]> jfr.im git - yt-dlp.git/blob - youtube_dl/YoutubeDL.py
94457188175fd52c298f4abb272812d9ffc02149
[yt-dlp.git] / youtube_dl / YoutubeDL.py
1 #!/usr/bin/env python
2 # -*- coding: utf-8 -*-
3
4 from __future__ import absolute_import, unicode_literals
5
6 import collections
7 import contextlib
8 import datetime
9 import errno
10 import fileinput
11 import io
12 import itertools
13 import json
14 import locale
15 import operator
16 import os
17 import platform
18 import re
19 import shutil
20 import subprocess
21 import socket
22 import sys
23 import time
24 import traceback
25
26 if os.name == 'nt':
27 import ctypes
28
29 from .compat import (
30 compat_basestring,
31 compat_cookiejar,
32 compat_expanduser,
33 compat_get_terminal_size,
34 compat_http_client,
35 compat_kwargs,
36 compat_str,
37 compat_urllib_error,
38 compat_urllib_request,
39 )
40 from .utils import (
41 escape_url,
42 ContentTooShortError,
43 date_from_str,
44 DateRange,
45 DEFAULT_OUTTMPL,
46 determine_ext,
47 DownloadError,
48 encodeFilename,
49 ExtractorError,
50 format_bytes,
51 formatSeconds,
52 locked_file,
53 make_HTTPS_handler,
54 MaxDownloadsReached,
55 PagedList,
56 parse_filesize,
57 PerRequestProxyHandler,
58 PostProcessingError,
59 platform_name,
60 preferredencoding,
61 render_table,
62 SameFileError,
63 sanitize_filename,
64 sanitize_path,
65 std_headers,
66 subtitles_filename,
67 takewhile_inclusive,
68 UnavailableVideoError,
69 url_basename,
70 version_tuple,
71 write_json_file,
72 write_string,
73 YoutubeDLHandler,
74 prepend_extension,
75 args_to_str,
76 age_restricted,
77 )
78 from .cache import Cache
79 from .extractor import get_info_extractor, gen_extractors
80 from .downloader import get_suitable_downloader
81 from .downloader.rtmp import rtmpdump_version
82 from .postprocessor import (
83 FFmpegFixupM4aPP,
84 FFmpegFixupStretchedPP,
85 FFmpegMergerPP,
86 FFmpegPostProcessor,
87 get_postprocessor,
88 )
89 from .version import __version__
90
91
92 class YoutubeDL(object):
93 """YoutubeDL class.
94
95 YoutubeDL objects are the ones responsible of downloading the
96 actual video file and writing it to disk if the user has requested
97 it, among some other tasks. In most cases there should be one per
98 program. As, given a video URL, the downloader doesn't know how to
99 extract all the needed information, task that InfoExtractors do, it
100 has to pass the URL to one of them.
101
102 For this, YoutubeDL objects have a method that allows
103 InfoExtractors to be registered in a given order. When it is passed
104 a URL, the YoutubeDL object handles it to the first InfoExtractor it
105 finds that reports being able to handle it. The InfoExtractor extracts
106 all the information about the video or videos the URL refers to, and
107 YoutubeDL process the extracted information, possibly using a File
108 Downloader to download the video.
109
110 YoutubeDL objects accept a lot of parameters. In order not to saturate
111 the object constructor with arguments, it receives a dictionary of
112 options instead. These options are available through the params
113 attribute for the InfoExtractors to use. The YoutubeDL also
114 registers itself as the downloader in charge for the InfoExtractors
115 that are added to it, so this is a "mutual registration".
116
117 Available options:
118
119 username: Username for authentication purposes.
120 password: Password for authentication purposes.
121 videopassword: Password for acces a video.
122 usenetrc: Use netrc for authentication instead.
123 verbose: Print additional info to stdout.
124 quiet: Do not print messages to stdout.
125 no_warnings: Do not print out anything for warnings.
126 forceurl: Force printing final URL.
127 forcetitle: Force printing title.
128 forceid: Force printing ID.
129 forcethumbnail: Force printing thumbnail URL.
130 forcedescription: Force printing description.
131 forcefilename: Force printing final filename.
132 forceduration: Force printing duration.
133 forcejson: Force printing info_dict as JSON.
134 dump_single_json: Force printing the info_dict of the whole playlist
135 (or video) as a single JSON line.
136 simulate: Do not download the video files.
137 format: Video format code. See options.py for more information.
138 format_limit: Highest quality format to try.
139 outtmpl: Template for output names.
140 restrictfilenames: Do not allow "&" and spaces in file names
141 ignoreerrors: Do not stop on download errors.
142 nooverwrites: Prevent overwriting files.
143 playliststart: Playlist item to start at.
144 playlistend: Playlist item to end at.
145 playlist_items: Specific indices of playlist to download.
146 playlistreverse: Download playlist items in reverse order.
147 matchtitle: Download only matching titles.
148 rejecttitle: Reject downloads for matching titles.
149 logger: Log messages to a logging.Logger instance.
150 logtostderr: Log messages to stderr instead of stdout.
151 writedescription: Write the video description to a .description file
152 writeinfojson: Write the video description to a .info.json file
153 writeannotations: Write the video annotations to a .annotations.xml file
154 writethumbnail: Write the thumbnail image to a file
155 write_all_thumbnails: Write all thumbnail formats to files
156 writesubtitles: Write the video subtitles to a file
157 writeautomaticsub: Write the automatic subtitles to a file
158 allsubtitles: Downloads all the subtitles of the video
159 (requires writesubtitles or writeautomaticsub)
160 listsubtitles: Lists all available subtitles for the video
161 subtitlesformat: The format code for subtitles
162 subtitleslangs: List of languages of the subtitles to download
163 keepvideo: Keep the video file after post-processing
164 daterange: A DateRange object, download only if the upload_date is in the range.
165 skip_download: Skip the actual download of the video file
166 cachedir: Location of the cache files in the filesystem.
167 False to disable filesystem cache.
168 noplaylist: Download single video instead of a playlist if in doubt.
169 age_limit: An integer representing the user's age in years.
170 Unsuitable videos for the given age are skipped.
171 min_views: An integer representing the minimum view count the video
172 must have in order to not be skipped.
173 Videos without view count information are always
174 downloaded. None for no limit.
175 max_views: An integer representing the maximum view count.
176 Videos that are more popular than that are not
177 downloaded.
178 Videos without view count information are always
179 downloaded. None for no limit.
180 download_archive: File name of a file where all downloads are recorded.
181 Videos already present in the file are not downloaded
182 again.
183 cookiefile: File name where cookies should be read from and dumped to.
184 nocheckcertificate:Do not verify SSL certificates
185 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
186 At the moment, this is only supported by YouTube.
187 proxy: URL of the proxy server to use
188 cn_verification_proxy: URL of the proxy to use for IP address verification
189 on Chinese sites. (Experimental)
190 socket_timeout: Time to wait for unresponsive hosts, in seconds
191 bidi_workaround: Work around buggy terminals without bidirectional text
192 support, using fridibi
193 debug_printtraffic:Print out sent and received HTTP traffic
194 include_ads: Download ads as well
195 default_search: Prepend this string if an input url is not valid.
196 'auto' for elaborate guessing
197 encoding: Use this encoding instead of the system-specified.
198 extract_flat: Do not resolve URLs, return the immediate result.
199 Pass in 'in_playlist' to only show this behavior for
200 playlist items.
201 postprocessors: A list of dictionaries, each with an entry
202 * key: The name of the postprocessor. See
203 youtube_dl/postprocessor/__init__.py for a list.
204 as well as any further keyword arguments for the
205 postprocessor.
206 progress_hooks: A list of functions that get called on download
207 progress, with a dictionary with the entries
208 * status: One of "downloading", "error", or "finished".
