list(itertools.islice(
963 ie_entries, playliststart, playlistend))
964 n_entries = len(entries)
965 report_download(n_entries)
966
967 if self.params.get('playlistreverse', False):
968 entries = entries[::-1]
969
970 if self.params.get('playlistrandom', False):
971 random.shuffle(entries)
972
973 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
974
975 for i, entry in enumerate(entries, 1):
976 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
977 # This __x_forwarded_for_ip thing is a bit ugly but requires
978 # minimal changes
979 if x_forwarded_for:
980 entry['__x_forwarded_for_ip'] = x_forwarded_for
981 extra = {
982 'n_entries': n_entries,
983 'playlist': playlist,
984 'playlist_id': ie_result.get('id'),
985 'playlist_title': ie_result.get('title'),
986 'playlist_uploader': ie_result.get('uploader'),
987 'playlist_uploader_id': ie_result.get('uploader_id'),
988 'playlist_index': i + playliststart,
989 'extractor': ie_result['extractor'],
990 'webpage_url': ie_result['webpage_url'],
991 'webpage_url_basename': url_basename(ie_result['webpage_url']),
992 'extractor_key': ie_result['extractor_key'],
993 }
994
995 reason = self._match_entry(entry, incomplete=True)
996 if reason is not None:
997 self.to_screen('[download] ' + reason)
998 continue
999
1000 entry_result = self.process_ie_result(entry,
1001 download=download,
1002 extra_info=extra)
1003 playlist_results.append(entry_result)
1004 ie_result['entries'] = playlist_results
1005 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1006 return ie_result
1007 elif result_type == 'compat_list':
1008 self.report_warning(
1009 'Extractor %s returned a compat_list result. '
1010 'It needs to be updated.' % ie_result.get('extractor'))
1011
1012 def _fixup(r):
1013 self.add_extra_info(
1014 r,
1015 {
1016 'extractor': ie_result['extractor'],
1017 'webpage_url': ie_result['webpage_url'],
1018 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1019 'extractor_key': ie_result['extractor_key'],
1020 }
1021 )
1022 return r
1023 ie_result['entries'] = [
1024 self.process_ie_result(_fixup(r), download, extra_info)
1025 for r in ie_result['entries']
1026 ]
1027 return ie_result
1028 else:
1029 raise Exception('Invalid result type: %s' % result_type)
1030
1031 def _build_format_filter(self, filter_spec):
1032 " Returns a function to filter the formats according to the filter_spec "
1033
1034 OPERATORS = {
1035 '<': operator.lt,
1036 '<=': operator.le,
1037 '>': operator.gt,
1038 '>=': operator.ge,
1039 '=': operator.eq,
1040 '!=': operator.ne,
1041 }
1042 operator_rex = re.compile(r'''(?x)\s*
1043 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)
1044 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1045 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1046 $
1047 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1048 m = operator_rex.search(filter_spec)
1049 if m:
1050 try:
1051 comparison_value = int(m.group('value'))
1052 except ValueError:
1053 comparison_value = parse_filesize(m.group('value'))
1054 if comparison_value is None:
1055 comparison_value = parse_filesize(m.group('value') + 'B')
1056 if comparison_value is None:
1057 raise ValueError(
1058 'Invalid value %r in format specification %r' % (
1059 m.group('value'), filter_spec))
1060 op = OPERATORS[m.group('op')]
1061
1062 if not m:
1063 STR_OPERATORS = {
1064 '=': operator.eq,
1065 '!=': operator.ne,
1066 '^=': lambda attr, value: attr.startswith(value),
1067 '$=': lambda attr, value: attr.endswith(value),
1068 '*=': lambda attr, value: value in attr,
1069 }
1070 str_operator_rex = re.compile(r'''(?x)
1071 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
1072 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
1073 \s*(?P<value>[a-zA-Z0-9._-]+)
1074 \s*$
1075 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1076 m = str_operator_rex.search(filter_spec)
1077 if m:
1078 comparison_value = m.group('value')
1079 op = STR_OPERATORS[m.group('op')]
1080
1081 if not m:
1082 raise ValueError('Invalid filter specification %r' % filter_spec)
1083
1084 def _filter(f):
1085 actual_value = f.get(m.group('key'))
1086 if actual_value is None:
1087 return m.group('none_inclusive')
1088 return op(actual_value, comparison_value)
1089 return _filter
1090
1091 def _default_format_spec(self, info_dict, download=True):
1092
1093 def can_merge():
1094 merger = FFmpegMergerPP(self)
1095 return merger.available and merger.can_merge()
1096
1097 def prefer_best():
1098 if self.params.get('simulate', False):
1099 return False
1100 if not download:
1101 return False
1102 if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-':
1103 return True
1104 if info_dict.get('is_live'):
1105 return True
1106 if not can_merge():
1107 return True
1108 return False
1109
1110 req_format_list = ['bestvideo+bestaudio', 'best']
1111 if prefer_best():
1112 req_format_list.reverse()
1113 return '/'.join(req_format_list)
1114
1115 def build_format_selector(self, format_spec):
1116 def syntax_error(note, start):
1117 message = (
1118 'Invalid format specification: '
1119 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1120 return SyntaxError(message)
1121
1122 PICKFIRST = 'PICKFIRST'
1123 MERGE = 'MERGE'
1124 SINGLE = 'SINGLE'
1125 GROUP = 'GROUP'
1126 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1127
1128 def _parse_filter(tokens):
1129 filter_parts = []
1130 for type, string, start, _, _ in tokens:
1131 if type == tokenize.OP and string == ']':
1132 return ''.join(filter_parts)
1133 else:
1134 filter_parts.append(string)
1135
1136 def _remove_unused_ops(tokens):
1137 # Remove operators that we don't use and join them with the surrounding strings
1138 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1139 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1140 last_string, last_start, last_end, last_line = None, None, None, None
1141 for type, string, start, end, line in tokens:
1142 if type == tokenize.OP and string == '[':
1143 if last_string:
1144 yield tokenize.NAME, last_string, last_start, last_end, last_line
1145 last_string = None
1146 yield type, string, start, end, line
1147 # everything inside brackets will be handled by _parse_filter
1148 for type, string, start, end, line in tokens:
1149 yield type, string, start, end, line
1150 if type == tokenize.OP and string == ']':
1151 break
1152 elif type == tokenize.OP and string in ALLOWED_OPS:
1153 if last_string:
1154 yield tokenize.NAME, last_string, last_start, last_end, last_line
1155 last_string = None
1156 yield type, string, start, end, line
1157 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1158 if not last_string:
1159 last_string = string
1160 last_start = start
1161 last_end = end
1162 else:
1163 last_string += string
1164 if last_string:
1165 yield tokenize.NAME, last_string, last_start, last_end, last_line
1166
1167 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1168 selectors = []
1169 current_selector = None
1170 for type, string, start, _, _ in tokens:
1171 # ENCODING is only defined in python 3.x
1172 if type == getattr(tokenize, 'ENCODING', None):
1173 continue
1174 elif type in [tokenize.NAME, tokenize.NUMBER]:
1175 current_selector = FormatSelector(SINGLE, string, [])
1176 elif type == tokenize.OP:
1177 if string == ')':
1178 if not inside_group:
1179 # ')' will be handled by the parentheses group
1180 tokens.restore_last_token()
1181 break
1182 elif inside_merge and string in ['/', ',']:
1183 tokens.restore_last_token()
1184 break
1185 elif inside_choice and string == ',':
1186 tokens.restore_last_token()
1187 break
1188 elif string == ',':
1189 if not current_selector:
1190 raise syntax_error('"," must follow a format selector', start)
1191 selectors.append(current_selector)
1192 current_selector = None
1193 elif string == '/':
1194 if not current_selector:
1195 raise syntax_error('"/" must follow a format selector', start)
1196 first_choice = current_selector
1197 second_choice = _parse_format_selection(tokens, inside_choice=True)
1198 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1199 elif string == '[':
1200 if not current_selector:
1201 current_selector = FormatSelector(SINGLE, 'best', [])
1202 format_filter = _parse_filter(tokens)
1203 current_selector.filters.append(format_filter)
1204 elif string == '(':
1205 if current_selector:
1206 raise syntax_error('Unexpected "("', start)
1207 group = _parse_format_selection(tokens, inside_group=True)
1208 current_selector = FormatSelector(GROUP, group, [])
1209 elif string == '+':
1210 video_selector = current_selector
1211 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1212 if not video_selector or not audio_selector:
1213 raise syntax_error('"+" must be between two format selectors', start)
1214 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1215 else:
1216 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1217 elif type == tokenize.ENDMARKER:
1218 break
1219 if current_selector:
1220 selectors.append(current_selector)
1221 return selectors
1222
1223 def _build_selector_function(selector):
1224 if isinstance(selector, list):
1225 fs = [_build_selector_function(s) for s in selector]
1226
1227 def selector_function(ctx):
1228 for f in fs:
1229 for format in f(ctx):
1230 yield format
1231 return selector_function
1232 elif selector.type == GROUP:
1233 selector_function = _build_selector_function(selector.selector)
1234 elif selector.type == PICKFIRST:
1235 fs = [_build_selector_function(s) for s in selector.selector]
1236
1237 def selector_function(ctx):
1238 for f in fs:
1239 picked_formats = list(f(ctx))
1240 if picked_formats:
1241 return picked_formats
1242 return []
1243 elif selector.type == SINGLE:
1244 format_spec = selector.selector
1245
1246 def selector_function(ctx):
1247 formats = list(ctx['formats'])
1248 if not formats:
1249 return
1250 if format_spec == 'all':
1251 for f in formats:
1252 yield f
1253 elif format_spec in ['best', 'worst', None]:
1254 format_idx = 0 if format_spec == 'worst' else -1
1255 audiovideo_formats = [
1256 f for f in formats
1257 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1258 if audiovideo_formats:
1259 yield audiovideo_formats[format_idx]
1260 # for extractors with incomplete formats (audio only (soundcloud)
1261 # or video only (imgur)) we will fallback to best/worst
1262 # {video,audio}-only format
1263 elif ctx['incomplete_formats']:
1264 yield formats[format_idx]
1265 elif format_spec == 'bestaudio':
1266 audio_formats = [
1267 f for f in formats
1268 if f.get('vcodec') == 'none']
1269 if audio_formats:
1270 yield audio_formats[-1]
1271 elif format_spec == 'worstaudio':
1272 audio_formats = [
1273 f for f in formats
1274 if f.get('vcodec') == 'none']
1275 if audio_formats:
1276 yield audio_formats[0]
1277 elif format_spec == 'bestvideo':
1278 video_formats = [
1279 f for f in formats
1280 if f.get('acodec') == 'none']
1281 if video_formats:
1282 yield video_formats[-1]
1283 elif format_spec == 'worstvideo':
1284 video_formats = [
1285 f for f in formats
1286 if f.get('acodec') == 'none']
1287 if video_formats:
1288 yield video_formats[0]
1289 else:
1290 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1291 if format_spec in extensions:
1292 filter_f = lambda f: f['ext'] == format_spec
1293 else:
1294 filter_f = lambda f: f['format_id'] == format_spec
1295 matches = list(filter(filter_f, formats))
1296 if matches:
1297 yield matches[-1]
1298 elif selector.type == MERGE:
1299 def _merge(formats_info):
1300 format_1, format_2 = [f['format_id'] for f in formats_info]
1301 # The first format must contain the video and the
1302 # second the audio
1303 if formats_info[0].get('vcodec') == 'none':
1304 self.report_error('The first format must '
1305 'contain the video, try using '
1306 '"-f %s+%s"' % (format_2, format_1))
1307 return
1308 # Formats must be opposite (video+audio)
1309 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1310 self.report_error(
1311 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1312 % (format_1, format_2))
1313 return
1314 output_ext = (
1315 formats_info[0]['ext']
1316 if self.params.get('merge_output_format') is None
1317 else self.params['merge_output_format'])
1318 return {
1319 'requested_formats': formats_info,
1320 'format': '%s+%s' % (formats_info[0].get('format'),
1321 formats_info[1].get('format')),
1322 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1323 formats_info[1].get('format_id')),
1324 'width': formats_info[0].get('width'),
1325 'height': formats_info[0].get('height'),
1326 'resolution': formats_info[0].get('resolution'),
1327 'fps': formats_info[0].get('fps'),
1328 'vcodec': formats_info[0].get('vcodec'),
1329 'vbr': formats_info[0].get('vbr'),
1330 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1331 'acodec': formats_info[1].get('acodec'),
