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