209 Check this first and ignore unknown values.
210
211 If status is one of "downloading", or "finished", the
212 following properties may also be present:
213 * filename: The final filename (always present)
214 * tmpfilename: The filename we're currently writing to
215 * downloaded_bytes: Bytes on disk
216 * total_bytes: Size of the whole file, None if unknown
217 * total_bytes_estimate: Guess of the eventual file size,
218 None if unavailable.
219 * elapsed: The number of seconds since download started.
220 * eta: The estimated time in seconds, None if unknown
221 * speed: The download speed in bytes/second, None if
222 unknown
223 * fragment_index: The counter of the currently
224 downloaded video fragment.
225 * fragment_count: The number of fragments (= individual
226 files that will be merged)
227
228 Progress hooks are guaranteed to be called at least once
229 (with status "finished") if the download is successful.
230 merge_output_format: Extension to use when merging formats.
231 fixup: Automatically correct known faults of the file.
232 One of:
233 - "never": do nothing
234 - "warn": only emit a warning
235 - "detect_or_warn": check whether we can do anything
236 about it, warn otherwise (default)
237 source_address: (Experimental) Client-side IP address to bind to.
238 call_home: Boolean, true iff we are allowed to contact the
239 youtube-dl servers for debugging.
240 sleep_interval: Number of seconds to sleep before each download.
241 listformats: Print an overview of available video formats and exit.
242 list_thumbnails: Print a table of all thumbnails and exit.
243 match_filter: A function that gets called with the info_dict of
244 every video.
245 If it returns a message, the video is ignored.
246 If it returns None, the video is downloaded.
247 match_filter_func in utils.py is one example for this.
248 no_color: Do not emit color codes in output.
249
250 The following options determine which downloader is picked:
251 external_downloader: Executable of the external downloader to call.
252 None or unset for standard (built-in) downloader.
253 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
254
255 The following parameters are not used by YoutubeDL itself, they are used by
256 the downloader (see youtube_dl/downloader/common.py):
257 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
258 noresizebuffer, retries, continuedl, noprogress, consoletitle,
259 xattr_set_filesize, external_downloader_args.
260
261 The following options are used by the post processors:
262 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
263 otherwise prefer avconv.
264 exec_cmd: Arbitrary command to run after downloading
265 """
266
267 params = None
268 _ies = []
269 _pps = []
270 _download_retcode = None
271 _num_downloads = None
272 _screen_file = None
273
274 def __init__(self, params=None, auto_init=True):
275 """Create a FileDownloader object with the given options."""
276 if params is None:
277 params = {}
278 self._ies = []
279 self._ies_instances = {}
280 self._pps = []
281 self._progress_hooks = []
282 self._download_retcode = 0
283 self._num_downloads = 0
284 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
285 self._err_file = sys.stderr
286 self.params = params
287 self.cache = Cache(self)
288
289 if params.get('bidi_workaround', False):
290 try:
291 import pty
292 master, slave = pty.openpty()
293 width = compat_get_terminal_size().columns
294 if width is None:
295 width_args = []
296 else:
297 width_args = ['-w', str(width)]
298 sp_kwargs = dict(
299 stdin=subprocess.PIPE,
300 stdout=slave,
301 stderr=self._err_file)
302 try:
303 self._output_process = subprocess.Popen(
304 ['bidiv'] + width_args, **sp_kwargs
305 )
306 except OSError:
307 self._output_process = subprocess.Popen(
308 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
309 self._output_channel = os.fdopen(master, 'rb')
310 except OSError as ose:
311 if ose.errno == 2:
312 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 $PATH.')
313 else:
314 raise
315
316 if (sys.version_info >= (3,) and sys.platform != 'win32' and
317 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
318 not params.get('restrictfilenames', False)):
319 # On Python 3, the Unicode filesystem API will throw errors (#1474)
320 self.report_warning(
321 'Assuming --restrict-filenames since file system encoding '
322 'cannot encode all characters. '
323 'Set the LC_ALL environment variable to fix this.')
324 self.params['restrictfilenames'] = True
325
326 if isinstance(params.get('outtmpl'), bytes):
327 self.report_warning(
328 'Parameter outtmpl is bytes, but should be a unicode string. '
329 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
330
331 self._setup_opener()
332
333 if auto_init:
334 self.print_debug_header()
335 self.add_default_info_extractors()
336
337 for pp_def_raw in self.params.get('postprocessors', []):
338 pp_class = get_postprocessor(pp_def_raw['key'])
339 pp_def = dict(pp_def_raw)
340 del pp_def['key']
341 pp = pp_class(self, **compat_kwargs(pp_def))
342 self.add_post_processor(pp)
343
344 for ph in self.params.get('progress_hooks', []):
345 self.add_progress_hook(ph)
346
347 def warn_if_short_id(self, argv):
348 # short YouTube ID starting with dash?
349 idxs = [
350 i for i, a in enumerate(argv)
351 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
352 if idxs:
353 correct_argv = (
354 ['youtube-dl'] +
355 [a for i, a in enumerate(argv) if i not in idxs] +
356 ['--'] + [argv[i] for i in idxs]
357 )
358 self.report_warning(
359 'Long argument string detected. '
360 'Use -- to separate parameters and URLs, like this:\n%s\n' %
361 args_to_str(correct_argv))
362
363 def add_info_extractor(self, ie):
364 """Add an InfoExtractor object to the end of the list."""
365 self._ies.append(ie)
366 self._ies_instances[ie.ie_key()] = ie
367 ie.set_downloader(self)
368
369 def get_info_extractor(self, ie_key):
370 """
371 Get an instance of an IE with name ie_key, it will try to get one from
372 the _ies list, if there's no instance it will create a new one and add
373 it to the extractor list.
374 """
375 ie = self._ies_instances.get(ie_key)
376 if ie is None:
377 ie = get_info_extractor(ie_key)()
378 self.add_info_extractor(ie)
379 return ie
380
381 def add_default_info_extractors(self):
382 """
383 Add the InfoExtractors returned by gen_extractors to the end of the list
384 """
385 for ie in gen_extractors():
386 self.add_info_extractor(ie)
387
388 def add_post_processor(self, pp):
389 """Add a PostProcessor object to the end of the chain."""
390 self._pps.append(pp)
391 pp.set_downloader(self)
392
393 def add_progress_hook(self, ph):
394 """Add the progress hook (currently only for the file downloader)"""
395 self._progress_hooks.append(ph)
396
397 def _bidi_workaround(self, message):
398 if not hasattr(self, '_output_channel'):
399 return message
400
401 assert hasattr(self, '_output_process')
402 assert isinstance(message, compat_str)
403 line_count = message.count('\n') + 1
404 self._output_process.stdin.write((message + '\n').encode('utf-8'))
405 self._output_process.stdin.flush()
406 res = ''.join(self._output_channel.readline().decode('utf-8')
407 for _ in range(line_count))
408 return res[:-len('\n')]
409
410 def to_screen(self, message, skip_eol=False):
411 """Print message to stdout if not in quiet mode."""
412 return self.to_stdout(message, skip_eol, check_quiet=True)
413
414 def _write_string(self, s, out=None):
415 write_string(s, out=out, encoding=self.params.get('encoding'))
416
417 def to_stdout(self, message, skip_eol=False, check_quiet=False):
418 """Print message to stdout if not in quiet mode."""
419 if self.params.get('logger'):
420 self.params['logger'].debug(message)
421 elif not check_quiet or not self.params.get('quiet', False):
422 message = self._bidi_workaround(message)
423 terminator = ['\n', ''][skip_eol]
424 output = message + terminator
425
426 self._write_string(output, self._screen_file)
427
428 def to_stderr(self, message):
429 """Print message to stderr."""