1332 'abr': formats_info[1].get('abr'),
1333 'ext': output_ext,
1334 }
1335 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1336
1337 def selector_function(ctx):
1338 for pair in itertools.product(
1339 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1340 yield _merge(pair)
1341
1342 filters = [self._build_format_filter(f) for f in selector.filters]
1343
1344 def final_selector(ctx):
1345 ctx_copy = copy.deepcopy(ctx)
1346 for _filter in filters:
1347 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1348 return selector_function(ctx_copy)
1349 return final_selector
1350
1351 stream = io.BytesIO(format_spec.encode('utf-8'))
1352 try:
1353 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1354 except tokenize.TokenError:
1355 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1356
1357 class TokenIterator(object):
1358 def __init__(self, tokens):
1359 self.tokens = tokens
1360 self.counter = 0
1361
1362 def __iter__(self):
1363 return self
1364
1365 def __next__(self):
1366 if self.counter >= len(self.tokens):
1367 raise StopIteration()
1368 value = self.tokens[self.counter]
1369 self.counter += 1
1370 return value
1371
1372 next = __next__
1373
1374 def restore_last_token(self):
1375 self.counter -= 1
1376
1377 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1378 return _build_selector_function(parsed_selector)
1379
1380 def _calc_headers(self, info_dict):
1381 res = std_headers.copy()
1382
1383 add_headers = info_dict.get('http_headers')
1384 if add_headers:
1385 res.update(add_headers)
1386
1387 cookies = self._calc_cookies(info_dict)
1388 if cookies:
1389 res['Cookie'] = cookies
1390
1391 if 'X-Forwarded-For' not in res:
1392 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1393 if x_forwarded_for_ip:
1394 res['X-Forwarded-For'] = x_forwarded_for_ip
1395
1396 return res
1397
1398 def _calc_cookies(self, info_dict):
1399 pr = sanitized_Request(info_dict['url'])
1400 self.cookiejar.add_cookie_header(pr)
1401 return pr.get_header('Cookie')
1402
1403 def process_video_result(self, info_dict, download=True):
1404 assert info_dict.get('_type', 'video') == 'video'
1405
1406 if 'id' not in info_dict:
1407 raise ExtractorError('Missing "id" field in extractor result')
1408 if 'title' not in info_dict:
1409 raise ExtractorError('Missing "title" field in extractor result')
1410
1411 def report_force_conversion(field, field_not, conversion):
1412 self.report_warning(
1413 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1414 % (field, field_not, conversion))
1415
1416 def sanitize_string_field(info, string_field):
1417 field = info.get(string_field)
1418 if field is None or isinstance(field, compat_str):
1419 return
1420 report_force_conversion(string_field, 'a string', 'string')
1421 info[string_field] = compat_str(field)
1422
1423 def sanitize_numeric_fields(info):
1424 for numeric_field in self._NUMERIC_FIELDS:
1425 field = info.get(numeric_field)
1426 if field is None or isinstance(field, compat_numeric_types):
1427 continue
1428 report_force_conversion(numeric_field, 'numeric', 'int')
1429 info[numeric_field] = int_or_none(field)
1430
1431 sanitize_string_field(info_dict, 'id')
1432 sanitize_numeric_fields(info_dict)
1433
1434 if 'playlist' not in info_dict:
1435 # It isn't part of a playlist
1436 info_dict['playlist'] = None
1437 info_dict['playlist_index'] = None
1438
1439 thumbnails = info_dict.get('thumbnails')
1440 if thumbnails is None:
1441 thumbnail = info_dict.get('thumbnail')
1442 if thumbnail:
1443 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1444 if thumbnails:
1445 thumbnails.sort(key=lambda t: (
1446 t.get('preference') if t.get('preference') is not None else -1,
1447 t.get('width') if t.get('width') is not None else -1,
1448 t.get('height') if t.get('height') is not None else -1,
1449 t.get('id') if t.get('id') is not None else '', t.get('url')))
1450 for i, t in enumerate(thumbnails):
1451 t['url'] = sanitize_url(t['url'])
1452 if t.get('width') and t.get('height'):
1453 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1454 if t.get('id') is None:
1455 t['id'] = '%d' % i
1456
1457 if self.params.get('list_thumbnails'):
1458 self.list_thumbnails(info_dict)
1459 return
1460
1461 thumbnail = info_dict.get('thumbnail')
1462 if thumbnail:
1463 info_dict['thumbnail'] = sanitize_url(thumbnail)
1464 elif thumbnails:
1465 info_dict['thumbnail'] = thumbnails[-1]['url']
1466
1467 if 'display_id' not in info_dict and 'id' in info_dict:
1468 info_dict['display_id'] = info_dict['id']
1469
1470 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1471 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1472 # see http://bugs.python.org/issue1646728)
1473 try:
1474 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1475 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1476 except (ValueError, OverflowError, OSError):
1477 pass
1478
1479 # Auto generate title fields corresponding to the *_number fields when missing
1480 # in order to always have clean titles. This is very common for TV series.
1481 for field in ('chapter', 'season', 'episode'):
1482 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1483 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1484
1485 for cc_kind in ('subtitles', 'automatic_captions'):
1486 cc = info_dict.get(cc_kind)
1487 if cc:
1488 for _, subtitle in cc.items():
1489 for subtitle_format in subtitle:
1490 if subtitle_format.get('url'):
1491 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1492 if subtitle_format.get('ext') is None:
1493 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1494
1495 automatic_captions = info_dict.get('automatic_captions')
1496 subtitles = info_dict.get('subtitles')
1497
1498 if self.params.get('listsubtitles', False):
1499 if 'automatic_captions' in info_dict:
1500 self.list_subtitles(
1501 info_dict['id'], automatic_captions, 'automatic captions')
1502 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1503 return
1504
1505 info_dict['requested_subtitles'] = self.process_subtitles(
1506 info_dict['id'], subtitles, automatic_captions)
1507
1508 # We now pick which formats have to be downloaded
1509 if info_dict.get('formats') is None:
1510 # There's only one format available
1511 formats = [info_dict]
1512 else:
1513 formats = info_dict['formats']
1514
1515 if not formats:
1516 raise ExtractorError('No video formats found!')
