]> jfr.im git - yt-dlp.git/blob - youtube_dl/YoutubeDL.py
[YoutubeDL] Introduce CSS3 like string operators
[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 tokenize
25 import traceback
26
27 if os.name == 'nt':
28 import ctypes
29
30 from .compat import (
31 compat_basestring,
32 compat_cookiejar,
33 compat_expanduser,
34 compat_get_terminal_size,
35 compat_http_client,
36 compat_kwargs,
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 ContentTooShortError,
45 date_from_str,
46 DateRange,
47 DEFAULT_OUTTMPL,
48 determine_ext,
49 DownloadError,
50 encode_compat_str,
51 encodeFilename,
52 error_to_compat_str,
53 ExtractorError,
54 format_bytes,
55 formatSeconds,
56 locked_file,
57 make_HTTPS_handler,
58 MaxDownloadsReached,
59 PagedList,
60 parse_filesize,
61 PerRequestProxyHandler,
62 PostProcessingError,
63 platform_name,
64 preferredencoding,
65 render_table,
66 SameFileError,
67 sanitize_filename,
68 sanitize_path,
69 sanitized_Request,
70 std_headers,
71 subtitles_filename,
72 UnavailableVideoError,
73 url_basename,
74 version_tuple,
75 write_json_file,
76 write_string,
77 YoutubeDLCookieProcessor,
78 YoutubeDLHandler,
79 prepend_extension,
80 replace_extension,
81 args_to_str,
82 age_restricted,
83 )
84 from .cache import Cache
85 from .extractor import get_info_extractor, gen_extractors
86 from .downloader import get_suitable_downloader
87 from .downloader.rtmp import rtmpdump_version
88 from .postprocessor import (
89 FFmpegFixupM4aPP,
90 FFmpegFixupStretchedPP,
91 FFmpegMergerPP,
92 FFmpegPostProcessor,
93 get_postprocessor,
94 )
95 from .version import __version__
96
97
98 class YoutubeDL(object):
99 """YoutubeDL class.
100
101 YoutubeDL objects are the ones responsible of downloading the
102 actual video file and writing it to disk if the user has requested
103 it, among some other tasks. In most cases there should be one per
104 program. As, given a video URL, the downloader doesn't know how to
105 extract all the needed information, task that InfoExtractors do, it
106 has to pass the URL to one of them.
107
108 For this, YoutubeDL objects have a method that allows
109 InfoExtractors to be registered in a given order. When it is passed
110 a URL, the YoutubeDL object handles it to the first InfoExtractor it
111 finds that reports being able to handle it. The InfoExtractor extracts
112 all the information about the video or videos the URL refers to, and
113 YoutubeDL process the extracted information, possibly using a File
114 Downloader to download the video.
115
116 YoutubeDL objects accept a lot of parameters. In order not to saturate
117 the object constructor with arguments, it receives a dictionary of
118 options instead. These options are available through the params
119 attribute for the InfoExtractors to use. The YoutubeDL also
120 registers itself as the downloader in charge for the InfoExtractors
121 that are added to it, so this is a "mutual registration".
122
123 Available options:
124
125 username: Username for authentication purposes.
126 password: Password for authentication purposes.
127 videopassword: Password for accessing a video.
128 usenetrc: Use netrc for authentication instead.
129 verbose: Print additional info to stdout.
130 quiet: Do not print messages to stdout.
131 no_warnings: Do not print out anything for warnings.
132 forceurl: Force printing final URL.
133 forcetitle: Force printing title.
134 forceid: Force printing ID.
135 forcethumbnail: Force printing thumbnail URL.
136 forcedescription: Force printing description.
137 forcefilename: Force printing final filename.
138 forceduration: Force printing duration.
139 forcejson: Force printing info_dict as JSON.
140 dump_single_json: Force printing the info_dict of the whole playlist
141 (or video) as a single JSON line.
142 simulate: Do not download the video files.
143 format: Video format code. See options.py for more information.
144 outtmpl: Template for output names.
145 restrictfilenames: Do not allow "&" and spaces in file names
146 ignoreerrors: Do not stop on download errors.
147 force_generic_extractor: Force downloader to use the generic extractor
148 nooverwrites: Prevent overwriting files.
149 playliststart: Playlist item to start at.
150 playlistend: Playlist item to end at.
151 playlist_items: Specific indices of playlist to download.
152 playlistreverse: Download playlist items in reverse order.
153 matchtitle: Download only matching titles.
154 rejecttitle: Reject downloads for matching titles.