430 assert isinstance(message, compat_str)
431 if self.params.get('logger'):
432 self.params['logger'].error(message)
433 else:
434 message = self._bidi_workaround(message)
435 output = message + '\n'
436 self._write_string(output, self._err_file)
437
438 def to_console_title(self, message):
439 if not self.params.get('consoletitle', False):
440 return
441 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
442 # c_wchar_p() might not be necessary if `message` is
443 # already of type unicode()
444 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
445 elif 'TERM' in os.environ:
446 self._write_string('\033]0;%s\007' % message, self._screen_file)
447
448 def save_console_title(self):
449 if not self.params.get('consoletitle', False):
450 return
451 if 'TERM' in os.environ:
452 # Save the title on stack
453 self._write_string('\033[22;0t', self._screen_file)
454
455 def restore_console_title(self):
456 if not self.params.get('consoletitle', False):
457 return
458 if 'TERM' in os.environ:
459 # Restore the title from stack
460 self._write_string('\033[23;0t', self._screen_file)
461
462 def __enter__(self):
463 self.save_console_title()
464 return self
465
466 def __exit__(self, *args):
467 self.restore_console_title()
468
469 if self.params.get('cookiefile') is not None:
470 self.cookiejar.save()
471
472 def trouble(self, message=None, tb=None):
473 """Determine action to take when a download problem appears.
474
475 Depending on if the downloader has been configured to ignore
476 download errors or not, this method may throw an exception or
477 not when errors are found, after printing the message.
478
479 tb, if given, is additional traceback information.
480 """
481 if message is not None:
482 self.to_stderr(message)
483 if self.params.get('verbose'):
484 if tb is None:
485 if sys.exc_info()[0]: # if .trouble has been called from an except block
486 tb = ''
487 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
488 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
489 tb += compat_str(traceback.format_exc())
490 else:
491 tb_data = traceback.format_list(traceback.extract_stack())
492 tb = ''.join(tb_data)
493 self.to_stderr(tb)
494 if not self.params.get('ignoreerrors', False):
495 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
496 exc_info = sys.exc_info()[1].exc_info
497 else:
498 exc_info = sys.exc_info()
499 raise DownloadError(message, exc_info)
500 self._download_retcode = 1
501
502 def report_warning(self, message):
503 '''
504 Print the message to stderr, it will be prefixed with 'WARNING:'
505 If stderr is a tty file the 'WARNING:' will be colored
506 '''
507 if self.params.get('logger') is not None:
508 self.params['logger'].warning(message)
509 else:
510 if self.params.get('no_warnings'):
511 return
512 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
513 _msg_header = '\033[0;33mWARNING:\033[0m'
514 else:
515 _msg_header = 'WARNING:'
516 warning_message = '%s %s' % (_msg_header, message)
517 self.to_stderr(warning_message)
518
519 def report_error(self, message, tb=None):
520 '''
521 Do the same as trouble, but prefixes the message with 'ERROR:', colored
522 in red if stderr is a tty file.
523 '''
524 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
525 _msg_header = '\033[0;31mERROR:\033[0m'
526 else:
527 _msg_header = 'ERROR:'
528 error_message = '%s %s' % (_msg_header, message)
529 self.trouble(error_message, tb)
530
531 def report_file_already_downloaded(self, file_name):
532 """Report file has already been fully downloaded."""
533 try:
534 self.to_screen('[download] %s has already been downloaded' % file_name)
535 except UnicodeEncodeError:
536 self.to_screen('[download] The file has already been downloaded')
537
538 def prepare_filename(self, info_dict):
539 """Generate the output filename."""
540 try:
541 template_dict = dict(info_dict)
542
543 template_dict['epoch'] = int(time.time())
544 autonumber_size = self.params.get('autonumber_size')
545 if autonumber_size is None:
546 autonumber_size = 5
547 autonumber_templ = '%0' + str(autonumber_size) + 'd'
548 template_dict['autonumber'] = autonumber_templ % self._num_downloads
549 if template_dict.get('playlist_index') is not None:
550 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
551 if template_dict.get('resolution') is None:
552 if template_dict.get('width') and template_dict.get('height'):
553 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
554 elif template_dict.get('height'):
555 template_dict['resolution'] = '%sp' % template_dict['height']
556 elif template_dict.get('width'):
557 template_dict['resolution'] = '?x%d' % template_dict['width']
558
559 sanitize = lambda k, v: sanitize_filename(
560 compat_str(v),
561 restricted=self.params.get('restrictfilenames'),
562 is_id=(k == 'id'))
563 template_dict = dict((k, sanitize(k, v))
564 for k, v in template_dict.items()
565 if v is not None)
566 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
567
568 outtmpl = sanitize_path(self.params.get('outtmpl', DEFAULT_OUTTMPL))
569 tmpl = compat_expanduser(outtmpl)
570 filename = tmpl % template_dict
571 # Temporary fix for #4787
572 # 'Treat' all problem characters by passing filename through preferredencoding
573 # to workaround encoding issues with subprocess on python2 @ Windows
574 if sys.version_info < (3, 0) and sys.platform == 'win32':
575 filename = encodeFilename(filename, True).decode(preferredencoding())
576 return filename
577 except ValueError as err:
578 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
579 return None
580
581 def _match_entry(self, info_dict, incomplete):
582 """ Returns None iff the file should be downloaded """
583
584 video_title = info_dict.get('title', info_dict.get('id', 'video'))
585 if 'title' in info_dict:
586 # This can happen when we're just evaluating the playlist
587 title = info_dict['title']
588 matchtitle = self.params.get('matchtitle', False)
589 if matchtitle:
590 if not re.search(matchtitle, title, re.IGNORECASE):
591 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
592 rejecttitle = self.params.get('rejecttitle', False)
593 if rejecttitle:
594 if re.search(rejecttitle, title, re.IGNORECASE):
595 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
596 date = info_dict.get('upload_date', None)
597 if date is not None:
598 dateRange = self.params.get('daterange', DateRange())
599 if date not in dateRange:
600 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
601 view_count = info_dict.get('view_count', None)
602 if view_count is not None:
603 min_views = self.params.get('min_views')
604 if min_views is not None and view_count < min_views:
605 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
606 max_views = self.params.get('max_views')
607 if max_views is not None and view_count > max_views:
608 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
609 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
610 return 'Skipping "%s" because it is age restricted' % video_title
611 if self.in_download_archive(info_dict):
612 return '%s has already been recorded in archive' % video_title
613
614 if not incomplete:
615 match_filter = self.params.get('match_filter')
616 if match_filter is not None:
617 ret = match_filter(info_dict)
618 if ret is not None:
619 return ret
620
621 return None
622
623 @staticmethod
624 def add_extra_info(info_dict, extra_info):
625 '''Set the keys from extra_info in info dict if they are missing'''
626 for key, value in extra_info.items():
627 info_dict.setdefault(key, value)
628
629 def extract_info(self, url, download=True, ie_key=None, extra_info={},
630 process=True):
631 '''
632 Returns a list with a dictionary for each video we find.
633 If 'download', also downloads the videos.
634 extra_info is a dict containing the extra values to add to each result
635 '''
636
637 if ie_key:
638 ies = [self.get_info_extractor(ie_key)]
639 else:
640 ies = self._ies
641
642 for ie in ies:
643 if not ie.suitable(url):
644 continue
645
646 if not ie.working():
647 self.report_warning('The program functionality for this site has been marked as broken, '
648 'and will probably not work.')
649
650 try:
651 ie_result = ie.extract(url)
652 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
653 break
654 if isinstance(ie_result, list):
655 # Backwards compatibility: old IE result format
656 ie_result = {
657 '_type': 'compat_list',
658 'entries': ie_result,
659 }
660 self.add_default_extra_info(ie_result, ie, url)
661 if process:
662 return self.process_ie_result(ie_result, download, extra_info)
663 else:
664 return ie_result
665 except ExtractorError as de: # An error we somewhat expected
666 self.report_error(compat_str(de), de.format_traceback())
667 break
668 except MaxDownloadsReached:
669 raise
670 except Exception as e:
671 if self.params.get('ignoreerrors', False):
672 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
673 break
674 else:
675 raise
676 else:
677 self.report_error('no suitable InfoExtractor for URL %s' % url)
678
679 def add_default_extra_info(self, ie_result, ie, url):
680 self.add_extra_info(ie_result, {
681 'extractor': ie.IE_NAME,
682 'webpage_url': url,
683 'webpage_url_basename': url_basename(url),
684 'extractor_key': ie.ie_key(),
685 })
686
687 def process_ie_result(self, ie_result, download=True, extra_info={}):
688 """
689 Take the result of the ie(may be modified) and resolve all unresolved
690 references (URLs, playlist items).
