]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/YoutubeDL.py
e844dc98a5b3915070ffae079395233de7ed04f7
[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
28 from .compat import (
29 compat_basestring,
30 compat_cookiejar,
31 compat_expanduser,
32 compat_get_terminal_size,
33 compat_http_client,
34 compat_kwargs,
35 compat_os_name,
36 compat_str,
37 compat_tokenize_tokenize,
38 compat_urllib_error,
39 compat_urllib_request,
40 compat_urllib_request_DataHandler,
41 )
42 from .utils import (
43 age_restricted,
44 args_to_str,
45 ContentTooShortError,
46 date_from_str,
47 DateRange,
48 DEFAULT_OUTTMPL,
49 determine_ext,
50 determine_protocol,
51 DownloadError,
52 encode_compat_str,
53 encodeFilename,
54 error_to_compat_str,
55 ExtractorError,
56 format_bytes,
57 formatSeconds,
58 locked_file,
59 make_HTTPS_handler,
60 MaxDownloadsReached,
61 PagedList,
62 parse_filesize,
63 PerRequestProxyHandler,
64 platform_name,
65 PostProcessingError,
66 preferredencoding,
67 prepend_extension,
68 register_socks_protocols,
69 render_table,
70 replace_extension,
71 SameFileError,
72 sanitize_filename,
73 sanitize_path,
74 sanitize_url,
75 sanitized_Request,
76 std_headers,
77 subtitles_filename,
78 UnavailableVideoError,
79 url_basename,
80 version_tuple,
81 write_json_file,
82 write_string,
83 YoutubeDLCookieProcessor,
84 YoutubeDLHandler,
85 )
86 from .cache import Cache
87 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
88 from .downloader import get_suitable_downloader
89 from .downloader.rtmp import rtmpdump_version
90 from .postprocessor import (
91 FFmpegFixupM3u8PP,
92 FFmpegFixupM4aPP,
93 FFmpegFixupStretchedPP,
94 FFmpegMergerPP,
95 FFmpegPostProcessor,
96 get_postprocessor,
97 )
98 from .version import __version__
99
100 if compat_os_name == 'nt':
101 import ctypes
102
103
104 class YoutubeDL(object):
105 """YoutubeDL class.
106
107 YoutubeDL objects are the ones responsible of downloading the
108 actual video file and writing it to disk if the user has requested
109 it, among some other tasks. In most cases there should be one per
110 program. As, given a video URL, the downloader doesn't know how to
111 extract all the needed information, task that InfoExtractors do, it
112 has to pass the URL to one of them.
113
114 For this, YoutubeDL objects have a method that allows
115 InfoExtractors to be registered in a given order. When it is passed
116 a URL, the YoutubeDL object handles it to the first InfoExtractor it
117 finds that reports being able to handle it. The InfoExtractor extracts
118 all the information about the video or videos the URL refers to, and
119 YoutubeDL process the extracted information, possibly using a File
120 Downloader to download the video.
121
122 YoutubeDL objects accept a lot of parameters. In order not to saturate
123 the object constructor with arguments, it receives a dictionary of
124 options instead. These options are available through the params
125 attribute for the InfoExtractors to use. The YoutubeDL also
126 registers itself as the downloader in charge for the InfoExtractors
127 that are added to it, so this is a "mutual registration".
128
129 Available options:
130
131 username: Username for authentication purposes.
132 password: Password for authentication purposes.
133 videopassword: Password for accessing a video.
134 usenetrc: Use netrc for authentication instead.
135 verbose: Print additional info to stdout.
136 quiet: Do not print messages to stdout.
137 no_warnings: Do not print out anything for warnings.
138 forceurl: Force printing final URL.
139 forcetitle: Force printing title.
140 forceid: Force printing ID.
141 forcethumbnail: Force printing thumbnail URL.
142 forcedescription: Force printing description.
143 forcefilename: Force printing final filename.
144 forceduration: Force printing duration.
145 forcejson: Force printing info_dict as JSON.
146 dump_single_json: Force printing the info_dict of the whole playlist
147 (or video) as a single JSON line.
148 simulate: Do not download the video files.
149 format: Video format code. See options.py for more information.
150 outtmpl: Template for output names.
151 restrictfilenames: Do not allow "&" and spaces in file names
152 ignoreerrors: Do not stop on download errors.
153 force_generic_extractor: Force downloader to use the generic extractor
154 nooverwrites: Prevent overwriting files.
155 playliststart: Playlist item to start at.
156 playlistend: Playlist item to end at.
157 playlist_items: Specific indices of playlist to download.
158 playlistreverse: Download playlist items in reverse order.
159 matchtitle: Download only matching titles.
160 rejecttitle: Reject downloads for matching titles.
161 logger: Log messages to a logging.Logger instance.
162 logtostderr: Log messages to stderr instead of stdout.
163 writedescription: Write the video description to a .description file
164 writeinfojson: Write the video description to a .info.json file
165 writeannotations: Write the video annotations to a .annotations.xml file
166 writethumbnail: Write the thumbnail image to a file
167 write_all_thumbnails: Write all thumbnail formats to files
168 writesubtitles: Write the video subtitles to a file
169 writeautomaticsub: Write the automatically generated subtitles to a file
170 allsubtitles: Downloads all the subtitles of the video
171 (requires writesubtitles or writeautomaticsub)
172 listsubtitles: Lists all available subtitles for the video
173 subtitlesformat: The format code for subtitles
174 subtitleslangs: List of languages of the subtitles to download
175 keepvideo: Keep the video file after post-processing
176 daterange: A DateRange object, download only if the upload_date is in the range.
177 skip_download: Skip the actual download of the video file
178 cachedir: Location of the cache files in the filesystem.
179 False to disable filesystem cache.
180 noplaylist: Download single video instead of a playlist if in doubt.
181 age_limit: An integer representing the user's age in years.
182 Unsuitable videos for the given age are skipped.
183 min_views: An integer representing the minimum view count the video
184 must have in order to not be skipped.
185 Videos without view count information are always
186 downloaded. None for no limit.
187 max_views: An integer representing the maximum view count.
188 Videos that are more popular than that are not
189 downloaded.
190 Videos without view count information are always
191 downloaded. None for no limit.
192 download_archive: File name of a file where all downloads are recorded.
193 Videos already present in the file are not downloaded
194 again.
195 cookiefile: File name where cookies should be read from and dumped to.
196 nocheckcertificate:Do not verify SSL certificates
197 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
198 At the moment, this is only supported by YouTube.
199 proxy: URL of the proxy server to use
200 geo_verification_proxy: URL of the proxy to use for IP address verification
201 on geo-restricted sites. (Experimental)
202 socket_timeout: Time to wait for unresponsive hosts, in seconds
203 bidi_workaround: Work around buggy terminals without bidirectional text
204 support, using fridibi
205 debug_printtraffic:Print out sent and received HTTP traffic
206 include_ads: Download ads as well
207 default_search: Prepend this string if an input url is not valid.
208 'auto' for elaborate guessing
209 encoding: Use this encoding instead of the system-specified.
210 extract_flat: Do not resolve URLs, return the immediate result.
211 Pass in 'in_playlist' to only show this behavior for
212 playlist items.
213 postprocessors: A list of dictionaries, each with an entry
214 * key: The name of the postprocessor. See
215 youtube_dl/postprocessor/__init__.py for a list.
216 as well as any further keyword arguments for the
217 postprocessor.
218 progress_hooks: A list of functions that get called on download
219 progress, with a dictionary with the entries
220 * status: One of "downloading", "error", or "finished".
221 Check this first and ignore unknown values.
222
223 If status is one of "downloading", or "finished", the
224 following properties may also be present:
225 * filename: The final filename (always present)
226 * tmpfilename: The filename we're currently writing to
227 * downloaded_bytes: Bytes on disk
228 * total_bytes: Size of the whole file, None if unknown
229 * total_bytes_estimate: Guess of the eventual file size,
230 None if unavailable.
231 * elapsed: The number of seconds since download started.
232 * eta: The estimated time in seconds, None if unknown
233 * speed: The download speed in bytes/second, None if
234 unknown
235 * fragment_index: The counter of the currently
236 downloaded video fragment.
237 * fragment_count: The number of fragments (= individual
238 files that will be merged)
239
240 Progress hooks are guaranteed to be called at least once
241 (with status "finished") if the download is successful.
242 merge_output_format: Extension to use when merging formats.
243 fixup: Automatically correct known faults of the file.
244 One of:
245 - "never": do nothing
246 - "warn": only emit a warning
247 - "detect_or_warn": check whether we can do anything
248 about it, warn otherwise (default)
249 source_address: (Experimental) Client-side IP address to bind to.
250 call_home: Boolean, true iff we are allowed to contact the
251 youtube-dl servers for debugging.
252 sleep_interval: Number of seconds to sleep before each download when
253 used alone or a lower bound of a range for randomized
254 sleep before each download (minimum possible number
255 of seconds to sleep) when used along with
256 max_sleep_interval.
257 max_sleep_interval:Upper bound of a range for randomized sleep before each
258 download (maximum possible number of seconds to sleep).
259 Must only be used along with sleep_interval.
260 Actual sleep time will be a random float from range
261 [sleep_interval; max_sleep_interval].
262 listformats: Print an overview of available video formats and exit.