1517
1518 def is_wellformed(f):
1519 url = f.get('url')
1520 if not url:
1521 self.report_warning(
1522 '"url" field is missing or empty - skipping format, '
1523 'there is an error in extractor')
1524 return False
1525 if isinstance(url, bytes):
1526 sanitize_string_field(f, 'url')
1527 return True
1528
1529 # Filter out malformed formats for better extraction robustness
1530 formats = list(filter(is_wellformed, formats))
1531
1532 formats_dict = {}
1533
1534 # We check that all the formats have the format and format_id fields
1535 for i, format in enumerate(formats):
1536 sanitize_string_field(format, 'format_id')
1537 sanitize_numeric_fields(format)
1538 format['url'] = sanitize_url(format['url'])
1539 if not format.get('format_id'):
1540 format['format_id'] = compat_str(i)
1541 else:
1542 # Sanitize format_id from characters used in format selector expression
1543 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1544 format_id = format['format_id']
1545 if format_id not in formats_dict:
1546 formats_dict[format_id] = []
1547 formats_dict[format_id].append(format)
1548
1549 # Make sure all formats have unique format_id
1550 for format_id, ambiguous_formats in formats_dict.items():
1551 if len(ambiguous_formats) > 1:
1552 for i, format in enumerate(ambiguous_formats):
1553 format['format_id'] = '%s-%d' % (format_id, i)
1554
1555 for i, format in enumerate(formats):
1556 if format.get('format') is None:
1557 format['format'] = '{id} - {res}{note}'.format(
1558 id=format['format_id'],
1559 res=self.format_resolution(format),
1560 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1561 )
1562 # Automatically determine file extension if missing
1563 if format.get('ext') is None:
1564 format['ext'] = determine_ext(format['url']).lower()
1565 # Automatically determine protocol if missing (useful for format
1566 # selection purposes)
1567 if format.get('protocol') is None:
1568 format['protocol'] = determine_protocol(format)
1569 # Add HTTP headers, so that external programs can use them from the
1570 # json output
1571 full_format_info = info_dict.copy()
1572 full_format_info.update(format)
1573 format['http_headers'] = self._calc_headers(full_format_info)
1574 # Remove private housekeeping stuff
1575 if '__x_forwarded_for_ip' in info_dict:
1576 del info_dict['__x_forwarded_for_ip']
1577
1578 # TODO Central sorting goes here
1579
1580 if formats[0] is not info_dict:
1581 # only set the 'formats' fields if the original info_dict list them
1582 # otherwise we end up with a circular reference, the first (and unique)
1583 # element in the 'formats' field in info_dict is info_dict itself,
1584 # which can't be exported to json
1585 info_dict['formats'] = formats
1586 if self.params.get('listformats'):
1587 self.list_formats(info_dict)
1588 return
1589
1590 req_format = self.params.get('format')
1591 if req_format is None:
1592 req_format = self._default_format_spec(info_dict, download=download)
1593 if self.params.get('verbose'):
1594 self.to_stdout('[debug] Default format spec: %s' % req_format)
1595
1596 format_selector = self.build_format_selector(req_format)
1597
1598 # While in format selection we may need to have an access to the original
1599 # format set in order to calculate some metrics or do some processing.
1600 # For now we need to be able to guess whether original formats provided
1601 # by extractor are incomplete or not (i.e. whether extractor provides only
1602 # video-only or audio-only formats) for proper formats selection for
1603 # extractors with such incomplete formats (see
1604 # https://github.com/rg3/youtube-dl/pull/5556).
1605 # Since formats may be filtered during format selection and may not match
1606 # the original formats the results may be incorrect. Thus original formats
1607 # or pre-calculated metrics should be passed to format selection routines
1608 # as well.
1609 # We will pass a context object containing all necessary additional data
1610 # instead of just formats.
1611 # This fixes incorrect format selection issue (see
1612 # https://github.com/rg3/youtube-dl/issues/10083).
1613 incomplete_formats = (
1614 # All formats are video-only or
1615 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1616 # all formats are audio-only
1617 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1618
1619 ctx = {
1620 'formats': formats,
1621 'incomplete_formats': incomplete_formats,
1622 }
1623
1624 formats_to_download = list(format_selector(ctx))
1625 if not formats_to_download:
1626 raise ExtractorError('requested format not available',
1627 expected=True)
1628
1629 if download:
1630 if len(formats_to_download) > 1:
1631 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1632 for format in formats_to_download:
1633 new_info = dict(info_dict)
1634 new_info.update(format)
1635 self.process_info(new_info)
1636 # We update the info dict with the best quality format (backwards compatibility)
1637 info_dict.update(formats_to_download[-1])
1638 return info_dict
1639
1640 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1641 """Select the requested subtitles and their format"""
1642 available_subs = {}
1643 if normal_subtitles and self.params.get('writesubtitles'):
1644 available_subs.update(normal_subtitles)
1645 if automatic_captions and self.params.get('writeautomaticsub'):
1646 for lang, cap_info in automatic_captions.items():
1647 if lang not in available_subs:
1648 available_subs[lang] = cap_info
1649
1650 if (not self.params.get('writesubtitles') and not
1651 self.params.get('writeautomaticsub') or not
1652 available_subs):
1653 return None
1654
1655 if self.params.get('allsubtitles', False):
1656 requested_langs = available_subs.keys()
1657 else:
1658 if self.params.get('subtitleslangs', False):
1659 requested_langs = self.params.get('subtitleslangs')
1660 elif 'en' in available_subs:
1661 requested_langs = ['en']
1662 else:
1663 requested_langs = [list(available_subs.keys())[0]]
1664
1665 formats_query = self.params.get('subtitlesformat', 'best')
1666 formats_preference = formats_query.split('/') if formats_query else []
1667 subs = {}
1668 for lang in requested_langs:
1669 formats = available_subs.get(lang)
1670 if formats is None:
1671 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1672 continue
1673 for ext in formats_preference:
1674 if ext == 'best':
1675 f = formats[-1]
1676 break
1677 matches = list(filter(lambda f: f['ext'] == ext, formats))
1678 if matches:
1679 f = matches[-1]
1680 break
1681 else:
1682 f = formats[-1]
1683 self.report_warning(
1684 'No subtitle format found matching "%s" for language %s, '
1685 'using %s' % (formats_query, lang, f['ext']))
1686 subs[lang] = f
1687 return subs
1688
1689 def process_info(self, info_dict):
1690 """Process a single resolved IE result."""