691
692 It will also download the videos if 'download'.
693 Returns the resolved ie_result.
694 """
695
696 result_type = ie_result.get('_type', 'video')
697
698 if result_type in ('url', 'url_transparent'):
699 extract_flat = self.params.get('extract_flat', False)
700 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
701 extract_flat is True):
702 if self.params.get('forcejson', False):
703 self.to_stdout(json.dumps(ie_result))
704 return ie_result
705
706 if result_type == 'video':
707 self.add_extra_info(ie_result, extra_info)
708 return self.process_video_result(ie_result, download=download)
709 elif result_type == 'url':
710 # We have to add extra_info to the results because it may be
711 # contained in a playlist
712 return self.extract_info(ie_result['url'],
713 download,
714 ie_key=ie_result.get('ie_key'),
715 extra_info=extra_info)
716 elif result_type == 'url_transparent':
717 # Use the information from the embedding page
718 info = self.extract_info(
719 ie_result['url'], ie_key=ie_result.get('ie_key'),
720 extra_info=extra_info, download=False, process=False)
721
722 force_properties = dict(
723 (k, v) for k, v in ie_result.items() if v is not None)
724 for f in ('_type', 'url'):
725 if f in force_properties:
726 del force_properties[f]
727 new_result = info.copy()
728 new_result.update(force_properties)
729
730 assert new_result.get('_type') != 'url_transparent'
731
732 return self.process_ie_result(
733 new_result, download=download, extra_info=extra_info)
734 elif result_type == 'playlist' or result_type == 'multi_video':
735 # We process each entry in the playlist
736 playlist = ie_result.get('title', None) or ie_result.get('id', None)
737 self.to_screen('[download] Downloading playlist: %s' % playlist)
738
739 playlist_results = []
740
741 playliststart = self.params.get('playliststart', 1) - 1
742 playlistend = self.params.get('playlistend', None)
743 # For backwards compatibility, interpret -1 as whole list
744 if playlistend == -1:
745 playlistend = None
746
747 playlistitems_str = self.params.get('playlist_items', None)
748 playlistitems = None
749 if playlistitems_str is not None:
750 def iter_playlistitems(format):
751 for string_segment in format.split(','):
752 if '-' in string_segment:
753 start, end = string_segment.split('-')
754 for item in range(int(start), int(end) + 1):
755 yield int(item)
756 else:
757 yield int(string_segment)
758 playlistitems = iter_playlistitems(playlistitems_str)
759
760 ie_entries = ie_result['entries']
761 if isinstance(ie_entries, list):
762 n_all_entries = len(ie_entries)
763 if playlistitems:
764 entries = [ie_entries[i - 1] for i in playlistitems]
765 else:
766 entries = ie_entries[playliststart:playlistend]
767 n_entries = len(entries)
768 self.to_screen(
769 "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
770 (ie_result['extractor'], playlist, n_all_entries, n_entries))
771 elif isinstance(ie_entries, PagedList):
772 if playlistitems:
773 entries = []
774 for item in playlistitems:
775 entries.extend(ie_entries.getslice(
776 item - 1, item
777 ))
778 else:
779 entries = ie_entries.getslice(
780 playliststart, playlistend)
781 n_entries = len(entries)
782 self.to_screen(
783 "[%s] playlist %s: Downloading %d videos" %
784 (ie_result['extractor'], playlist, n_entries))
785 else: # iterable
786 if playlistitems:
787 entry_list = list(ie_entries)
788 entries = [entry_list[i - 1] for i in playlistitems]
789 else:
790 entries = list(itertools.islice(
791 ie_entries, playliststart, playlistend))
792 n_entries = len(entries)
793 self.to_screen(
794 "[%s] playlist %s: Downloading %d videos" %
795 (ie_result['extractor'], playlist, n_entries))
796
797 if self.params.get('playlistreverse', False):
798 entries = entries[::-1]
799
800 for i, entry in enumerate(entries, 1):
801 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
802 extra = {
803 'n_entries': n_entries,
804 'playlist': playlist,
805 'playlist_id': ie_result.get('id'),
806 'playlist_title': ie_result.get('title'),
807 'playlist_index': i + playliststart,
808 'extractor': ie_result['extractor'],
809 'webpage_url': ie_result['webpage_url'],
810 'webpage_url_basename': url_basename(ie_result['webpage_url']),
811 'extractor_key': ie_result['extractor_key'],
812 }
813
814 reason = self._match_entry(entry, incomplete=True)
815 if reason is not None:
816 self.to_screen('[download] ' + reason)
817 continue
818
819 entry_result = self.process_ie_result(entry,
820 download=download,
821 extra_info=extra)
822 playlist_results.append(entry_result)
823 ie_result['entries'] = playlist_results
824 return ie_result
825 elif result_type == 'compat_list':
826 self.report_warning(
827 'Extractor %s returned a compat_list result. '
828 'It needs to be updated.' % ie_result.get('extractor'))
829
830 def _fixup(r):
831 self.add_extra_info(
832 r,
833 {
834 'extractor': ie_result['extractor'],
835 'webpage_url': ie_result['webpage_url'],
836 'webpage_url_basename': url_basename(ie_result['webpage_url']),
837 'extractor_key': ie_result['extractor_key'],
838 }
839 )
840 return r
841 ie_result['entries'] = [
842 self.process_ie_result(_fixup(r), download, extra_info)
843 for r in ie_result['entries']
844 ]
845 return ie_result
846 else:
847 raise Exception('Invalid result type: %s' % result_type)
848
849 def _apply_format_filter(self, format_spec, available_formats):
850 " Returns a tuple of the remaining format_spec and filtered formats "
851
852 OPERATORS = {
853 '<': operator.lt,
854 '<=': operator.le,
855 '>': operator.gt,
856 '>=': operator.ge,
857 '=': operator.eq,
858 '!=': operator.ne,
859 }
860 operator_rex = re.compile(r'''(?x)\s*\[
861 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
862 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
863 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
864 \]$
865 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
866 m = operator_rex.search(format_spec)
867 if m:
868 try:
869 comparison_value = int(m.group('value'))
870 except ValueError:
871 comparison_value = parse_filesize(m.group('value'))
872 if comparison_value is None:
873 comparison_value = parse_filesize(m.group('value') + 'B')
874 if comparison_value is None:
875 raise ValueError(
876 'Invalid value %r in format specification %r' % (
877 m.group('value'), format_spec))
878 op = OPERATORS[m.group('op')]
879
880 if not m:
881 STR_OPERATORS = {
882 '=': operator.eq,
883 '!=': operator.ne,
884 }
885 str_operator_rex = re.compile(r'''(?x)\s*\[
886 \s*(?P<key>ext|acodec|vcodec|container|protocol)
887 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
888 \s*(?P<value>[a-zA-Z0-9_-]+)
889 \s*\]$
890 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
891 m = str_operator_rex.search(format_spec)
892 if m:
893 comparison_value = m.group('value')
894 op = STR_OPERATORS[m.group('op')]
895
896 if not m:
897 raise ValueError('Invalid format specification %r' % format_spec)
898
899 def _filter(f):
900 actual_value = f.get(m.group('key'))
901 if actual_value is None:
902 return m.group('none_inclusive')
903 return op(actual_value, comparison_value)
904 new_formats = [f for f in available_formats if _filter(f)]
905
906 new_format_spec = format_spec[:-len(m.group(0))]
907 if not new_format_spec:
908 new_format_spec = 'best'
909
910 return (new_format_spec, new_formats)
911
912 def select_format(self, format_spec, available_formats):
913 while format_spec.endswith(']'):
914 format_spec, available_formats = self._apply_format_filter(
915 format_spec, available_formats)
916 if not available_formats:
917 return None
918
919 if format_spec == 'best' or format_spec is None:
920 return available_formats[-1]
921 elif format_spec == 'worst':
922 audiovideo_formats = [
923 f for f in available_formats
924 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
925 if audiovideo_formats:
926 return audiovideo_formats[0]
927 return available_formats[0]
928 elif format_spec == 'bestaudio':
929 audio_formats = [