263 list_thumbnails: Print a table of all thumbnails and exit.
264 match_filter: A function that gets called with the info_dict of
265 every video.
266 If it returns a message, the video is ignored.
267 If it returns None, the video is downloaded.
268 match_filter_func in utils.py is one example for this.
269 no_color: Do not emit color codes in output.
270
271 The following options determine which downloader is picked:
272 external_downloader: Executable of the external downloader to call.
273 None or unset for standard (built-in) downloader.
274 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
275 if True, otherwise use ffmpeg/avconv if False, otherwise
276 use downloader suggested by extractor if None.
277
278 The following parameters are not used by YoutubeDL itself, they are used by
279 the downloader (see youtube_dl/downloader/common.py):
280 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
281 noresizebuffer, retries, continuedl, noprogress, consoletitle,
282 xattr_set_filesize, external_downloader_args, hls_use_mpegts.
283
284 The following options are used by the post processors:
285 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
286 otherwise prefer avconv.
287 postprocessor_args: A list of additional command-line arguments for the
288 postprocessor.
289 """
290
291 params = None
292 _ies = []
293 _pps = []
294 _download_retcode = None
295 _num_downloads = None
296 _screen_file = None
297
298 def __init__(self, params=None, auto_init=True):
299 """Create a FileDownloader object with the given options."""
300 if params is None:
301 params = {}
302 self._ies = []
303 self._ies_instances = {}
304 self._pps = []
305 self._progress_hooks = []
306 self._download_retcode = 0
307 self._num_downloads = 0
308 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
309 self._err_file = sys.stderr
310 self.params = {
311 # Default parameters
312 'nocheckcertificate': False,
313 }
314 self.params.update(params)
315 self.cache = Cache(self)
316
317 if self.params.get('cn_verification_proxy') is not None:
318 self.report_warning('--cn-verification-proxy is deprecated. Use --geo-verification-proxy instead.')
319 if self.params.get('geo_verification_proxy') is None:
320 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
321
322 if params.get('bidi_workaround', False):
323 try:
324 import pty
325 master, slave = pty.openpty()
326 width = compat_get_terminal_size().columns
327 if width is None:
328 width_args = []
329 else:
330 width_args = ['-w', str(width)]
331 sp_kwargs = dict(
332 stdin=subprocess.PIPE,
333 stdout=slave,
334 stderr=self._err_file)
335 try:
336 self._output_process = subprocess.Popen(
337 ['bidiv'] + width_args, **sp_kwargs
338 )
339 except OSError:
340 self._output_process = subprocess.Popen(
341 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
342 self._output_channel = os.fdopen(master, 'rb')
343 except OSError as ose:
344 if ose.errno == errno.ENOENT:
345 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 $PATH.')
346 else:
347 raise
348
349 if (sys.version_info >= (3,) and sys.platform != 'win32' and
350 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
351 not params.get('restrictfilenames', False)):
352 # On Python 3, the Unicode filesystem API will throw errors (#1474)
353 self.report_warning(
354 'Assuming --restrict-filenames since file system encoding '
355 'cannot encode all characters. '
356 'Set the LC_ALL environment variable to fix this.')
357 self.params['restrictfilenames'] = True
358
359 if isinstance(params.get('outtmpl'), bytes):
360 self.report_warning(
361 'Parameter outtmpl is bytes, but should be a unicode string. '
362 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
363
364 self._setup_opener()
365
366 if auto_init:
367 self.print_debug_header()
368 self.add_default_info_extractors()
369
370 for pp_def_raw in self.params.get('postprocessors', []):
371 pp_class = get_postprocessor(pp_def_raw['key'])
372 pp_def = dict(pp_def_raw)
373 del pp_def['key']
374 pp = pp_class(self, **compat_kwargs(pp_def))
375 self.add_post_processor(pp)
376
377 for ph in self.params.get('progress_hooks', []):
378 self.add_progress_hook(ph)
379
380 register_socks_protocols()
381
382 def warn_if_short_id(self, argv):
383 # short YouTube ID starting with dash?
384 idxs = [
385 i for i, a in enumerate(argv)
386 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
387 if idxs:
388 correct_argv = (
389 ['youtube-dl'] +
390 [a for i, a in enumerate(argv) if i not in idxs] +
391 ['--'] + [argv[i] for i in idxs]
392 )
393 self.report_warning(
394 'Long argument string detected. '
395 'Use -- to separate parameters and URLs, like this:\n%s\n' %
396 args_to_str(correct_argv))
397
398 def add_info_extractor(self, ie):
399 """Add an InfoExtractor object to the end of the list."""
400 self._ies.append(ie)
401 if not isinstance(ie, type):
402 self._ies_instances[ie.ie_key()] = ie
403 ie.set_downloader(self)
404
405 def get_info_extractor(self, ie_key):
406 """
407 Get an instance of an IE with name ie_key, it will try to get one from
408 the _ies list, if there's no instance it will create a new one and add
409 it to the extractor list.
410 """
411 ie = self._ies_instances.get(ie_key)
412 if ie is None:
413 ie = get_info_extractor(ie_key)()
414 self.add_info_extractor(ie)
415 return ie
416
417 def add_default_info_extractors(self):
418 """
419 Add the InfoExtractors returned by gen_extractors to the end of the list
420 """
421 for ie in gen_extractor_classes():
422 self.add_info_extractor(ie)
423
424 def add_post_processor(self, pp):
425 """Add a PostProcessor object to the end of the chain."""
426 self._pps.append(pp)
427 pp.set_downloader(self)
428
429 def add_progress_hook(self, ph):
430 """Add the progress hook (currently only for the file downloader)"""
431 self._progress_hooks.append(ph)
432
433 def _bidi_workaround(self, message):
434 if not hasattr(self, '_output_channel'):
435 return message
436
437 assert hasattr(self, '_output_process')
438 assert isinstance(message, compat_str)
439 line_count = message.count('\n') + 1
440 self._output_process.stdin.write((message + '\n').encode('utf-8'))
441 self._output_process.stdin.flush()
442 res = ''.join(self._output_channel.readline().decode('utf-8')
443 for _ in range(line_count))
444 return res[:-len('\n')]
445
446 def to_screen(self, message, skip_eol=False):
447 """Print message to stdout if not in quiet mode."""
448 return self.to_stdout(message, skip_eol, check_quiet=True)
449
450 def _write_string(self, s, out=None):
451 write_string(s, out=out, encoding=self.params.get('encoding'))
452
453 def to_stdout(self, message, skip_eol=False, check_quiet=False):
454 """Print message to stdout if not in quiet mode."""
455 if self.params.get('logger'):
456 self.params['logger'].debug(message)
457 elif not check_quiet or not self.params.get('quiet', False):
458 message = self._bidi_workaround(message)
459 terminator = ['\n', ''][skip_eol]
460 output = message + terminator
461
462 self._write_string(output, self._screen_file)
463
464 def to_stderr(self, message):
465 """Print message to stderr."""
466 assert isinstance(message, compat_str)
467 if self.params.get('logger'):
468 self.params['logger'].error(message)
469 else:
470 message = self._bidi_workaround(message)
471 output = message + '\n'
472 self._write_string(output, self._err_file)
473
474 def to_console_title(self, message):
475 if not self.params.get('consoletitle', False):
476 return
477 if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
478 # c_wchar_p() might not be necessary if `message` is
479 # already of type unicode()
480 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
481 elif 'TERM' in os.environ:
482 self._write_string('\033]0;%s\007' % message, self._screen_file)
483
484 def save_console_title(self):
485 if not self.params.get('consoletitle', False):
486 return
487 if 'TERM' in os.environ:
488 # Save the title on stack
489 self._write_string('\033[22;0t', self._screen_file)
490
491 def restore_console_title(self):
492 if not self.params.get('consoletitle', False):
493 return
494 if 'TERM' in os.environ:
495 # Restore the title from stack
496 self._write_string('\033[23;0t', self._screen_file)
497
498 def __enter__(self):
499 self.save_console_title()
500 return self
501
502 def __exit__(self, *args):
503 self.restore_console_title()
504
505 if self.params.get('cookiefile') is not None:
506 self.cookiejar.save()
507
508 def trouble(self, message=None, tb=None):
509 """Determine action to take when a download problem appears.
510
511 Depending on if the downloader has been configured to ignore
512 download errors or not, this method may throw an exception or
513 not when errors are found, after printing the message.
514
515 tb, if given, is additional traceback information.
516 """
517 if message is not None:
518 self.to_stderr(message)
519 if self.params.get('verbose'):
520 if tb is None:
521 if sys.exc_info()[0]: # if .trouble has been called from an except block
522 tb = ''
523 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
524 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
525 tb += encode_compat_str(traceback.format_exc())
526 else:
527 tb_data = traceback.format_list(traceback.extract_stack())
528 tb = ''.join(tb_data)
529 self.to_stderr(tb)
530 if not self.params.get('ignoreerrors', False):
531 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
532 exc_info = sys.exc_info()[1].exc_info
533 else:
534 exc_info = sys.exc_info()
535 raise DownloadError(message, exc_info)
536 self._download_retcode = 1
537
538 def report_warning(self, message):
539 '''
540 Print the message to stderr, it will be prefixed with 'WARNING:'
541 If stderr is a tty file the 'WARNING:' will be colored
542 '''
543 if self.params.get('logger') is not None:
544 self.params['logger'].warning(message)
545 else:
546 if self.params.get('no_warnings'):
547 return
548 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
549 _msg_header = '\033[0;33mWARNING:\033[0m'
550 else:
551 _msg_header = 'WARNING:'
552 warning_message = '%s %s' % (_msg_header, message)
553 self.to_stderr(warning_message)
554
555 def report_error(self, message, tb=None):
556 '''
557 Do the same as trouble, but prefixes the message with 'ERROR:', colored
558 in red if stderr is a tty file.