1691
1692 assert info_dict.get('_type', 'video') == 'video'
1693
1694 max_downloads = self.params.get('max_downloads')
1695 if max_downloads is not None:
1696 if self._num_downloads >= int(max_downloads):
1697 raise MaxDownloadsReached()
1698
1699 info_dict['fulltitle'] = info_dict['title']
1700 if len(info_dict['title']) > 200:
1701 info_dict['title'] = info_dict['title'][:197] + '...'
1702
1703 if 'format' not in info_dict:
1704 info_dict['format'] = info_dict['ext']
1705
1706 reason = self._match_entry(info_dict, incomplete=False)
1707 if reason is not None:
1708 self.to_screen('[download] ' + reason)
1709 return
1710
1711 self._num_downloads += 1
1712
1713 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1714
1715 # Forced printings
1716 if self.params.get('forcetitle', False):
1717 self.to_stdout(info_dict['fulltitle'])
1718 if self.params.get('forceid', False):
1719 self.to_stdout(info_dict['id'])
1720 if self.params.get('forceurl', False):
1721 if info_dict.get('requested_formats') is not None:
1722 for f in info_dict['requested_formats']:
1723 self.to_stdout(f['url'] + f.get('play_path', ''))
1724 else:
1725 # For RTMP URLs, also include the playpath
1726 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1727 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1728 self.to_stdout(info_dict['thumbnail'])
1729 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1730 self.to_stdout(info_dict['description'])
1731 if self.params.get('forcefilename', False) and filename is not None:
1732 self.to_stdout(filename)
1733 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1734 self.to_stdout(formatSeconds(info_dict['duration']))
1735 if self.params.get('forceformat', False):
1736 self.to_stdout(info_dict['format'])
1737 if self.params.get('forcejson', False):
1738 self.to_stdout(json.dumps(info_dict))
1739
1740 # Do nothing else if in simulate mode
1741 if self.params.get('simulate', False):
1742 return
1743
1744 if filename is None:
1745 return
1746
1747 def ensure_dir_exists(path):
1748 try:
1749 dn = os.path.dirname(path)
1750 if dn and not os.path.exists(dn):
1751 os.makedirs(dn)
1752 return True
1753 except (OSError, IOError) as err:
1754 self.report_error('unable to create directory ' + error_to_compat_str(err))
1755 return False
1756
1757 if not ensure_dir_exists(sanitize_path(encodeFilename(filename))):
1758 return
1759
1760 if self.params.get('writedescription', False):
1761 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1762 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1763 self.to_screen('[info] Video description is already present')
1764 elif info_dict.get('description') is None:
1765 self.report_warning('There\'s no description to write.')
1766 else:
1767 try:
1768 self.to_screen('[info] Writing video description to: ' + descfn)
1769 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1770 descfile.write(info_dict['description'])
1771 except (OSError, IOError):
1772 self.report_error('Cannot write description file ' + descfn)
1773 return
1774
1775 if self.params.get('writeannotations', False):
1776 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1777 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1778 self.to_screen('[info] Video annotations are already present')
1779 else:
1780 try:
1781 self.to_screen('[info] Writing video annotations to: ' + annofn)
1782 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1783 annofile.write(info_dict['annotations'])
1784 except (KeyError, TypeError):
1785 self.report_warning('There are no annotations to write.')
1786 except (OSError, IOError):
1787 self.report_error('Cannot write annotations file: ' + annofn)
1788 return
1789
1790 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1791 self.params.get('writeautomaticsub')])
1792
1793 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1794 # subtitles download errors are already managed as troubles in relevant IE
1795 # that way it will silently go on when used with unsupporting IE
1796 subtitles = info_dict['requested_subtitles']
1797 ie = self.get_info_extractor(info_dict['extractor_key'])
1798 for sub_lang, sub_info in subtitles.items():
1799 sub_format = sub_info['ext']
1800 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1801 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1802 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
1803 else:
1804 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1805 if sub_info.get('data') is not None:
1806 try:
1807 # Use newline='' to prevent conversion of newline characters
1808 # See https://github.com/rg3/youtube-dl/issues/10268
1809 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1810 subfile.write(sub_info['data'])
1811 except (OSError, IOError):
1812 self.report_error('Cannot write subtitles file ' + sub_filename)
1813 return
1814 else:
1815 try:
1816 sub_data = ie._request_webpage(
1817 sub_info['url'], info_dict['id'], note=False).read()
1818 with io.open(encodeFilename(sub_filename), 'wb') as subfile:
1819 subfile.write(sub_data)
1820 except (ExtractorError, IOError, OSError, ValueError) as err:
1821 self.report_warning('Unable to download subtitle for "%s": %s' %
1822 (sub_lang, error_to_compat_str(err)))
1823 continue
1824
1825 if self.params.get('writeinfojson', False):
1826 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1827 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1828 self.to_screen('[info] Video description metadata is already present')
1829 else:
1830 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1831 try:
1832 write_json_file(self.filter_requested_info(info_dict), infofn)
1833 except (OSError, IOError):
1834 self.report_error('Cannot write metadata to JSON file ' + infofn)
1835 return
1836
1837 self._write_thumbnails(info_dict, filename)
1838
1839 if not self.params.get('skip_download', False):
1840 try:
1841 def dl(name, info):
1842 fd = get_suitable_downloader(info, self.params)(self, self.params)
1843 for ph in self._progress_hooks:
1844 fd.add_progress_hook(ph)
1845 if self.params.get('verbose'):
1846 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1847 return fd.download(name, info)
1848
1849 if info_dict.get('requested_formats') is not None:
1850 downloaded = []
1851 success = True
1852 merger = FFmpegMergerPP(self)
1853 if not merger.available:
1854 postprocessors = []
1855 self.report_warning('You have requested multiple '
1856 'formats but ffmpeg or avconv are not installed.'