930 f for f in available_formats
931 if f.get('vcodec') == 'none']
932 if audio_formats:
933 return audio_formats[-1]
934 elif format_spec == 'worstaudio':
935 audio_formats = [
936 f for f in available_formats
937 if f.get('vcodec') == 'none']
938 if audio_formats:
939 return audio_formats[0]
940 elif format_spec == 'bestvideo':
941 video_formats = [
942 f for f in available_formats
943 if f.get('acodec') == 'none']
944 if video_formats:
945 return video_formats[-1]
946 elif format_spec == 'worstvideo':
947 video_formats = [
948 f for f in available_formats
949 if f.get('acodec') == 'none']
950 if video_formats:
951 return video_formats[0]
952 else:
953 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
954 if format_spec in extensions:
955 filter_f = lambda f: f['ext'] == format_spec
956 else:
957 filter_f = lambda f: f['format_id'] == format_spec
958 matches = list(filter(filter_f, available_formats))
959 if matches:
960 return matches[-1]
961 return None
962
963 def _calc_headers(self, info_dict):
964 res = std_headers.copy()
965
966 add_headers = info_dict.get('http_headers')
967 if add_headers:
968 res.update(add_headers)
969
970 cookies = self._calc_cookies(info_dict)
971 if cookies:
972 res['Cookie'] = cookies
973
974 return res
975
976 def _calc_cookies(self, info_dict):
977 pr = compat_urllib_request.Request(info_dict['url'])
978 self.cookiejar.add_cookie_header(pr)
979 return pr.get_header('Cookie')
980
981 def process_video_result(self, info_dict, download=True):
982 assert info_dict.get('_type', 'video') == 'video'
983
984 if 'id' not in info_dict:
985 raise ExtractorError('Missing "id" field in extractor result')
986 if 'title' not in info_dict:
987 raise ExtractorError('Missing "title" field in extractor result')
988
989 if 'playlist' not in info_dict:
990 # It isn't part of a playlist
991 info_dict['playlist'] = None
992 info_dict['playlist_index'] = None
993
994 thumbnails = info_dict.get('thumbnails')
995 if thumbnails is None:
996 thumbnail = info_dict.get('thumbnail')
997 if thumbnail:
998 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
999 if thumbnails:
1000 thumbnails.sort(key=lambda t: (
1001 t.get('preference'), t.get('width'), t.get('height'),
1002 t.get('id'), t.get('url')))
1003 for i, t in enumerate(thumbnails):
1004 if 'width' in t and 'height' in t:
1005 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1006 if t.get('id') is None:
1007 t['id'] = '%d' % i
1008
1009 if thumbnails and 'thumbnail' not in info_dict:
1010 info_dict['thumbnail'] = thumbnails[-1]['url']
1011
1012 if 'display_id' not in info_dict and 'id' in info_dict:
1013 info_dict['display_id'] = info_dict['id']
1014
1015 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1016 # Working around negative timestamps in Windows
1017 # (see http://bugs.python.org/issue1646728)
1018 if info_dict['timestamp'] < 0 and os.name == 'nt':
1019 info_dict['timestamp'] = 0
1020 upload_date = datetime.datetime.utcfromtimestamp(
1021 info_dict['timestamp'])
1022 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1023
1024 if self.params.get('listsubtitles', False):
1025 if 'automatic_captions' in info_dict:
1026 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1027 self.list_subtitles(info_dict['id'], info_dict.get('subtitles'), 'subtitles')
1028 return
1029 info_dict['requested_subtitles'] = self.process_subtitles(
1030 info_dict['id'], info_dict.get('subtitles'),
1031 info_dict.get('automatic_captions'))
1032
1033 # This extractors handle format selection themselves
1034 if info_dict['extractor'] in ['Youku']:
1035 if download:
1036 self.process_info(info_dict)
1037 return info_dict
1038
1039 # We now pick which formats have to be downloaded
1040 if info_dict.get('formats') is None:
1041 # There's only one format available
1042 formats = [info_dict]
1043 else:
1044 formats = info_dict['formats']
1045
1046 if not formats:
1047 raise ExtractorError('No video formats found!')
1048
1049 # We check that all the formats have the format and format_id fields
1050 for i, format in enumerate(formats):
1051 if 'url' not in format:
1052 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1053
1054 if format.get('format_id') is None:
1055 format['format_id'] = compat_str(i)
1056 if format.get('format') is None:
1057 format['format'] = '{id} - {res}{note}'.format(
1058 id=format['format_id'],
1059 res=self.format_resolution(format),
1060 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1061 )
1062 # Automatically determine file extension if missing
1063 if 'ext' not in format:
1064 format['ext'] = determine_ext(format['url']).lower()
1065 # Add HTTP headers, so that external programs can use them from the
1066 # json output
1067 full_format_info = info_dict.copy()
1068 full_format_info.update(format)
1069 format['http_headers'] = self._calc_headers(full_format_info)
1070
1071 format_limit = self.params.get('format_limit', None)
1072 if format_limit:
1073 formats = list(takewhile_inclusive(
1074 lambda f: f['format_id'] != format_limit, formats
1075 ))
1076
1077 # TODO Central sorting goes here
1078
1079 if formats[0] is not info_dict:
1080 # only set the 'formats' fields if the original info_dict list them
1081 # otherwise we end up with a circular reference, the first (and unique)
1082 # element in the 'formats' field in info_dict is info_dict itself,
1083 # wich can't be exported to json
1084 info_dict['formats'] = formats
1085 if self.params.get('listformats'):
1086 self.list_formats(info_dict)
1087 return
1088 if self.params.get('list_thumbnails'):
1089 self.list_thumbnails(info_dict)
1090 return
1091
1092 req_format = self.params.get('format')
1093 if req_format is None:
1094 req_format = 'best'
1095 formats_to_download = []
1096 if req_format == 'all':
1097 formats_to_download = formats
1098 else:
1099 for rfstr in req_format.split(','):
1100 # We can accept formats requested in the format: 34/5/best, we pick
1101 # the first that is available, starting from left
1102 req_formats = rfstr.split('/')
1103 for rf in req_formats:
1104 if re.match(r'.+?\+.+?', rf) is not None:
1105 # Two formats have been requested like '137+139'
1106 format_1, format_2 = rf.split('+')
1107 formats_info = (self.select_format(format_1, formats),
1108 self.select_format(format_2, formats))
1109 if all(formats_info):
1110 # The first format must contain the video and the
1111 # second the audio
1112 if formats_info[0].get('vcodec') == 'none':
1113 self.report_error('The first format must '
1114 'contain the video, try using '
1115 '"-f %s+%s"' % (format_2, format_1))
1116 return
1117 output_ext = (
1118 formats_info[0]['ext']
1119 if self.params.get('merge_output_format') is None
1120 else self.params['merge_output_format'])
1121 selected_format = {
1122 'requested_formats': formats_info,
1123 'format': '%s+%s' % (formats_info[0].get('format'),
1124 formats_info[1].get('format')),
1125 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1126 formats_info[1].get('format_id')),
1127 'width': formats_info[0].get('width'),
1128 'height': formats_info[0].get('height'),
1129 'resolution': formats_info[0].get('resolution'),
1130 'fps': formats_info[0].get('fps'),
1131 'vcodec': formats_info[0].get('vcodec'),
1132 'vbr': formats_info[0].get('vbr'),
1133 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1134 'acodec': formats_info[1].get('acodec'),
1135 'abr': formats_info[1].get('abr'),
1136 'ext': output_ext,
1137 }
1138 else:
1139 selected_format = None
1140 else:
1141 selected_format = self.select_format(rf, formats)
1142 if selected_format is not None:
1143 formats_to_download.append(selected_format)
1144 break
1145 if not formats_to_download:
1146 raise ExtractorError('requested format not available',
1147 expected=True)
1148
1149 if download:
1150 if len(formats_to_download) > 1:
1151 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1152 for format in formats_to_download:
1153 new_info = dict(info_dict)
1154 new_info.update(format)
1155 self.process_info(new_info)