559 '''
560 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
561 _msg_header = '\033[0;31mERROR:\033[0m'
562 else:
563 _msg_header = 'ERROR:'
564 error_message = '%s %s' % (_msg_header, message)
565 self.trouble(error_message, tb)
566
567 def report_file_already_downloaded(self, file_name):
568 """Report file has already been fully downloaded."""
569 try:
570 self.to_screen('[download] %s has already been downloaded' % file_name)
571 except UnicodeEncodeError:
572 self.to_screen('[download] The file has already been downloaded')
573
574 def prepare_filename(self, info_dict):
575 """Generate the output filename."""
576 try:
577 template_dict = dict(info_dict)
578
579 template_dict['epoch'] = int(time.time())
580 autonumber_size = self.params.get('autonumber_size')
581 if autonumber_size is None:
582 autonumber_size = 5
583 autonumber_templ = '%0' + str(autonumber_size) + 'd'
584 template_dict['autonumber'] = autonumber_templ % self._num_downloads
585 if template_dict.get('playlist_index') is not None:
586 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
587 if template_dict.get('resolution') is None:
588 if template_dict.get('width') and template_dict.get('height'):
589 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
590 elif template_dict.get('height'):
591 template_dict['resolution'] = '%sp' % template_dict['height']
592 elif template_dict.get('width'):
593 template_dict['resolution'] = '%dx?' % template_dict['width']
594
595 sanitize = lambda k, v: sanitize_filename(
596 compat_str(v),
597 restricted=self.params.get('restrictfilenames'),
598 is_id=(k == 'id'))
599 template_dict = dict((k, sanitize(k, v))
600 for k, v in template_dict.items()
601 if v is not None and not isinstance(v, (list, tuple, dict)))
602 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
603
604 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
605 tmpl = compat_expanduser(outtmpl)
606 filename = tmpl % template_dict
607 # Temporary fix for #4787
608 # 'Treat' all problem characters by passing filename through preferredencoding
609 # to workaround encoding issues with subprocess on python2 @ Windows
610 if sys.version_info < (3, 0) and sys.platform == 'win32':
611 filename = encodeFilename(filename, True).decode(preferredencoding())
612 return sanitize_path(filename)
613 except ValueError as err:
614 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
615 return None
616
617 def _match_entry(self, info_dict, incomplete):
618 """ Returns None iff the file should be downloaded """
619
620 video_title = info_dict.get('title', info_dict.get('id', 'video'))
621 if 'title' in info_dict:
622 # This can happen when we're just evaluating the playlist
623 title = info_dict['title']
624 matchtitle = self.params.get('matchtitle', False)
625 if matchtitle:
626 if not re.search(matchtitle, title, re.IGNORECASE):
627 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
628 rejecttitle = self.params.get('rejecttitle', False)
629 if rejecttitle:
630 if re.search(rejecttitle, title, re.IGNORECASE):
631 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
632 date = info_dict.get('upload_date')
633 if date is not None:
634 dateRange = self.params.get('daterange', DateRange())
635 if date not in dateRange:
636 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
637 view_count = info_dict.get('view_count')
638 if view_count is not None:
639 min_views = self.params.get('min_views')
640 if min_views is not None and view_count < min_views:
641 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
642 max_views = self.params.get('max_views')
643 if max_views is not None and view_count > max_views:
644 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
645 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
646 return 'Skipping "%s" because it is age restricted' % video_title
647 if self.in_download_archive(info_dict):
648 return '%s has already been recorded in archive' % video_title
649
650 if not incomplete:
651 match_filter = self.params.get('match_filter')
652 if match_filter is not None:
653 ret = match_filter(info_dict)
654 if ret is not None:
655 return ret
656
657 return None
658
659 @staticmethod
660 def add_extra_info(info_dict, extra_info):
661 '''Set the keys from extra_info in info dict if they are missing'''
662 for key, value in extra_info.items():
663 info_dict.setdefault(key, value)
664
665 def extract_info(self, url, download=True, ie_key=None, extra_info={},
666 process=True, force_generic_extractor=False):
667 '''
668 Returns a list with a dictionary for each video we find.
669 If 'download', also downloads the videos.
670 extra_info is a dict containing the extra values to add to each result
671 '''
672
673 if not ie_key and force_generic_extractor:
674 ie_key = 'Generic'
675
676 if ie_key:
677 ies = [self.get_info_extractor(ie_key)]
678 else:
679 ies = self._ies
680
681 for ie in ies:
682 if not ie.suitable(url):
683 continue
684
685 ie = self.get_info_extractor(ie.ie_key())
686 if not ie.working():
687 self.report_warning('The program functionality for this site has been marked as broken, '
688 'and will probably not work.')
689
690 try:
691 ie_result = ie.extract(url)
692 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
693 break
694 if isinstance(ie_result, list):
695 # Backwards compatibility: old IE result format
696 ie_result = {
697 '_type': 'compat_list',
698 'entries': ie_result,
699 }
700 self.add_default_extra_info(ie_result, ie, url)
701 if process:
702 return self.process_ie_result(ie_result, download, extra_info)
703 else:
704 return ie_result
705 except ExtractorError as e: # An error we somewhat expected
706 self.report_error(compat_str(e), e.format_traceback())
707 break
708 except MaxDownloadsReached:
709 raise
710 except Exception as e:
711 if self.params.get('ignoreerrors', False):
712 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
713 break
714 else:
715 raise
716 else:
717 self.report_error('no suitable InfoExtractor for URL %s' % url)
718
719 def add_default_extra_info(self, ie_result, ie, url):
720 self.add_extra_info(ie_result, {
721 'extractor': ie.IE_NAME,
722 'webpage_url': url,
723 'webpage_url_basename': url_basename(url),
724 'extractor_key': ie.ie_key(),
725 })
726
727 def process_ie_result(self, ie_result, download=True, extra_info={}):
728 """
729 Take the result of the ie(may be modified) and resolve all unresolved
730 references (URLs, playlist items).