1857 ' The formats won\'t be merged.')
1858 else:
1859 postprocessors = [merger]
1860
1861 def compatible_formats(formats):
1862 video, audio = formats
1863 # Check extension
1864 video_ext, audio_ext = video.get('ext'), audio.get('ext')
1865 if video_ext and audio_ext:
1866 COMPATIBLE_EXTS = (
1867 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
1868 ('webm')
1869 )
1870 for exts in COMPATIBLE_EXTS:
1871 if video_ext in exts and audio_ext in exts:
1872 return True
1873 # TODO: Check acodec/vcodec
1874 return False
1875
1876 filename_real_ext = os.path.splitext(filename)[1][1:]
1877 filename_wo_ext = (
1878 os.path.splitext(filename)[0]
1879 if filename_real_ext == info_dict['ext']
1880 else filename)
1881 requested_formats = info_dict['requested_formats']
1882 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1883 info_dict['ext'] = 'mkv'
1884 self.report_warning(
1885 'Requested formats are incompatible for merge and will be merged into mkv.')
1886 # Ensure filename always has a correct extension for successful merge
1887 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1888 if os.path.exists(encodeFilename(filename)):
1889 self.to_screen(
1890 '[download] %s has already been downloaded and '
1891 'merged' % filename)
1892 else:
1893 for f in requested_formats:
1894 new_info = dict(info_dict)
1895 new_info.update(f)
1896 fname = prepend_extension(
1897 self.prepare_filename(new_info),
1898 'f%s' % f['format_id'], new_info['ext'])
1899 if not ensure_dir_exists(fname):
1900 return
1901 downloaded.append(fname)
1902 partial_success = dl(fname, new_info)
1903 success = success and partial_success
1904 info_dict['__postprocessors'] = postprocessors
1905 info_dict['__files_to_merge'] = downloaded
1906 else:
1907 # Just a single file
1908 success = dl(filename, info_dict)
1909 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1910 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1911 return
1912 except (OSError, IOError) as err:
1913 raise UnavailableVideoError(err)
1914 except (ContentTooShortError, ) as err:
1915 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1916 return
1917
1918 if success and filename != '-':
1919 # Fixup content
1920 fixup_policy = self.params.get('fixup')
1921 if fixup_policy is None:
1922 fixup_policy = 'detect_or_warn'
1923
1924 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1925
1926 stretched_ratio = info_dict.get('stretched_ratio')
1927 if stretched_ratio is not None and stretched_ratio != 1:
1928 if fixup_policy == 'warn':
1929 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1930 info_dict['id'], stretched_ratio))
1931 elif fixup_policy == 'detect_or_warn':
1932 stretched_pp = FFmpegFixupStretchedPP(self)
1933 if stretched_pp.available:
1934 info_dict.setdefault('__postprocessors', [])
1935 info_dict['__postprocessors'].append(stretched_pp)
1936 else:
1937 self.report_warning(
1938 '%s: Non-uniform pixel ratio (%s). %s'
1939 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1940 else:
1941 assert fixup_policy in ('ignore', 'never')
1942
1943 if (info_dict.get('requested_formats') is None and
1944 info_dict.get('container') == 'm4a_dash'):
1945 if fixup_policy == 'warn':
1946 self.report_warning(
1947 '%s: writing DASH m4a. '
1948 'Only some players support this container.'
1949 % info_dict['id'])
1950 elif fixup_policy == 'detect_or_warn':
1951 fixup_pp = FFmpegFixupM4aPP(self)
1952 if fixup_pp.available:
1953 info_dict.setdefault('__postprocessors', [])
1954 info_dict['__postprocessors'].append(fixup_pp)
1955 else:
1956 self.report_warning(
1957 '%s: writing DASH m4a. '
1958 'Only some players support this container. %s'
1959 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1960 else:
1961 assert fixup_policy in ('ignore', 'never')
1962
1963 if (info_dict.get('protocol') == 'm3u8_native' or
1964 info_dict.get('protocol') == 'm3u8' and
1965 self.params.get('hls_prefer_native')):
1966 if fixup_policy == 'warn':
1967 self.report_warning('%s: malformed AAC bitstream detected.' % (
1968 info_dict['id']))
1969 elif fixup_policy == 'detect_or_warn':
1970 fixup_pp = FFmpegFixupM3u8PP(self)
1971 if fixup_pp.available:
1972 info_dict.setdefault('__postprocessors', [])
1973 info_dict['__postprocessors'].append(fixup_pp)
1974 else:
1975 self.report_warning(
1976 '%s: malformed AAC bitstream detected. %s'
1977 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1978 else:
1979 assert fixup_policy in ('ignore', 'never')
1980
1981 try:
1982 self.post_process(filename, info_dict)
1983 except (PostProcessingError) as err:
1984 self.report_error('postprocessing: %s' % str(err))
1985 return
1986 self.record_download_archive(info_dict)
1987
1988 def download(self, url_list):
1989 """Download a given list of URLs."""
1990 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1991 if (len(url_list) > 1 and
1992 outtmpl != '-' and
1993 '%' not in outtmpl and
1994 self.params.get('max_downloads') != 1):
1995 raise SameFileError(outtmpl)
1996
1997 for url in url_list:
1998 try:
1999 # It also downloads the videos
2000 res = self.extract_info(
2001 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2002 except UnavailableVideoError:
2003 self.report_error('unable to download video')
2004 except MaxDownloadsReached:
2005 self.to_screen('[info] Maximum number of downloaded files reached.')