1156 # We update the info dict with the best quality format (backwards compatibility)
1157 info_dict.update(formats_to_download[-1])
1158 return info_dict
1159
1160 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1161 """Select the requested subtitles and their format"""
1162 available_subs = {}
1163 if normal_subtitles and self.params.get('writesubtitles'):
1164 available_subs.update(normal_subtitles)
1165 if automatic_captions and self.params.get('writeautomaticsub'):
1166 for lang, cap_info in automatic_captions.items():
1167 if lang not in available_subs:
1168 available_subs[lang] = cap_info
1169
1170 if (not self.params.get('writesubtitles') and not
1171 self.params.get('writeautomaticsub') or not
1172 available_subs):
1173 return None
1174
1175 if self.params.get('allsubtitles', False):
1176 requested_langs = available_subs.keys()
1177 else:
1178 if self.params.get('subtitleslangs', False):
1179 requested_langs = self.params.get('subtitleslangs')
1180 elif 'en' in available_subs:
1181 requested_langs = ['en']
1182 else:
1183 requested_langs = [list(available_subs.keys())[0]]
1184
1185 formats_query = self.params.get('subtitlesformat', 'best')
1186 formats_preference = formats_query.split('/') if formats_query else []
1187 subs = {}
1188 for lang in requested_langs:
1189 formats = available_subs.get(lang)
1190 if formats is None:
1191 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1192 continue
1193 for ext in formats_preference:
1194 if ext == 'best':
1195 f = formats[-1]
1196 break
1197 matches = list(filter(lambda f: f['ext'] == ext, formats))
1198 if matches:
1199 f = matches[-1]
1200 break
1201 else:
1202 f = formats[-1]
1203 self.report_warning(
1204 'No subtitle format found matching "%s" for language %s, '
1205 'using %s' % (formats_query, lang, f['ext']))
1206 subs[lang] = f
1207 return subs
1208
1209 def process_info(self, info_dict):
1210 """Process a single resolved IE result."""
1211
1212 assert info_dict.get('_type', 'video') == 'video'
1213
1214 max_downloads = self.params.get('max_downloads')
1215 if max_downloads is not None:
1216 if self._num_downloads >= int(max_downloads):
1217 raise MaxDownloadsReached()
1218
1219 info_dict['fulltitle'] = info_dict['title']
1220 if len(info_dict['title']) > 200:
1221 info_dict['title'] = info_dict['title'][:197] + '...'
1222
1223 if 'format' not in info_dict:
1224 info_dict['format'] = info_dict['ext']
1225
1226 reason = self._match_entry(info_dict, incomplete=False)
1227 if reason is not None:
1228 self.to_screen('[download] ' + reason)
1229 return
1230
1231 self._num_downloads += 1
1232
1233 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1234
1235 # Forced printings
1236 if self.params.get('forcetitle', False):
1237 self.to_stdout(info_dict['fulltitle'])
1238 if self.params.get('forceid', False):
1239 self.to_stdout(info_dict['id'])
1240 if self.params.get('forceurl', False):
1241 if info_dict.get('requested_formats') is not None:
1242 for f in info_dict['requested_formats']:
1243 self.to_stdout(f['url'] + f.get('play_path', ''))
1244 else:
1245 # For RTMP URLs, also include the playpath
1246 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1247 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1248 self.to_stdout(info_dict['thumbnail'])
1249 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1250 self.to_stdout(info_dict['description'])
1251 if self.params.get('forcefilename', False) and filename is not None:
1252 self.to_stdout(filename)
1253 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1254 self.to_stdout(formatSeconds(info_dict['duration']))
1255 if self.params.get('forceformat', False):
1256 self.to_stdout(info_dict['format'])
1257 if self.params.get('forcejson', False):
1258 self.to_stdout(json.dumps(info_dict))
1259
1260 # Do nothing else if in simulate mode
1261 if self.params.get('simulate', False):
1262 return
1263
1264 if filename is None:
1265 return
1266
1267 try:
1268 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1269 if dn and not os.path.exists(dn):
1270 os.makedirs(dn)
1271 except (OSError, IOError) as err:
1272 self.report_error('unable to create directory ' + compat_str(err))
1273 return
1274
1275 if self.params.get('writedescription', False):
1276 descfn = filename + '.description'
1277 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1278 self.to_screen('[info] Video description is already present')
1279 elif info_dict.get('description') is None:
1280 self.report_warning('There\'s no description to write.')
1281 else:
1282 try:
1283 self.to_screen('[info] Writing video description to: ' + descfn)
1284 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1285 descfile.write(info_dict['description'])
1286 except (OSError, IOError):
1287 self.report_error('Cannot write description file ' + descfn)
1288 return
1289
1290 if self.params.get('writeannotations', False):
1291 annofn = filename + '.annotations.xml'
1292 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1293 self.to_screen('[info] Video annotations are already present')
1294 else:
1295 try:
1296 self.to_screen('[info] Writing video annotations to: ' + annofn)
1297 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1298 annofile.write(info_dict['annotations'])
1299 except (KeyError, TypeError):
1300 self.report_warning('There are no annotations to write.')
1301 except (OSError, IOError):
1302 self.report_error('Cannot write annotations file: ' + annofn)
1303 return
1304
1305 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1306 self.params.get('writeautomaticsub')])
1307
1308 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1309 # subtitles download errors are already managed as troubles in relevant IE
1310 # that way it will silently go on when used with unsupporting IE
1311 subtitles = info_dict['requested_subtitles']
1312 ie = self.get_info_extractor(info_dict['extractor_key'])
1313 for sub_lang, sub_info in subtitles.items():
1314 sub_format = sub_info['ext']
1315 if sub_info.get('data') is not None:
1316 sub_data = sub_info['data']
1317 else:
1318 try:
1319 sub_data = ie._download_webpage(
1320 sub_info['url'], info_dict['id'], note=False)
1321 except ExtractorError as err:
1322 self.report_warning('Unable to download subtitle for "%s": %s' %
1323 (sub_lang, compat_str(err.cause)))
1324 continue
1325 try:
1326 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1327 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1328 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1329 else:
1330 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1331 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1332 subfile.write(sub_data)
1333 except (OSError, IOError):
1334 self.report_error('Cannot write subtitles file ' + sub_filename)
1335 return
1336
1337 if self.params.get('writeinfojson', False):
1338 infofn = os.path.splitext(filename)[0] + '.info.json'
1339 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1340 self.to_screen('[info] Video description metadata is already present')
1341 else:
1342 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1343 try:
1344 write_json_file(info_dict, infofn)
1345 except (OSError, IOError):
1346 self.report_error('Cannot write metadata to JSON file ' + infofn)
1347 return
1348
1349 self._write_thumbnails(info_dict, filename)
1350
1351 if not self.params.get('skip_download', False):
1352 try:
1353 def dl(name, info):
1354 fd = get_suitable_downloader(info, self.params)(self, self.params)
1355 for ph in self._progress_hooks:
1356 fd.add_progress_hook(ph)
1357 if self.params.get('verbose'):
1358 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1359 return fd.download(name, info)
1360
1361 if info_dict.get('requested_formats') is not None:
1362 downloaded = []
1363 success = True
1364 merger = FFmpegMergerPP(self)
1365 if not merger.available:
1366 postprocessors = []
1367 self.report_warning('You have requested multiple '
1368 'formats but ffmpeg or avconv are not installed.'