731
732 It will also download the videos if 'download'.
733 Returns the resolved ie_result.
734 """
735 result_type = ie_result.get('_type', 'video')
736
737 if result_type in ('url', 'url_transparent'):
738 ie_result['url'] = sanitize_url(ie_result['url'])
739 extract_flat = self.params.get('extract_flat', False)
740 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
741 extract_flat is True):
742 if self.params.get('forcejson', False):
743 self.to_stdout(json.dumps(ie_result))
744 return ie_result
745
746 if result_type == 'video':
747 self.add_extra_info(ie_result, extra_info)
748 return self.process_video_result(ie_result, download=download)
749 elif result_type == 'url':
750 # We have to add extra_info to the results because it may be
751 # contained in a playlist
752 return self.extract_info(ie_result['url'],
753 download,
754 ie_key=ie_result.get('ie_key'),
755 extra_info=extra_info)
756 elif result_type == 'url_transparent':
757 # Use the information from the embedding page
758 info = self.extract_info(
759 ie_result['url'], ie_key=ie_result.get('ie_key'),
760 extra_info=extra_info, download=False, process=False)
761
762 force_properties = dict(
763 (k, v) for k, v in ie_result.items() if v is not None)
764 for f in ('_type', 'url', 'ie_key'):
765 if f in force_properties:
766 del force_properties[f]
767 new_result = info.copy()
768 new_result.update(force_properties)
769
770 assert new_result.get('_type') != 'url_transparent'
771
772 return self.process_ie_result(
773 new_result, download=download, extra_info=extra_info)
774 elif result_type == 'playlist' or result_type == 'multi_video':
775 # We process each entry in the playlist
776 playlist = ie_result.get('title') or ie_result.get('id')
777 self.to_screen('[download] Downloading playlist: %s' % playlist)
778
779 playlist_results = []
780
781 playliststart = self.params.get('playliststart', 1) - 1
782 playlistend = self.params.get('playlistend')
783 # For backwards compatibility, interpret -1 as whole list
784 if playlistend == -1:
785 playlistend = None
786
787 playlistitems_str = self.params.get('playlist_items')
788 playlistitems = None
789 if playlistitems_str is not None:
790 def iter_playlistitems(format):
791 for string_segment in format.split(','):
792 if '-' in string_segment:
793 start, end = string_segment.split('-')
794 for item in range(int(start), int(end) + 1):
795 yield int(item)
796 else:
797 yield int(string_segment)
798 playlistitems = iter_playlistitems(playlistitems_str)
799
800 ie_entries = ie_result['entries']
801 if isinstance(ie_entries, list):
802 n_all_entries = len(ie_entries)
803 if playlistitems:
804 entries = [
805 ie_entries[i - 1] for i in playlistitems
806 if -n_all_entries <= i - 1 < n_all_entries]
807 else:
808 entries = ie_entries[playliststart:playlistend]
809 n_entries = len(entries)
810 self.to_screen(
811 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
812 (ie_result['extractor'], playlist, n_all_entries, n_entries))
813 elif isinstance(ie_entries, PagedList):
814 if playlistitems:
815 entries = []
816 for item in playlistitems:
817 entries.extend(ie_entries.getslice(
818 item - 1, item
819 ))
820 else:
821 entries = ie_entries.getslice(
822 playliststart, playlistend)
823 n_entries = len(entries)
824 self.to_screen(
825 '[%s] playlist %s: Downloading %d videos' %
826 (ie_result['extractor'], playlist, n_entries))
827 else: # iterable
828 if playlistitems:
829 entry_list = list(ie_entries)
830 entries = [entry_list[i - 1] for i in playlistitems]
831 else:
832 entries = list(itertools.islice(
833 ie_entries, playliststart, playlistend))
834 n_entries = len(entries)
835 self.to_screen(
836 '[%s] playlist %s: Downloading %d videos' %
837 (ie_result['extractor'], playlist, n_entries))
838
839 if self.params.get('playlistreverse', False):
840 entries = entries[::-1]
841
842 for i, entry in enumerate(entries, 1):
843 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
844 extra = {
845 'n_entries': n_entries,
846 'playlist': playlist,
847 'playlist_id': ie_result.get('id'),
848 'playlist_title': ie_result.get('title'),
849 'playlist_index': i + playliststart,
850 'extractor': ie_result['extractor'],
851 'webpage_url': ie_result['webpage_url'],
852 'webpage_url_basename': url_basename(ie_result['webpage_url']),
853 'extractor_key': ie_result['extractor_key'],
854 }
855
856 reason = self._match_entry(entry, incomplete=True)
857 if reason is not None:
858 self.to_screen('[download] ' + reason)
859 continue
860
861 entry_result = self.process_ie_result(entry,
862 download=download,
863 extra_info=extra)
864 playlist_results.append(entry_result)
865 ie_result['entries'] = playlist_results
866 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
867 return ie_result
868 elif result_type == 'compat_list':
869 self.report_warning(
870 'Extractor %s returned a compat_list result. '
871 'It needs to be updated.' % ie_result.get('extractor'))
872
873 def _fixup(r):
874 self.add_extra_info(
875 r,
876 {
877 'extractor': ie_result['extractor'],
878 'webpage_url': ie_result['webpage_url'],
879 'webpage_url_basename': url_basename(ie_result['webpage_url']),
880 'extractor_key': ie_result['extractor_key'],
881 }
882 )
883 return r
884 ie_result['entries'] = [
885 self.process_ie_result(_fixup(r), download, extra_info)
886 for r in ie_result['entries']
887 ]
888 return ie_result
889 else:
890 raise Exception('Invalid result type: %s' % result_type)
891
892 def _build_format_filter(self, filter_spec):
893 " Returns a function to filter the formats according to the filter_spec "
894
895 OPERATORS = {
896 '<': operator.lt,
897 '<=': operator.le,
898 '>': operator.gt,
899 '>=': operator.ge,
900 '=': operator.eq,
901 '!=': operator.ne,
902 }
903 operator_rex = re.compile(r'''(?x)\s*
904 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
905 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
906 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
907 $
908 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
909 m = operator_rex.search(filter_spec)
910 if m:
911 try:
912 comparison_value = int(m.group('value'))
913 except ValueError:
914 comparison_value = parse_filesize(m.group('value'))
915 if comparison_value is None:
916 comparison_value = parse_filesize(m.group('value') + 'B')
917 if comparison_value is None:
918 raise ValueError(
919 'Invalid value %r in format specification %r' % (
920 m.group('value'), filter_spec))
921 op = OPERATORS[m.group('op')]
922
923 if not m:
924 STR_OPERATORS = {
925 '=': operator.eq,
926 '!=': operator.ne,
927 '^=': lambda attr, value: attr.startswith(value),
928 '$=': lambda attr, value: attr.endswith(value),
929 '*=': lambda attr, value: value in attr,
930 }
931 str_operator_rex = re.compile(r'''(?x)
932 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
933 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
934 \s*(?P<value>[a-zA-Z0-9._-]+)
935 \s*$
936 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
937 m = str_operator_rex.search(filter_spec)
938 if m:
939 comparison_value = m.group('value')
940 op = STR_OPERATORS[m.group('op')]
941
942 if not m:
943 raise ValueError('Invalid filter specification %r' % filter_spec)
944
945 def _filter(f):
946 actual_value = f.get(m.group('key'))
947 if actual_value is None:
948 return m.group('none_inclusive')
949 return op(actual_value, comparison_value)
950 return _filter
951
952 def build_format_selector(self, format_spec):
953 def syntax_error(note, start):
954 message = (
955 'Invalid format specification: '
956 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
957 return SyntaxError(message)
958
959 PICKFIRST = 'PICKFIRST'
960 MERGE = 'MERGE'
961 SINGLE = 'SINGLE'
962 GROUP = 'GROUP'
963 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
964
965 def _parse_filter(tokens):
966 filter_parts = []
967 for type, string, start, _, _ in tokens:
968 if type == tokenize.OP and string == ']':
969 return ''.join(filter_parts)
970 else:
971 filter_parts.append(string)
972
973 def _remove_unused_ops(tokens):
974 # Remove operators that we don't use and join them with the surrounding strings
975 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
976 ALLOWED_OPS = ('/', '+', ',', '(', ')')
977 last_string, last_start, last_end, last_line = None, None, None, None
978 for type, string, start, end, line in tokens:
979 if type == tokenize.OP and string == '[':
980 if last_string:
981 yield tokenize.NAME, last_string, last_start, last_end, last_line
982 last_string = None
983 yield type, string, start, end, line
984 # everything inside brackets will be handled by _parse_filter
985 for type, string, start, end, line in tokens:
986 yield type, string, start, end, line
987 if type == tokenize.OP and string == ']':
988 break
989 elif type == tokenize.OP and string in ALLOWED_OPS:
990 if last_string:
991 yield tokenize.NAME, last_string, last_start, last_end, last_line
992 last_string = None
993 yield type, string, start, end, line
994 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
995 if not last_string:
996 last_string = string
997 last_start = start
998 last_end = end
999 else:
1000 last_string += string
1001 if last_string:
1002 yield tokenize.NAME, last_string, last_start, last_end, last_line
1003
1004 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1005 selectors = []
1006 current_selector = None
1007 for type, string, start, _, _ in tokens:
1008 # ENCODING is only defined in python 3.x
1009 if type == getattr(tokenize, 'ENCODING', None):
1010 continue
1011 elif type in [tokenize.NAME, tokenize.NUMBER]:
1012 current_selector = FormatSelector(SINGLE, string, [])
1013 elif type == tokenize.OP:
1014 if string == ')':
1015 if not inside_group:
1016 # ')' will be handled by the parentheses group
1017 tokens.restore_last_token()
1018 break
1019 elif inside_merge and string in ['/', ',']:
1020 tokens.restore_last_token()
1021 break
1022 elif inside_choice and string == ',':
1023 tokens.restore_last_token()
1024 break
1025 elif string == ',':
1026 if not current_selector:
1027 raise syntax_error('"," must follow a format selector', start)
1028 selectors.append(current_selector)
1029 current_selector = None
1030 elif string == '/':
1031 if not current_selector:
1032 raise syntax_error('"/" must follow a format selector', start)
1033 first_choice = current_selector
1034 second_choice = _parse_format_selection(tokens, inside_choice=True)
1035 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1036 elif string == '[':
1037 if not current_selector:
1038 current_selector = FormatSelector(SINGLE, 'best', [])
1039 format_filter = _parse_filter(tokens)
1040 current_selector.filters.append(format_filter)
1041 elif string == '(':
1042 if current_selector:
1043 raise syntax_error('Unexpected "("', start)
1044 group = _parse_format_selection(tokens, inside_group=True)
1045 current_selector = FormatSelector(GROUP, group, [])
1046 elif string == '+':
1047 video_selector = current_selector
1048 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1049 if not video_selector or not audio_selector:
1050 raise syntax_error('"+" must be between two format selectors', start)
1051 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1052 else:
1053 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1054 elif type == tokenize.ENDMARKER:
1055 break
1056 if current_selector:
1057 selectors.append(current_selector)
1058 return selectors
1059
1060 def _build_selector_function(selector):
1061 if isinstance(selector, list):
1062 fs = [_build_selector_function(s) for s in selector]
1063
1064 def selector_function(ctx):
1065 for f in fs:
1066 for format in f(ctx):