2006 raise
2007 else:
2008 if self.params.get('dump_single_json', False):
2009 self.to_stdout(json.dumps(res))
2010
2011 return self._download_retcode
2012
2013 def download_with_info_file(self, info_filename):
2014 with contextlib.closing(fileinput.FileInput(
2015 [info_filename], mode='r',
2016 openhook=fileinput.hook_encoded('utf-8'))) as f:
2017 # FileInput doesn't have a read method, we can't call json.load
2018 info = self.filter_requested_info(json.loads('\n'.join(f)))
2019 try:
2020 self.process_ie_result(info, download=True)
2021 except DownloadError:
2022 webpage_url = info.get('webpage_url')
2023 if webpage_url is not None:
2024 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2025 return self.download([webpage_url])
2026 else:
2027 raise
2028 return self._download_retcode
2029
2030 @staticmethod
2031 def filter_requested_info(info_dict):
2032 return dict(
2033 (k, v) for k, v in info_dict.items()
2034 if k not in ['requested_formats', 'requested_subtitles'])
2035
2036 def post_process(self, filename, ie_info):
2037 """Run all the postprocessors on the given file."""
2038 info = dict(ie_info)
2039 info['filepath'] = filename
2040 pps_chain = []
2041 if ie_info.get('__postprocessors') is not None:
2042 pps_chain.extend(ie_info['__postprocessors'])
2043 pps_chain.extend(self._pps)
2044 for pp in pps_chain:
2045 files_to_delete = []
2046 try:
2047 files_to_delete, info = pp.run(info)
2048 except PostProcessingError as e:
2049 self.report_error(e.msg)
2050 if files_to_delete and not self.params.get('keepvideo', False):
2051 for old_filename in files_to_delete:
2052 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2053 try:
2054 os.remove(encodeFilename(old_filename))
2055 except (IOError, OSError):
2056 self.report_warning('Unable to remove downloaded original file')
2057
2058 def _make_archive_id(self, info_dict):
2059 # Future-proof against any change in case
2060 # and backwards compatibility with prior versions
2061 extractor = info_dict.get('extractor_key')
2062 if extractor is None:
2063 if 'id' in info_dict:
2064 extractor = info_dict.get('ie_key') # key in a playlist
2065 if extractor is None:
2066 return None # Incomplete video information
2067 return extractor.lower() + ' ' + info_dict['id']
2068
2069 def in_download_archive(self, info_dict):
2070 fn = self.params.get('download_archive')
2071 if fn is None:
2072 return False
2073
2074 vid_id = self._make_archive_id(info_dict)
2075 if vid_id is None:
2076 return False # Incomplete video information
2077
2078 try:
2079 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
2080 for line in archive_file:
2081 if line.strip() == vid_id:
2082 return True
2083 except IOError as ioe:
2084 if ioe.errno != errno.ENOENT:
2085 raise
2086 return False
2087
2088 def record_download_archive(self, info_dict):
2089 fn = self.params.get('download_archive')
2090 if fn is None:
2091 return
2092 vid_id = self._make_archive_id(info_dict)
2093 assert vid_id
2094 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2095 archive_file.write(vid_id + '\n')
2096
2097 @staticmethod
2098 def format_resolution(format, default='unknown'):
2099 if format.get('vcodec') == 'none':
2100 return 'audio only'
2101 if format.get('resolution') is not None:
2102 return format['resolution']
2103 if format.get('height') is not None:
2104 if format.get('width') is not None:
2105 res = '%sx%s' % (format['width'], format['height'])
2106 else:
2107 res = '%sp' % format['height']
2108 elif format.get('width') is not None:
2109 res = '%dx?' % format['width']
2110 else:
2111 res = default
2112 return res
2113
2114 def _format_note(self, fdict):
2115 res = ''
2116 if fdict.get('ext') in ['f4f', 'f4m']:
2117 res += '(unsupported) '
2118 if fdict.get('language'):
2119 if res:
2120 res += ' '
2121 res += '[%s] ' % fdict['language']
2122 if fdict.get('format_note') is not None:
2123 res += fdict['format_note'] + ' '
2124 if fdict.get('tbr') is not None:
2125 res += '%4dk ' % fdict['tbr']
2126 if fdict.get('container') is not None:
2127 if res:
2128 res += ', '
2129 res += '%s container' % fdict['container']
2130 if (fdict.get('vcodec') is not None and
2131 fdict.get('vcodec') != 'none'):
2132 if res:
2133 res += ', '
2134 res += fdict['vcodec']
2135 if fdict.get('vbr') is not None:
2136 res += '@'
2137 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2138 res += 'video@'
2139 if fdict.get('vbr') is not None:
2140 res += '%4dk' % fdict['vbr']
2141 if fdict.get('fps') is not None:
2142 if res:
2143 res += ', '
2144 res += '%sfps' % fdict['fps']
2145 if fdict.get('acodec') is not None:
2146 if res:
2147 res += ', '
2148 if fdict['acodec'] == 'none':
2149 res += 'video only'
2150 else:
2151 res += '%-5s' % fdict['acodec']
2152 elif fdict.get('abr') is not None:
2153 if res:
2154 res += ', '
2155 res += 'audio'
2156 if fdict.get('abr') is not None:
2157 res += '@%3dk' % fdict['abr']
2158 if fdict.get('asr') is not None:
2159 res += ' (%5dHz)' % fdict['asr']
2160 if fdict.get('filesize') is not None:
2161 if res:
2162 res += ', '
2163 res += format_bytes(fdict['filesize'])
2164 elif fdict.get('filesize_approx') is not None:
2165 if res:
2166 res += ', '
2167 res += '~' + format_bytes(fdict['filesize_approx'])
2168 return res
2169
2170 def list_formats(self, info_dict):
2171 formats = info_dict.get('formats', [info_dict])
2172 table = [
2173 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
2174 for f in formats
2175 if f.get('preference') is None or f['preference'] >= -1000]
2176 if len(formats) > 1:
2177 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2178
2179 header_line = ['format code', 'extension', 'resolution', 'note']
2180 self.to_screen(
2181 '[info] Available formats for %s:\n%s' %
2182 (info_dict['id'], render_table(header_line, table)))
2183
2184 def list_thumbnails(self, info_dict):
2185 thumbnails = info_dict.get('thumbnails')
2186 if not thumbnails:
2187 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2188 return
2189
2190 self.to_screen(
2191 '[info] Thumbnails for %s:' % info_dict['id'])
2192 self.to_screen(render_table(
2193 ['ID', 'width', 'height', 'URL'],
2194 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2195
2196 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2197 if not subtitles:
2198 self.to_screen('%s has no %s' % (video_id, name))
2199 return
2200 self.to_screen(
2201 'Available %s for %s:' % (name, video_id))
2202 self.to_screen(render_table(
2203 ['Language', 'formats'],
2204 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2205 for lang, formats in subtitles.items()]))
2206
2207 def urlopen(self, req):
2208 """ Start an HTTP download """
2209 if isinstance(req, compat_basestring):