1369 ' The formats won\'t be merged')
1370 else:
1371 postprocessors = [merger]
1372 for f in info_dict['requested_formats']:
1373 new_info = dict(info_dict)
1374 new_info.update(f)
1375 fname = self.prepare_filename(new_info)
1376 fname = prepend_extension(fname, 'f%s' % f['format_id'])
1377 downloaded.append(fname)
1378 partial_success = dl(fname, new_info)
1379 success = success and partial_success
1380 info_dict['__postprocessors'] = postprocessors
1381 info_dict['__files_to_merge'] = downloaded
1382 else:
1383 # Just a single file
1384 success = dl(filename, info_dict)
1385 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1386 self.report_error('unable to download video data: %s' % str(err))
1387 return
1388 except (OSError, IOError) as err:
1389 raise UnavailableVideoError(err)
1390 except (ContentTooShortError, ) as err:
1391 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1392 return
1393
1394 if success:
1395 # Fixup content
1396 fixup_policy = self.params.get('fixup')
1397 if fixup_policy is None:
1398 fixup_policy = 'detect_or_warn'
1399
1400 stretched_ratio = info_dict.get('stretched_ratio')
1401 if stretched_ratio is not None and stretched_ratio != 1:
1402 if fixup_policy == 'warn':
1403 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1404 info_dict['id'], stretched_ratio))
1405 elif fixup_policy == 'detect_or_warn':
1406 stretched_pp = FFmpegFixupStretchedPP(self)
1407 if stretched_pp.available:
1408 info_dict.setdefault('__postprocessors', [])
1409 info_dict['__postprocessors'].append(stretched_pp)
1410 else:
1411 self.report_warning(
1412 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1413 info_dict['id'], stretched_ratio))
1414 else:
1415 assert fixup_policy in ('ignore', 'never')
1416
1417 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1418 if fixup_policy == 'warn':
1419 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1420 info_dict['id']))
1421 elif fixup_policy == 'detect_or_warn':
1422 fixup_pp = FFmpegFixupM4aPP(self)
1423 if fixup_pp.available:
1424 info_dict.setdefault('__postprocessors', [])
1425 info_dict['__postprocessors'].append(fixup_pp)
1426 else:
1427 self.report_warning(
1428 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1429 info_dict['id']))
1430 else:
1431 assert fixup_policy in ('ignore', 'never')
1432
1433 try:
1434 self.post_process(filename, info_dict)
1435 except (PostProcessingError) as err:
1436 self.report_error('postprocessing: %s' % str(err))
1437 return
1438 self.record_download_archive(info_dict)
1439
1440 def download(self, url_list):
1441 """Download a given list of URLs."""
1442 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1443 if (len(url_list) > 1 and
1444 '%' not in outtmpl and
1445 self.params.get('max_downloads') != 1):
1446 raise SameFileError(outtmpl)
1447
1448 for url in url_list:
1449 try:
1450 # It also downloads the videos
1451 res = self.extract_info(url)
1452 except UnavailableVideoError:
1453 self.report_error('unable to download video')
1454 except MaxDownloadsReached:
1455 self.to_screen('[info] Maximum number of downloaded files reached.')
1456 raise
1457 else:
1458 if self.params.get('dump_single_json', False):
1459 self.to_stdout(json.dumps(res))
1460
1461 return self._download_retcode
1462
1463 def download_with_info_file(self, info_filename):
1464 with contextlib.closing(fileinput.FileInput(
1465 [info_filename], mode='r',
1466 openhook=fileinput.hook_encoded('utf-8'))) as f:
1467 # FileInput doesn't have a read method, we can't call json.load
1468 info = json.loads('\n'.join(f))
1469 try:
1470 self.process_ie_result(info, download=True)
1471 except DownloadError:
1472 webpage_url = info.get('webpage_url')
1473 if webpage_url is not None:
1474 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1475 return self.download([webpage_url])
1476 else:
1477 raise
1478 return self._download_retcode
1479
1480 def post_process(self, filename, ie_info):
1481 """Run all the postprocessors on the given file."""
1482 info = dict(ie_info)
1483 info['filepath'] = filename
1484 pps_chain = []
1485 if ie_info.get('__postprocessors') is not None:
1486 pps_chain.extend(ie_info['__postprocessors'])
1487 pps_chain.extend(self._pps)
1488 for pp in pps_chain:
1489 old_filename = info['filepath']
1490 try:
1491 files_to_delete, info = pp.run(info)
1492 except PostProcessingError as e:
1493 self.report_error(e.msg)
1494 if files_to_delete and not self.params.get('keepvideo', False):
1495 for old_filename in files_to_delete:
1496 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1497 try:
1498 os.remove(encodeFilename(old_filename))
1499 except (IOError, OSError):
1500 self.report_warning('Unable to remove downloaded original file')
1501
1502 def _make_archive_id(self, info_dict):
1503 # Future-proof against any change in case
1504 # and backwards compatibility with prior versions
1505 extractor = info_dict.get('extractor_key')
1506 if extractor is None:
1507 if 'id' in info_dict:
1508 extractor = info_dict.get('ie_key') # key in a playlist
1509 if extractor is None:
1510 return None # Incomplete video information
1511 return extractor.lower() + ' ' + info_dict['id']
1512
1513 def in_download_archive(self, info_dict):
1514 fn = self.params.get('download_archive')
1515 if fn is None:
1516 return False
1517
1518 vid_id = self._make_archive_id(info_dict)
1519 if vid_id is None:
1520 return False # Incomplete video information
1521
1522 try:
1523 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1524 for line in archive_file:
1525 if line.strip() == vid_id:
1526 return True
1527 except IOError as ioe:
1528 if ioe.errno != errno.ENOENT:
1529 raise
1530 return False
1531
1532 def record_download_archive(self, info_dict):
1533 fn = self.params.get('download_archive')
1534 if fn is None:
1535 return
1536 vid_id = self._make_archive_id(info_dict)
1537 assert vid_id
1538 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1539 archive_file.write(vid_id + '\n')
1540
1541 @staticmethod
1542 def format_resolution(format, default='unknown'):
1543 if format.get('vcodec') == 'none':
1544 return 'audio only'
1545 if format.get('resolution') is not None:
1546 return format['resolution']
1547 if format.get('height') is not None:
1548 if format.get('width') is not None:
1549 res = '%sx%s' % (format['width'], format['height'])
1550 else:
1551 res = '%sp' % format['height']
1552 elif format.get('width') is not None:
1553 res = '?x%d' % format['width']
1554 else:
1555 res = default
1556 return res
1557
1558 def _format_note(self, fdict):
1559 res = ''
1560 if fdict.get('ext') in ['f4f', 'f4m']:
1561 res += '(unsupported) '
1562 if fdict.get('format_note') is not None:
1563 res += fdict['format_note'] + ' '
1564 if fdict.get('tbr') is not None:
1565 res += '%4dk ' % fdict['tbr']
1566 if fdict.get('container') is not None:
1567 if res:
1568 res += ', '
1569 res += '%s container' % fdict['container']
1570 if (fdict.get('vcodec') is not None and
1571 fdict.get('vcodec') != 'none'):
1572 if res:
1573 res += ', '
1574 res += fdict['vcodec']
1575 if fdict.get('vbr') is not None:
1576 res += '@'
1577 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1578 res += 'video@'
1579 if fdict.get('vbr') is not None:
1580 res += '%4dk' % fdict['vbr']
1581 if fdict.get('fps') is not None:
1582 res += ', %sfps' % fdict['fps']
1583 if fdict.get('acodec') is not None:
1584 if res:
1585 res += ', '
1586 if fdict['acodec'] == 'none':
1587 res += 'video only'
1588 else:
1589 res += '%-5s' % fdict['acodec']
1590 elif fdict.get('abr') is not None:
1591 if res:
1592 res += ', '
1593 res += 'audio'
1594 if fdict.get('abr') is not None:
1595 res += '@%3dk' % fdict['abr']
1596 if fdict.get('asr') is not None:
1597 res += ' (%5dHz)' % fdict['asr']
1598 if fdict.get('filesize') is not None:
1599 if res:
1600 res += ', '
1601 res += format_bytes(fdict['filesize'])
1602 elif fdict.get('filesize_approx') is not None:
1603 if res:
1604 res += ', '
1605 res += '~' + format_bytes(fdict['filesize_approx'])
1606 return res
1607
1608 def list_formats(self, info_dict):