1067 yield format
1068 return selector_function
1069 elif selector.type == GROUP:
1070 selector_function = _build_selector_function(selector.selector)
1071 elif selector.type == PICKFIRST:
1072 fs = [_build_selector_function(s) for s in selector.selector]
1073
1074 def selector_function(ctx):
1075 for f in fs:
1076 picked_formats = list(f(ctx))
1077 if picked_formats:
1078 return picked_formats
1079 return []
1080 elif selector.type == SINGLE:
1081 format_spec = selector.selector
1082
1083 def selector_function(ctx):
1084 formats = list(ctx['formats'])
1085 if not formats:
1086 return
1087 if format_spec == 'all':
1088 for f in formats:
1089 yield f
1090 elif format_spec in ['best', 'worst', None]:
1091 format_idx = 0 if format_spec == 'worst' else -1
1092 audiovideo_formats = [
1093 f for f in formats
1094 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1095 if audiovideo_formats:
1096 yield audiovideo_formats[format_idx]
1097 # for extractors with incomplete formats (audio only (soundcloud)
1098 # or video only (imgur)) we will fallback to best/worst
1099 # {video,audio}-only format
1100 elif ctx['incomplete_formats']:
1101 yield formats[format_idx]
1102 elif format_spec == 'bestaudio':
1103 audio_formats = [
1104 f for f in formats
1105 if f.get('vcodec') == 'none']
1106 if audio_formats:
1107 yield audio_formats[-1]
1108 elif format_spec == 'worstaudio':
1109 audio_formats = [
1110 f for f in formats
1111 if f.get('vcodec') == 'none']
1112 if audio_formats:
1113 yield audio_formats[0]
1114 elif format_spec == 'bestvideo':
1115 video_formats = [
1116 f for f in formats
1117 if f.get('acodec') == 'none']
1118 if video_formats:
1119 yield video_formats[-1]
1120 elif format_spec == 'worstvideo':
1121 video_formats = [
1122 f for f in formats
1123 if f.get('acodec') == 'none']
1124 if video_formats:
1125 yield video_formats[0]
1126 else:
1127 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1128 if format_spec in extensions:
1129 filter_f = lambda f: f['ext'] == format_spec
1130 else:
1131 filter_f = lambda f: f['format_id'] == format_spec
1132 matches = list(filter(filter_f, formats))
1133 if matches:
1134 yield matches[-1]
1135 elif selector.type == MERGE:
1136 def _merge(formats_info):
1137 format_1, format_2 = [f['format_id'] for f in formats_info]
1138 # The first format must contain the video and the
1139 # second the audio
1140 if formats_info[0].get('vcodec') == 'none':
1141 self.report_error('The first format must '
1142 'contain the video, try using '
1143 '"-f %s+%s"' % (format_2, format_1))
1144 return
1145 # Formats must be opposite (video+audio)
1146 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1147 self.report_error(
1148 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1149 % (format_1, format_2))
1150 return
1151 output_ext = (
1152 formats_info[0]['ext']
1153 if self.params.get('merge_output_format') is None
1154 else self.params['merge_output_format'])
1155 return {
1156 'requested_formats': formats_info,
1157 'format': '%s+%s' % (formats_info[0].get('format'),
1158 formats_info[1].get('format')),
1159 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1160 formats_info[1].get('format_id')),
1161 'width': formats_info[0].get('width'),
1162 'height': formats_info[0].get('height'),
1163 'resolution': formats_info[0].get('resolution'),
1164 'fps': formats_info[0].get('fps'),
1165 'vcodec': formats_info[0].get('vcodec'),
1166 'vbr': formats_info[0].get('vbr'),
1167 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1168 'acodec': formats_info[1].get('acodec'),
1169 'abr': formats_info[1].get('abr'),
1170 'ext': output_ext,
1171 }
1172 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1173
1174 def selector_function(ctx):
1175 for pair in itertools.product(
1176 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1177 yield _merge(pair)
1178
1179 filters = [self._build_format_filter(f) for f in selector.filters]
1180
1181 def final_selector(ctx):
1182 ctx_copy = copy.deepcopy(ctx)
1183 for _filter in filters:
1184 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1185 return selector_function(ctx_copy)
1186 return final_selector
1187
1188 stream = io.BytesIO(format_spec.encode('utf-8'))
1189 try:
1190 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1191 except tokenize.TokenError:
1192 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1193
1194 class TokenIterator(object):
1195 def __init__(self, tokens):
1196 self.tokens = tokens
1197 self.counter = 0
1198
1199 def __iter__(self):
1200 return self
1201
1202 def __next__(self):
1203 if self.counter >= len(self.tokens):
1204 raise StopIteration()
1205 value = self.tokens[self.counter]
1206 self.counter += 1
1207 return value
1208
1209 next = __next__
1210
1211 def restore_last_token(self):
1212 self.counter -= 1
1213
1214 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1215 return _build_selector_function(parsed_selector)
1216
1217 def _calc_headers(self, info_dict):
1218 res = std_headers.copy()
1219
1220 add_headers = info_dict.get('http_headers')
1221 if add_headers:
1222 res.update(add_headers)
1223
1224 cookies = self._calc_cookies(info_dict)
1225 if cookies:
1226 res['Cookie'] = cookies
1227
1228 return res
1229
1230 def _calc_cookies(self, info_dict):
1231 pr = sanitized_Request(info_dict['url'])
1232 self.cookiejar.add_cookie_header(pr)
1233 return pr.get_header('Cookie')
1234
1235 def process_video_result(self, info_dict, download=True):
1236 assert info_dict.get('_type', 'video') == 'video'
1237
1238 if 'id' not in info_dict:
1239 raise ExtractorError('Missing "id" field in extractor result')
1240 if 'title' not in info_dict:
1241 raise ExtractorError('Missing "title" field in extractor result')
1242
1243 if not isinstance(info_dict['id'], compat_str):
1244 self.report_warning('"id" field is not a string - forcing string conversion')
1245 info_dict['id'] = compat_str(info_dict['id'])
1246
1247 if 'playlist' not in info_dict:
1248 # It isn't part of a playlist
1249 info_dict['playlist'] = None
1250 info_dict['playlist_index'] = None
1251
1252 thumbnails = info_dict.get('thumbnails')
1253 if thumbnails is None:
1254 thumbnail = info_dict.get('thumbnail')
1255 if thumbnail:
1256 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1257 if thumbnails:
1258 thumbnails.sort(key=lambda t: (
1259 t.get('preference'), t.get('width'), t.get('height'),
1260 t.get('id'), t.get('url')))
1261 for i, t in enumerate(thumbnails):
1262 t['url'] = sanitize_url(t['url'])
1263 if t.get('width') and t.get('height'):
1264 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1265 if t.get('id') is None:
1266 t['id'] = '%d' % i
1267
1268 if self.params.get('list_thumbnails'):
1269 self.list_thumbnails(info_dict)
1270 return
1271
1272 thumbnail = info_dict.get('thumbnail')
1273 if thumbnail:
1274 info_dict['thumbnail'] = sanitize_url(thumbnail)
1275 elif thumbnails:
1276 info_dict['thumbnail'] = thumbnails[-1]['url']
1277
1278 if 'display_id' not in info_dict and 'id' in info_dict:
1279 info_dict['display_id'] = info_dict['id']
1280
1281 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1282 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1283 # see http://bugs.python.org/issue1646728)
1284 try:
1285 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1286 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1287 except (ValueError, OverflowError, OSError):
1288 pass
1289
1290 # Auto generate title fields corresponding to the *_number fields when missing
1291 # in order to always have clean titles. This is very common for TV series.
1292 for field in ('chapter', 'season', 'episode'):
1293 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1294 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1295
1296 subtitles = info_dict.get('subtitles')
1297 if subtitles:
1298 for _, subtitle in subtitles.items():
1299 for subtitle_format in subtitle:
1300 if subtitle_format.get('url'):
1301 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1302 if 'ext' not in subtitle_format:
1303 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1304
1305 if self.params.get('listsubtitles', False):
1306 if 'automatic_captions' in info_dict:
1307 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1308 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1309 return
1310 info_dict['requested_subtitles'] = self.process_subtitles(
1311 info_dict['id'], subtitles,
1312 info_dict.get('automatic_captions'))
1313
1314 # We now pick which formats have to be downloaded
1315 if info_dict.get('formats') is None:
1316 # There's only one format available
1317 formats = [info_dict]
1318 else:
1319 formats = info_dict['formats']
1320
1321 if not formats:
1322 raise ExtractorError('No video formats found!')
1323
1324 formats_dict = {}
1325
1326 # We check that all the formats have the format and format_id fields
1327 for i, format in enumerate(formats):
1328 if 'url' not in format:
1329 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1330
1331 format['url'] = sanitize_url(format['url'])
1332
1333 if format.get('format_id') is None:
1334 format['format_id'] = compat_str(i)
1335 else:
1336 # Sanitize format_id from characters used in format selector expression
1337 format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
1338 format_id = format['format_id']
1339 if format_id not in formats_dict:
1340 formats_dict[format_id] = []
1341 formats_dict[format_id].append(format)
1342
1343 # Make sure all formats have unique format_id
1344 for format_id, ambiguous_formats in formats_dict.items():
1345 if len(ambiguous_formats) > 1:
1346 for i, format in enumerate(ambiguous_formats):
1347 format['format_id'] = '%s-%d' % (format_id, i)
1348
1349 for i, format in enumerate(formats):
1350 if format.get('format') is None:
1351 format['format'] = '{id} - {res}{note}'.format(
1352 id=format['format_id'],
1353 res=self.format_resolution(format),
1354 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1355 )
1356 # Automatically determine file extension if missing
1357 if 'ext' not in format:
1358 format['ext'] = determine_ext(format['url']).lower()
1359 # Automatically determine protocol if missing (useful for format
1360 # selection purposes)
1361 if 'protocol' not in format:
1362 format['protocol'] = determine_protocol(format)
1363 # Add HTTP headers, so that external programs can use them from the
1364 # json output
1365 full_format_info = info_dict.copy()
1366 full_format_info.update(format)
1367 format['http_headers'] = self._calc_headers(full_format_info)
1368
1369 # TODO Central sorting goes here
1370
1371 if formats[0] is not info_dict:
1372 # only set the 'formats' fields if the original info_dict list them
1373 # otherwise we end up with a circular reference, the first (and unique)
1374 # element in the 'formats' field in info_dict is info_dict itself,
1375 # which can't be exported to json
1376 info_dict['formats'] = formats
1377 if self.params.get('listformats'):
1378 self.list_formats(info_dict)
1379 return
1380
1381 req_format = self.params.get('format')
1382 if req_format is None:
1383 req_format_list = []
1384 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1385 not info_dict.get('is_live')):
1386 merger = FFmpegMergerPP(self)
1387 if merger.available and merger.can_merge():
1388 req_format_list.append('bestvideo+bestaudio')
1389 req_format_list.append('best')
1390 req_format = '/'.join(req_format_list)
1391 format_selector = self.build_format_selector(req_format)
1392
1393 # While in format selection we may need to have an access to the original
1394 # format set in order to calculate some metrics or do some processing.