2210 req = sanitized_Request(req)
2211 return self._opener.open(req, timeout=self._socket_timeout)
2212
2213 def print_debug_header(self):
2214 if not self.params.get('verbose'):
2215 return
2216
2217 if type('') is not compat_str:
2218 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
2219 self.report_warning(
2220 'Your Python is broken! Update to a newer and supported version')
2221
2222 stdout_encoding = getattr(
2223 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2224 encoding_str = (
2225 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2226 locale.getpreferredencoding(),
2227 sys.getfilesystemencoding(),
2228 stdout_encoding,
2229 self.get_encoding()))
2230 write_string(encoding_str, encoding=None)
2231
2232 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2233 if _LAZY_LOADER:
2234 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2235 try:
2236 sp = subprocess.Popen(
2237 ['git', 'rev-parse', '--short', 'HEAD'],
2238 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2239 cwd=os.path.dirname(os.path.abspath(__file__)))
2240 out, err = sp.communicate()
2241 out = out.decode().strip()
2242 if re.match('[0-9a-f]+', out):
2243 self._write_string('[debug] Git HEAD: ' + out + '\n')
2244 except Exception:
2245 try:
2246 sys.exc_clear()
2247 except Exception:
2248 pass
2249
2250 def python_implementation():
2251 impl_name = platform.python_implementation()
2252 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
2253 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
2254 return impl_name
2255
2256 self._write_string('[debug] Python version %s (%s) - %s\n' % (
2257 platform.python_version(), python_implementation(),
2258 platform_name()))
2259
2260 exe_versions = FFmpegPostProcessor.get_versions(self)
2261 exe_versions['rtmpdump'] = rtmpdump_version()
2262 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2263 exe_str = ', '.join(
2264 '%s %s' % (exe, v)
2265 for exe, v in sorted(exe_versions.items())
2266 if v
2267 )
2268 if not exe_str:
2269 exe_str = 'none'
2270 self._write_string('[debug] exe versions: %s\n' % exe_str)
2271
2272 proxy_map = {}
2273 for handler in self._opener.handlers:
2274 if hasattr(handler, 'proxies'):
2275 proxy_map.update(handler.proxies)
2276 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2277
2278 if self.params.get('call_home', False):
2279 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2280 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2281 latest_version = self.urlopen(
2282 'https://yt-dl.org/latest/version').read().decode('utf-8')
2283 if version_tuple(latest_version) > version_tuple(__version__):
2284 self.report_warning(
2285 'You are using an outdated version (newest version: %s)! '
2286 'See https://yt-dl.org/update if you need help updating.' %
2287 latest_version)
2288
2289 def _setup_opener(self):
2290 timeout_val = self.params.get('socket_timeout')
2291 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2292
2293 opts_cookiefile = self.params.get('cookiefile')
2294 opts_proxy = self.params.get('proxy')
2295
2296 if opts_cookiefile is None:
2297 self.cookiejar = compat_cookiejar.CookieJar()
2298 else:
2299 opts_cookiefile = expand_path(opts_cookiefile)
2300 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2301 opts_cookiefile)
2302 if os.access(opts_cookiefile, os.R_OK):
2303 self.cookiejar.load()
2304
2305 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2306 if opts_proxy is not None:
2307 if opts_proxy == '':
2308 proxies = {}
2309 else:
2310 proxies = {'http': opts_proxy, 'https': opts_proxy}
2311 else:
2312 proxies = compat_urllib_request.getproxies()
2313 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2314 if 'http' in proxies and 'https' not in proxies:
2315 proxies['https'] = proxies['http']
2316 proxy_handler = PerRequestProxyHandler(proxies)
2317
2318 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2319 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2320 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2321 data_handler = compat_urllib_request_DataHandler()
2322
2323 # When passing our own FileHandler instance, build_opener won't add the
2324 # default FileHandler and allows us to disable the file protocol, which
2325 # can be used for malicious purposes (see
2326 # https://github.com/rg3/youtube-dl/issues/8227)
2327 file_handler = compat_urllib_request.FileHandler()
2328
2329 def file_open(*args, **kwargs):
2330 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2331 file_handler.file_open = file_open
2332
2333 opener = compat_urllib_request.build_opener(
2334 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2335
2336 # Delete the default user-agent header, which would otherwise apply in
2337 # cases where our custom HTTP handler doesn't come into play
2338 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2339 opener.addheaders = []
2340 self._opener = opener
2341
2342 def encode(self, s):
2343 if isinstance(s, bytes):
2344 return s # Already encoded
2345
2346 try:
2347 return s.encode(self.get_encoding())
2348 except UnicodeEncodeError as err:
2349 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2350 raise
2351
2352 def get_encoding(self):
2353 encoding = self.params.get('encoding')
2354 if encoding is None:
2355 encoding = preferredencoding()
2356 return encoding
2357
2358 def _write_thumbnails(self, info_dict, filename):
2359 if self.params.get('writethumbnail', False):
2360 thumbnails = info_dict.get('thumbnails')
2361 if thumbnails:
2362 thumbnails = [thumbnails[-1]]
2363 elif self.params.get('write_all_thumbnails', False):
2364 thumbnails = info_dict.get('thumbnails')
2365 else:
2366 return
2367
2368 if not thumbnails:
2369 # No thumbnails present, so return immediately
2370 return
2371
2372 for t in thumbnails:
2373 thumb_ext = determine_ext(t['url'], 'jpg')
2374 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2375 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2376 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2377
2378 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2379 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2380 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2381 else:
2382 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2383 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2384 try:
2385 uf = self.urlopen(t['url'])
2386 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2387 shutil.copyfileobj(uf, thumbf)
2388 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2389 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2390 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2391 self.report_warning('Unable to download thumbnail "%s": %s' %
2392 (t['url'], error_to_compat_str(err)))