1609 formats = info_dict.get('formats', [info_dict])
1610 table = [
1611 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1612 for f in formats
1613 if f.get('preference') is None or f['preference'] >= -1000]
1614 if len(formats) > 1:
1615 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1616
1617 header_line = ['format code', 'extension', 'resolution', 'note']
1618 self.to_screen(
1619 '[info] Available formats for %s:\n%s' %
1620 (info_dict['id'], render_table(header_line, table)))
1621
1622 def list_thumbnails(self, info_dict):
1623 thumbnails = info_dict.get('thumbnails')
1624 if not thumbnails:
1625 tn_url = info_dict.get('thumbnail')
1626 if tn_url:
1627 thumbnails = [{'id': '0', 'url': tn_url}]
1628 else:
1629 self.to_screen(
1630 '[info] No thumbnails present for %s' % info_dict['id'])
1631 return
1632
1633 self.to_screen(
1634 '[info] Thumbnails for %s:' % info_dict['id'])
1635 self.to_screen(render_table(
1636 ['ID', 'width', 'height', 'URL'],
1637 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1638
1639 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1640 if not subtitles:
1641 self.to_screen('%s has no %s' % (video_id, name))
1642 return
1643 self.to_screen(
1644 'Available %s for %s:' % (name, video_id))
1645 self.to_screen(render_table(
1646 ['Language', 'formats'],
1647 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1648 for lang, formats in subtitles.items()]))
1649
1650 def urlopen(self, req):
1651 """ Start an HTTP download """
1652
1653 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1654 # always respected by websites, some tend to give out URLs with non percent-encoded
1655 # non-ASCII characters (see telemb.py, ard.py [#3412])
1656 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1657 # To work around aforementioned issue we will replace request's original URL with
1658 # percent-encoded one
1659 req_is_string = isinstance(req, compat_basestring)
1660 url = req if req_is_string else req.get_full_url()
1661 url_escaped = escape_url(url)
1662
1663 # Substitute URL if any change after escaping
1664 if url != url_escaped:
1665 if req_is_string:
1666 req = url_escaped
1667 else:
1668 req = compat_urllib_request.Request(
1669 url_escaped, data=req.data, headers=req.headers,
1670 origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
1671
1672 return self._opener.open(req, timeout=self._socket_timeout)
1673
1674 def print_debug_header(self):
1675 if not self.params.get('verbose'):
1676 return
1677
1678 if type('') is not compat_str:
1679 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1680 self.report_warning(
1681 'Your Python is broken! Update to a newer and supported version')
1682
1683 stdout_encoding = getattr(
1684 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1685 encoding_str = (
1686 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1687 locale.getpreferredencoding(),
1688 sys.getfilesystemencoding(),
1689 stdout_encoding,
1690 self.get_encoding()))
1691 write_string(encoding_str, encoding=None)
1692
1693 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1694 try:
1695 sp = subprocess.Popen(
1696 ['git', 'rev-parse', '--short', 'HEAD'],
1697 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1698 cwd=os.path.dirname(os.path.abspath(__file__)))
1699 out, err = sp.communicate()
1700 out = out.decode().strip()
1701 if re.match('[0-9a-f]+', out):
1702 self._write_string('[debug] Git HEAD: ' + out + '\n')
1703 except Exception:
1704 try:
1705 sys.exc_clear()
1706 except Exception:
1707 pass
1708 self._write_string('[debug] Python version %s - %s\n' % (
1709 platform.python_version(), platform_name()))
1710
1711 exe_versions = FFmpegPostProcessor.get_versions(self)
1712 exe_versions['rtmpdump'] = rtmpdump_version()
1713 exe_str = ', '.join(
1714 '%s %s' % (exe, v)
1715 for exe, v in sorted(exe_versions.items())
1716 if v
1717 )
1718 if not exe_str:
1719 exe_str = 'none'
1720 self._write_string('[debug] exe versions: %s\n' % exe_str)
1721
1722 proxy_map = {}
1723 for handler in self._opener.handlers:
1724 if hasattr(handler, 'proxies'):
1725 proxy_map.update(handler.proxies)
1726 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1727
1728 if self.params.get('call_home', False):
1729 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1730 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1731 latest_version = self.urlopen(
1732 'https://yt-dl.org/latest/version').read().decode('utf-8')
1733 if version_tuple(latest_version) > version_tuple(__version__):
1734 self.report_warning(
1735 'You are using an outdated version (newest version: %s)! '
1736 'See https://yt-dl.org/update if you need help updating.' %
1737 latest_version)
1738
1739 def _setup_opener(self):
1740 timeout_val = self.params.get('socket_timeout')
1741 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1742
1743 opts_cookiefile = self.params.get('cookiefile')
1744 opts_proxy = self.params.get('proxy')
1745
1746 if opts_cookiefile is None:
1747 self.cookiejar = compat_cookiejar.CookieJar()
1748 else:
1749 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1750 opts_cookiefile)
1751 if os.access(opts_cookiefile, os.R_OK):
1752 self.cookiejar.load()
1753
1754 cookie_processor = compat_urllib_request.HTTPCookieProcessor(
1755 self.cookiejar)
1756 if opts_proxy is not None:
1757 if opts_proxy == '':
1758 proxies = {}
1759 else:
1760 proxies = {'http': opts_proxy, 'https': opts_proxy}
1761 else:
1762 proxies = compat_urllib_request.getproxies()
1763 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1764 if 'http' in proxies and 'https' not in proxies:
1765 proxies['https'] = proxies['http']
1766 proxy_handler = PerRequestProxyHandler(proxies)
1767
1768 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
1769 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
1770 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
1771 opener = compat_urllib_request.build_opener(
1772 proxy_handler, https_handler, cookie_processor, ydlh)
1773
1774 # Delete the default user-agent header, which would otherwise apply in
1775 # cases where our custom HTTP handler doesn't come into play
1776 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
1777 opener.addheaders = []
1778 self._opener = opener
1779
1780 def encode(self, s):
1781 if isinstance(s, bytes):
1782 return s # Already encoded
1783
1784 try:
1785 return s.encode(self.get_encoding())
1786 except UnicodeEncodeError as err:
1787 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
1788 raise
1789
1790 def get_encoding(self):
1791 encoding = self.params.get('encoding')
1792 if encoding is None:
1793 encoding = preferredencoding()
1794 return encoding
1795
1796 def _write_thumbnails(self, info_dict, filename):
1797 if self.params.get('writethumbnail', False):
1798 thumbnails = info_dict.get('thumbnails')
1799 if thumbnails:
1800 thumbnails = [thumbnails[-1]]
1801 elif self.params.get('write_all_thumbnails', False):
1802 thumbnails = info_dict.get('thumbnails')
1803 else:
1804 return
1805
1806 if not thumbnails:
1807 # No thumbnails present, so return immediately
1808 return
1809
1810 for t in thumbnails:
1811 thumb_ext = determine_ext(t['url'], 'jpg')
1812 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
1813 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
1814 thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
1815
1816 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
1817 self.to_screen('[%s] %s: Thumbnail %sis already present' %
1818 (info_dict['extractor'], info_dict['id'], thumb_display_id))
1819 else:
1820 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
1821 (info_dict['extractor'], info_dict['id'], thumb_display_id))
1822 try:
1823 uf = self.urlopen(t['url'])
1824 with open(thumb_filename, 'wb') as thumbf:
1825 shutil.copyfileobj(uf, thumbf)
1826 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
1827 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
1828 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1829 self.report_warning('Unable to download thumbnail "%s": %s' %
1830 (t['url'], compat_str(err)))