1395 # For now we need to be able to guess whether original formats provided
1396 # by extractor are incomplete or not (i.e. whether extractor provides only
1397 # video-only or audio-only formats) for proper formats selection for
1398 # extractors with such incomplete formats (see
1399 # https://github.com/rg3/youtube-dl/pull/5556).
1400 # Since formats may be filtered during format selection and may not match
1401 # the original formats the results may be incorrect. Thus original formats
1402 # or pre-calculated metrics should be passed to format selection routines
1403 # as well.
1404 # We will pass a context object containing all necessary additional data
1405 # instead of just formats.
1406 # This fixes incorrect format selection issue (see
1407 # https://github.com/rg3/youtube-dl/issues/10083).
1408 incomplete_formats = (
1409 # All formats are video-only or
1410 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
1411 # all formats are audio-only
1412 all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1413
1414 ctx = {
1415 'formats': formats,
1416 'incomplete_formats': incomplete_formats,
1417 }
1418
1419 formats_to_download = list(format_selector(ctx))
1420 if not formats_to_download:
1421 raise ExtractorError('requested format not available',
1422 expected=True)
1423
1424 if download:
1425 if len(formats_to_download) > 1:
1426 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1427 for format in formats_to_download:
1428 new_info = dict(info_dict)
1429 new_info.update(format)
1430 self.process_info(new_info)
1431 # We update the info dict with the best quality format (backwards compatibility)
1432 info_dict.update(formats_to_download[-1])
1433 return info_dict
1434
1435 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1436 """Select the requested subtitles and their format"""
1437 available_subs = {}
1438 if normal_subtitles and self.params.get('writesubtitles'):
1439 available_subs.update(normal_subtitles)
1440 if automatic_captions and self.params.get('writeautomaticsub'):
1441 for lang, cap_info in automatic_captions.items():
1442 if lang not in available_subs:
1443 available_subs[lang] = cap_info
1444
1445 if (not self.params.get('writesubtitles') and not
1446 self.params.get('writeautomaticsub') or not
1447 available_subs):
1448 return None
1449
1450 if self.params.get('allsubtitles', False):
1451 requested_langs = available_subs.keys()
1452 else:
1453 if self.params.get('subtitleslangs', False):
1454 requested_langs = self.params.get('subtitleslangs')
1455 elif 'en' in available_subs:
1456 requested_langs = ['en']
1457 else:
1458 requested_langs = [list(available_subs.keys())[0]]
1459
1460 formats_query = self.params.get('subtitlesformat', 'best')
1461 formats_preference = formats_query.split('/') if formats_query else []
1462 subs = {}
1463 for lang in requested_langs:
1464 formats = available_subs.get(lang)
1465 if formats is None:
1466 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1467 continue
1468 for ext in formats_preference:
1469 if ext == 'best':
1470 f = formats[-1]
1471 break
1472 matches = list(filter(lambda f: f['ext'] == ext, formats))
1473 if matches:
1474 f = matches[-1]
1475 break
1476 else:
1477 f = formats[-1]
1478 self.report_warning(
1479 'No subtitle format found matching "%s" for language %s, '
1480 'using %s' % (formats_query, lang, f['ext']))
1481 subs[lang] = f
1482 return subs
1483
1484 def process_info(self, info_dict):
1485 """Process a single resolved IE result."""
1486
1487 assert info_dict.get('_type', 'video') == 'video'
1488
1489 max_downloads = self.params.get('max_downloads')
1490 if max_downloads is not None:
1491 if self._num_downloads >= int(max_downloads):
1492 raise MaxDownloadsReached()
1493
1494 info_dict['fulltitle'] = info_dict['title']
1495 if len(info_dict['title']) > 200:
1496 info_dict['title'] = info_dict['title'][:197] + '...'
1497
1498 if 'format' not in info_dict:
1499 info_dict['format'] = info_dict['ext']
1500
1501 reason = self._match_entry(info_dict, incomplete=False)
1502 if reason is not None:
1503 self.to_screen('[download] ' + reason)
1504 return
1505
1506 self._num_downloads += 1
1507
1508 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1509
1510 # Forced printings
1511 if self.params.get('forcetitle', False):
1512 self.to_stdout(info_dict['fulltitle'])
1513 if self.params.get('forceid', False):
1514 self.to_stdout(info_dict['id'])
1515 if self.params.get('forceurl', False):
1516 if info_dict.get('requested_formats') is not None:
1517 for f in info_dict['requested_formats']:
1518 self.to_stdout(f['url'] + f.get('play_path', ''))
1519 else:
1520 # For RTMP URLs, also include the playpath
1521 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1522 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1523 self.to_stdout(info_dict['thumbnail'])
1524 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1525 self.to_stdout(info_dict['description'])
1526 if self.params.get('forcefilename', False) and filename is not None:
1527 self.to_stdout(filename)
1528 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1529 self.to_stdout(formatSeconds(info_dict['duration']))
1530 if self.params.get('forceformat', False):
1531 self.to_stdout(info_dict['format'])
1532 if self.params.get('forcejson', False):
1533 self.to_stdout(json.dumps(info_dict))
1534
1535 # Do nothing else if in simulate mode
1536 if self.params.get('simulate', False):
1537 return
1538
1539 if filename is None:
1540 return
1541
1542 try:
1543 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1544 if dn and not os.path.exists(dn):
1545 os.makedirs(dn)
1546 except (OSError, IOError) as err:
1547 self.report_error('unable to create directory ' + error_to_compat_str(err))
1548 return
1549
1550 if self.params.get('writedescription', False):
1551 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1552 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1553 self.to_screen('[info] Video description is already present')
1554 elif info_dict.get('description') is None:
1555 self.report_warning('There\'s no description to write.')
1556 else:
1557 try:
1558 self.to_screen('[info] Writing video description to: ' + descfn)
1559 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1560 descfile.write(info_dict['description'])
1561 except (OSError, IOError):
1562 self.report_error('Cannot write description file ' + descfn)
1563 return
1564
1565 if self.params.get('writeannotations', False):
1566 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1567 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1568 self.to_screen('[info] Video annotations are already present')
1569 else:
1570 try:
1571 self.to_screen('[info] Writing video annotations to: ' + annofn)
1572 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1573 annofile.write(info_dict['annotations'])
1574 except (KeyError, TypeError):
1575 self.report_warning('There are no annotations to write.')
1576 except (OSError, IOError):
1577 self.report_error('Cannot write annotations file: ' + annofn)
1578 return
1579
1580 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1581 self.params.get('writeautomaticsub')])
1582
1583 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1584 # subtitles download errors are already managed as troubles in relevant IE
1585 # that way it will silently go on when used with unsupporting IE
1586 subtitles = info_dict['requested_subtitles']
1587 ie = self.get_info_extractor(info_dict['extractor_key'])
1588 for sub_lang, sub_info in subtitles.items():
1589 sub_format = sub_info['ext']
1590 if sub_info.get('data') is not None:
1591 sub_data = sub_info['data']
1592 else:
1593 try:
1594 sub_data = ie._download_webpage(
1595 sub_info['url'], info_dict['id'], note=False)
1596 except ExtractorError as err:
1597 self.report_warning('Unable to download subtitle for "%s": %s' %
1598 (sub_lang, error_to_compat_str(err.cause)))
1599 continue
1600 try:
1601 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1602 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1603 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1604 else:
1605 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1606 # Use newline='' to prevent conversion of newline characters
1607 # See https://github.com/rg3/youtube-dl/issues/10268
1608 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1609 subfile.write(sub_data)
1610 except (OSError, IOError):
1611 self.report_error('Cannot write subtitles file ' + sub_filename)
1612 return
1613
1614 if self.params.get('writeinfojson', False):
1615 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1616 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1617 self.to_screen('[info] Video description metadata is already present')
1618 else:
1619 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1620 try:
1621 write_json_file(self.filter_requested_info(info_dict), infofn)
1622 except (OSError, IOError):
1623 self.report_error('Cannot write metadata to JSON file ' + infofn)
1624 return
1625
1626 self._write_thumbnails(info_dict, filename)
1627
1628 if not self.params.get('skip_download', False):
1629 try:
1630 def dl(name, info):
1631 fd = get_suitable_downloader(info, self.params)(self, self.params)
1632 for ph in self._progress_hooks:
1633 fd.add_progress_hook(ph)
1634 if self.params.get('verbose'):
1635 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1636 return fd.download(name, info)
1637
1638 if info_dict.get('requested_formats') is not None:
1639 downloaded = []
1640 success = True
1641 merger = FFmpegMergerPP(self)
1642 if not merger.available:
1643 postprocessors = []
1644 self.report_warning('You have requested multiple '
1645 'formats but ffmpeg or avconv are not installed.'
1646 ' The formats won\'t be merged.')
1647 else:
1648 postprocessors = [merger]
1649
1650 def compatible_formats(formats):
1651 video, audio = formats
1652 # Check extension
1653 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1654 if video_ext and audio_ext:
1655 COMPATIBLE_EXTS = (
1656 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1657 ('webm')
1658 )
1659 for exts in COMPATIBLE_EXTS:
1660 if video_ext in exts and audio_ext in exts:
1661 return True
1662 # TODO: Check acodec/vcodec
1663 return False
1664
1665 filename_real_ext = os.path.splitext(filename)[1][1:]
1666 filename_wo_ext = (
1667 os.path.splitext(filename)[0]
1668 if filename_real_ext == info_dict['ext']
1669 else filename)
1670 requested_formats = info_dict['requested_formats']
1671 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1672 info_dict['ext'] = 'mkv'
1673 self.report_warning(
1674 'Requested formats are incompatible for merge and will be merged into mkv.')
1675 # Ensure filename always has a correct extension for successful merge
1676 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1677 if os.path.exists(encodeFilename(filename)):
1678 self.to_screen(
1679 '[download] %s has already been downloaded and '
1680 'merged' % filename)
1681 else:
1682 for f in requested_formats:
1683 new_info = dict(info_dict)
1684 new_info.update(f)
1685 fname = self.prepare_filename(new_info)
1686 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1687 downloaded.append(fname)
1688 partial_success = dl(fname, new_info)
1689 success = success and partial_success
1690 info_dict['__postprocessors'] = postprocessors
1691 info_dict['__files_to_merge'] = downloaded
1692 else:
1693 # Just a single file
1694 success = dl(filename, info_dict)
1695 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1696 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1697 return
1698 except (OSError, IOError) as err:
1699 raise UnavailableVideoError(err)
1700 except (ContentTooShortError, ) as err:
1701 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1702 return
1703
1704 if success and filename != '-':
1705 # Fixup content
1706 fixup_policy = self.params.get('fixup')
1707 if fixup_policy is None:
1708 fixup_policy = 'detect_or_warn'
1709
1710 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1711
1712 stretched_ratio = info_dict.get('stretched_ratio')
1713 if stretched_ratio is not None and stretched_ratio != 1:
1714 if fixup_policy == 'warn':
1715 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1716 info_dict['id'], stretched_ratio))
1717 elif fixup_policy == 'detect_or_warn':
1718 stretched_pp = FFmpegFixupStretchedPP(self)
1719 if stretched_pp.available:
1720 info_dict.setdefault('__postprocessors', [])
1721 info_dict['__postprocessors'].append(stretched_pp)
1722 else:
1723 self.report_warning(
1724 '%s: Non-uniform pixel ratio (%s). %s'
1725 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1726 else:
1727 assert fixup_policy in ('ignore', 'never')
1728
1729 if (info_dict.get('requested_formats') is None and
1730 info_dict.get('container') == 'm4a_dash'):
1731 if fixup_policy == 'warn':
1732 self.report_warning(
1733 '%s: writing DASH m4a. '
1734 'Only some players support this container.'
1735 % info_dict['id'])
1736 elif fixup_policy == 'detect_or_warn':
1737 fixup_pp = FFmpegFixupM4aPP(self)
1738 if fixup_pp.available:
1739 info_dict.setdefault('__postprocessors', [])
1740 info_dict['__postprocessors'].append(fixup_pp)
1741 else:
1742 self.report_warning(
1743 '%s: writing DASH m4a. '
1744 'Only some players support this container. %s'
1745 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1746 else:
1747 assert fixup_policy in ('ignore', 'never')
1748
1749 if (info_dict.get('protocol') == 'm3u8_native' or
1750 info_dict.get('protocol') == 'm3u8' and
1751 self.params.get('hls_prefer_native')):
1752 if fixup_policy == 'warn':
1753 self.report_warning('%s: malformated aac bitstream.' % (
1754 info_dict['id']))
1755 elif fixup_policy == 'detect_or_warn':
1756 fixup_pp = FFmpegFixupM3u8PP(self)
1757 if fixup_pp.available:
1758 info_dict.setdefault('__postprocessors', [])
1759 info_dict['__postprocessors'].append(fixup_pp)
1760 else:
1761 self.report_warning(
1762 '%s: malformated aac bitstream. %s'
1763 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1764 else:
1765 assert fixup_policy in ('ignore', 'never')
1766
1767 try:
1768 self.post_process(filename, info_dict)
1769 except (PostProcessingError) as err:
1770 self.report_error('postprocessing: %s' % str(err))
1771 return
1772 self.record_download_archive(info_dict)
1773
1774 def download(self, url_list):
1775 """Download a given list of URLs."""
1776 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1777 if (len(url_list) > 1 and
1778 '%' not in outtmpl and
1779 self.params.get('max_downloads') != 1):
1780 raise SameFileError(outtmpl)
1781
1782 for url in url_list:
1783 try:
1784 # It also downloads the videos
1785 res = self.extract_info(
1786 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1787 except UnavailableVideoError:
1788 self.report_error('unable to download video')
1789 except MaxDownloadsReached:
1790 self.to_screen('[info] Maximum number of downloaded files reached.')
1791 raise
1792 else:
1793 if self.params.get('dump_single_json', False):
1794 self.to_stdout(json.dumps(res))
1795
1796 return self._download_retcode
1797
1798 def download_with_info_file(self, info_filename):
1799 with contextlib.closing(fileinput.FileInput(
1800 [info_filename], mode='r',
1801 openhook=fileinput.hook_encoded('utf-8'))) as f:
1802 # FileInput doesn't have a read method, we can't call json.load
1803 info = self.filter_requested_info(json.loads('\n'.join(f)))
1804 try:
1805 self.process_ie_result(info, download=True)
1806 except DownloadError:
1807 webpage_url = info.get('webpage_url')
1808 if webpage_url is not None:
1809 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1810 return self.download([webpage_url])
1811 else:
1812 raise
1813 return self._download_retcode
1814
1815 @staticmethod
1816 def filter_requested_info(info_dict):
1817 return dict(
1818 (k, v) for k, v in info_dict.items()
1819 if k not in ['requested_formats', 'requested_subtitles'])
1820
1821 def post_process(self, filename, ie_info):
1822 """Run all the postprocessors on the given file."""
1823 info = dict(ie_info)
1824 info['filepath'] = filename
1825 pps_chain = []
1826 if ie_info.get('__postprocessors') is not None:
1827 pps_chain.extend(ie_info['__postprocessors'])
1828 pps_chain.extend(self._pps)
1829 for pp in pps_chain:
1830 files_to_delete = []
1831 try:
1832 files_to_delete, info = pp.run(info)
1833 except PostProcessingError as e:
1834 self.report_error(e.msg)
1835 if files_to_delete and not self.params.get('keepvideo', False):
1836 for old_filename in files_to_delete:
1837 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1838 try:
1839 os.remove(encodeFilename(old_filename))
1840 except (IOError, OSError):
1841 self.report_warning('Unable to remove downloaded original file')
1842
1843 def _make_archive_id(self, info_dict):
1844 # Future-proof against any change in case
1845 # and backwards compatibility with prior versions
1846 extractor = info_dict.get('extractor_key')
1847 if extractor is None:
1848 if 'id' in info_dict:
1849 extractor = info_dict.get('ie_key') # key in a playlist
1850 if extractor is None:
1851 return None # Incomplete video information
1852 return extractor.lower() + ' ' + info_dict['id']
1853
1854 def in_download_archive(self, info_dict):
1855 fn = self.params.get('download_archive')
1856 if fn is None:
1857 return False
1858
1859 vid_id = self._make_archive_id(info_dict)
1860 if vid_id is None:
1861 return False # Incomplete video information
1862
1863 try:
1864 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1865 for line in archive_file:
1866 if line.strip() == vid_id:
1867 return True
1868 except IOError as ioe:
1869 if ioe.errno != errno.ENOENT:
1870 raise
1871 return False
1872
1873 def record_download_archive(self, info_dict):
1874 fn = self.params.get('download_archive')
1875 if fn is None:
1876 return
1877 vid_id = self._make_archive_id(info_dict)
1878 assert vid_id
1879 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1880 archive_file.write(vid_id + '\n')
1881
1882 @staticmethod
1883 def format_resolution(format, default='unknown'):
1884 if format.get('vcodec') == 'none':
1885 return 'audio only'
1886 if format.get('resolution') is not None:
1887 return format['resolution']
1888 if format.get('height') is not None:
1889 if format.get('width') is not None:
1890 res = '%sx%s' % (format['width'], format['height'])
1891 else:
1892 res = '%sp' % format['height']
1893 elif format.get('width') is not None:
1894 res = '%dx?' % format['width']
1895 else:
1896 res = default
1897 return res
1898
1899 def _format_note(self, fdict):
1900 res = ''
1901 if fdict.get('ext') in ['f4f', 'f4m']:
1902 res += '(unsupported) '
1903 if fdict.get('language'):
1904 if res:
1905 res += ' '
1906 res += '[%s] ' % fdict['language']
1907 if fdict.get('format_note') is not None:
1908 res += fdict['format_note'] + ' '
1909 if fdict.get('tbr') is not None:
1910 res += '%4dk ' % fdict['tbr']
1911 if fdict.get('container') is not None:
1912 if res:
1913 res += ', '
1914 res += '%s container' % fdict['container']
1915 if (fdict.get('vcodec') is not None and
1916 fdict.get('vcodec') != 'none'):
1917 if res:
1918 res += ', '
1919 res += fdict['vcodec']
1920 if fdict.get('vbr') is not None:
1921 res += '@'
1922 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1923 res += 'video@'
1924 if fdict.get('vbr') is not None:
1925 res += '%4dk' % fdict['vbr']
1926 if fdict.get('fps') is not None:
1927 if res:
1928 res += ', '
1929 res += '%sfps' % fdict['fps']
1930 if fdict.get('acodec') is not None:
1931 if res:
1932 res += ', '
1933 if fdict['acodec'] == 'none':
1934 res += 'video only'
1935 else:
1936 res += '%-5s' % fdict['acodec']
1937 elif fdict.get('abr') is not None:
1938 if res:
1939 res += ', '
1940 res += 'audio'
1941 if fdict.get('abr') is not None:
1942 res += '@%3dk' % fdict['abr']
1943 if fdict.get('asr') is not None:
1944 res += ' (%5dHz)' % fdict['asr']
1945 if fdict.get('filesize') is not None:
1946 if res:
1947 res += ', '
1948 res += format_bytes(fdict['filesize'])
1949 elif fdict.get('filesize_approx') is not None:
1950 if res:
1951 res += ', '
1952 res += '~' + format_bytes(fdict['filesize_approx'])
1953 return res
1954
1955 def list_formats(self, info_dict):
1956 formats = info_dict.get('formats', [info_dict])
1957 table = [
1958 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1959 for f in formats
1960 if f.get('preference') is None or f['preference'] >= -1000]
1961 if len(formats) > 1:
1962 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1963
1964 header_line = ['format code', 'extension', 'resolution', 'note']
1965 self.to_screen(
1966 '[info] Available formats for %s:\n%s' %
1967 (info_dict['id'], render_table(header_line, table)))
1968
1969 def list_thumbnails(self, info_dict):
1970 thumbnails = info_dict.get('thumbnails')
1971 if not thumbnails:
1972 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
1973 return
1974
1975 self.to_screen(
1976 '[info] Thumbnails for %s:' % info_dict['id'])
1977 self.to_screen(render_table(
1978 ['ID', 'width', 'height', 'URL'],
1979 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1980
1981 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1982 if not subtitles:
1983 self.to_screen('%s has no %s' % (video_id, name))
1984 return
1985 self.to_screen(
1986 'Available %s for %s:' % (name, video_id))
1987 self.to_screen(render_table(
1988 ['Language', 'formats'],
1989 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1990 for lang, formats in subtitles.items()]))
1991
1992 def urlopen(self, req):
1993 """ Start an HTTP download """
1994 if isinstance(req, compat_basestring):
1995 req = sanitized_Request(req)
1996 return self._opener.open(req, timeout=self._socket_timeout)
1997
1998 def print_debug_header(self):
1999 if not self.params.get('verbose'):
2000 return
2001
2002 if type('') is not compat_str:
2003 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
2004 self.report_warning(
2005 'Your Python is broken! Update to a newer and supported version')
2006
2007 stdout_encoding = getattr(
2008 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2009 encoding_str = (
2010 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2011 locale.getpreferredencoding(),
2012 sys.getfilesystemencoding(),
2013 stdout_encoding,
2014 self.get_encoding()))
2015 write_string(encoding_str, encoding=None)
2016
2017 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2018 if _LAZY_LOADER:
2019 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2020 try:
2021 sp = subprocess.Popen(
2022 ['git', 'rev-parse', '--short', 'HEAD'],
2023 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2024 cwd=os.path.dirname(os.path.abspath(__file__)))
2025 out, err = sp.communicate()
2026 out = out.decode().strip()
2027 if re.match('[0-9a-f]+', out):
2028 self._write_string('[debug] Git HEAD: ' + out + '\n')
2029 except Exception:
2030 try:
2031 sys.exc_clear()
2032 except Exception:
2033 pass
2034 self._write_string('[debug] Python version %s - %s\n' % (
2035 platform.python_version(), platform_name()))
2036
2037 exe_versions = FFmpegPostProcessor.get_versions(self)
2038 exe_versions['rtmpdump'] = rtmpdump_version()
2039 exe_str = ', '.join(
2040 '%s %s' % (exe, v)
2041 for exe, v in sorted(exe_versions.items())
2042 if v
2043 )
2044 if not exe_str:
2045 exe_str = 'none'
2046 self._write_string('[debug] exe versions: %s\n' % exe_str)
2047
2048 proxy_map = {}
2049 for handler in self._opener.handlers:
2050 if hasattr(handler, 'proxies'):
2051 proxy_map.update(handler.proxies)
2052 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2053
2054 if self.params.get('call_home', False):
2055 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2056 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2057 latest_version = self.urlopen(
2058 'https://yt-dl.org/latest/version').read().decode('utf-8')
2059 if version_tuple(latest_version) > version_tuple(__version__):
2060 self.report_warning(
2061 'You are using an outdated version (newest version: %s)! '
2062 'See https://yt-dl.org/update if you need help updating.' %
2063 latest_version)
2064
2065 def _setup_opener(self):
2066 timeout_val = self.params.get('socket_timeout')
2067 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2068
2069 opts_cookiefile = self.params.get('cookiefile')
2070 opts_proxy = self.params.get('proxy')
2071
2072 if opts_cookiefile is None:
2073 self.cookiejar = compat_cookiejar.CookieJar()
2074 else:
2075 opts_cookiefile = compat_expanduser(opts_cookiefile)
2076 self.cookiejar = compat_cookiejar.MozillaCookieJar(
2077 opts_cookiefile)
2078 if os.access(opts_cookiefile, os.R_OK):
2079 self.cookiejar.load()
2080
2081 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2082 if opts_proxy is not None:
2083 if opts_proxy == '':
2084 proxies = {}
2085 else:
2086 proxies = {'http': opts_proxy, 'https': opts_proxy}
2087 else:
2088 proxies = compat_urllib_request.getproxies()
2089 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
2090 if 'http' in proxies and 'https' not in proxies:
2091 proxies['https'] = proxies['http']
2092 proxy_handler = PerRequestProxyHandler(proxies)
2093
2094 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2095 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2096 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2097 data_handler = compat_urllib_request_DataHandler()
2098
2099 # When passing our own FileHandler instance, build_opener won't add the
2100 # default FileHandler and allows us to disable the file protocol, which
2101 # can be used for malicious purposes (see
2102 # https://github.com/rg3/youtube-dl/issues/8227)
2103 file_handler = compat_urllib_request.FileHandler()
2104
2105 def file_open(*args, **kwargs):
2106 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2107 file_handler.file_open = file_open
2108
2109 opener = compat_urllib_request.build_opener(
2110 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2111
2112 # Delete the default user-agent header, which would otherwise apply in
2113 # cases where our custom HTTP handler doesn't come into play
2114 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2115 opener.addheaders = []
2116 self._opener = opener
2117
2118 def encode(self, s):
2119 if isinstance(s, bytes):
2120 return s # Already encoded
2121
2122 try:
2123 return s.encode(self.get_encoding())
2124 except UnicodeEncodeError as err:
2125 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2126 raise
2127
2128 def get_encoding(self):
2129 encoding = self.params.get('encoding')
2130 if encoding is None:
2131 encoding = preferredencoding()
2132 return encoding
2133
2134 def _write_thumbnails(self, info_dict, filename):
2135 if self.params.get('writethumbnail', False):
2136 thumbnails = info_dict.get('thumbnails')
2137 if thumbnails:
2138 thumbnails = [thumbnails[-1]]
2139 elif self.params.get('write_all_thumbnails', False):
2140 thumbnails = info_dict.get('thumbnails')
2141 else:
2142 return
2143
2144 if not thumbnails:
2145 # No thumbnails present, so return immediately
2146 return
2147
2148 for t in thumbnails:
2149 thumb_ext = determine_ext(t['url'], 'jpg')
2150 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2151 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2152 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2153
2154 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2155 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2156 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2157 else:
2158 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2159 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2160 try:
2161 uf = self.urlopen(t['url'])
2162 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2163 shutil.copyfileobj(uf, thumbf)
2164 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2165 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2166 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2167 self.report_warning('Unable to download thumbnail "%s": %s' %
2168 (t['url'], error_to_compat_str(err)))