155 logger: Log messages to a logging.Logger instance.
156 logtostderr: Log messages to stderr instead of stdout.
157 writedescription: Write the video description to a .description file
158 writeinfojson: Write the video description to a .info.json file
159 writeannotations: Write the video annotations to a .annotations.xml file
160 writethumbnail: Write the thumbnail image to a file
161 write_all_thumbnails: Write all thumbnail formats to files
162 writesubtitles: Write the video subtitles to a file
163 writeautomaticsub: Write the automatically generated subtitles to a file
164 allsubtitles: Downloads all the subtitles of the video
165 (requires writesubtitles or writeautomaticsub)
166 listsubtitles: Lists all available subtitles for the video
167 subtitlesformat: The format code for subtitles
168 subtitleslangs: List of languages of the subtitles to download
169 keepvideo: Keep the video file after post-processing
170 daterange: A DateRange object, download only if the upload_date is in the range.
171 skip_download: Skip the actual download of the video file
172 cachedir: Location of the cache files in the filesystem.
173 False to disable filesystem cache.
174 noplaylist: Download single video instead of a playlist if in doubt.
175 age_limit: An integer representing the user's age in years.
176 Unsuitable videos for the given age are skipped.
177 min_views: An integer representing the minimum view count the video
178 must have in order to not be skipped.
179 Videos without view count information are always
180 downloaded. None for no limit.
181 max_views: An integer representing the maximum view count.
182 Videos that are more popular than that are not
183 downloaded.
184 Videos without view count information are always
185 downloaded. None for no limit.
186 download_archive: File name of a file where all downloads are recorded.
187 Videos already present in the file are not downloaded
188 again.
189 cookiefile: File name where cookies should be read from and dumped to.
190 nocheckcertificate:Do not verify SSL certificates
191 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
192 At the moment, this is only supported by YouTube.
193 proxy: URL of the proxy server to use
194 cn_verification_proxy: URL of the proxy to use for IP address verification
195 on Chinese sites. (Experimental)
196 socket_timeout: Time to wait for unresponsive hosts, in seconds
197 bidi_workaround: Work around buggy terminals without bidirectional text
198 support, using fridibi
199 debug_printtraffic:Print out sent and received HTTP traffic
200 include_ads: Download ads as well
201 default_search: Prepend this string if an input url is not valid.
202 'auto' for elaborate guessing
203 encoding: Use this encoding instead of the system-specified.
204 extract_flat: Do not resolve URLs, return the immediate result.
205 Pass in 'in_playlist' to only show this behavior for
206 playlist items.
207 postprocessors: A list of dictionaries, each with an entry
208 * key: The name of the postprocessor. See
209 youtube_dl/postprocessor/__init__.py for a list.
210 as well as any further keyword arguments for the
211 postprocessor.
212 progress_hooks: A list of functions that get called on download
213 progress, with a dictionary with the entries
214 * status: One of "downloading", "error", or "finished".
215 Check this first and ignore unknown values.
216
217 If status is one of "downloading", or "finished", the
218 following properties may also be present:
219 * filename: The final filename (always present)
220 * tmpfilename: The filename we're currently writing to
221 * downloaded_bytes: Bytes on disk
222 * total_bytes: Size of the whole file, None if unknown
223 * total_bytes_estimate: Guess of the eventual file size,
224 None if unavailable.
225 * elapsed: The number of seconds since download started.
226 * eta: The estimated time in seconds, None if unknown
227 * speed: The download speed in bytes/second, None if
228 unknown
229 * fragment_index: The counter of the currently
230 downloaded video fragment.
231 * fragment_count: The number of fragments (= individual
232 files that will be merged)
233
234 Progress hooks are guaranteed to be called at least once
235 (with status "finished") if the download is successful.
236 merge_output_format: Extension to use when merging formats.
237 fixup: Automatically correct known faults of the file.
238 One of:
239 - "never": do nothing
240 - "warn": only emit a warning
241 - "detect_or_warn": check whether we can do anything
242 about it, warn otherwise (default)
243 source_address: (Experimental) Client-side IP address to bind to.
244 call_home: Boolean, true iff we are allowed to contact the
245 youtube-dl servers for debugging.
246 sleep_interval: Number of seconds to sleep before each download.
247 listformats: Print an overview of available video formats and exit.
248 list_thumbnails: Print a table of all thumbnails and exit.
249 match_filter: A function that gets called with the info_dict of
250 every video.
251 If it returns a message, the video is ignored.
252 If it returns None, the video is downloaded.
253 match_filter_func in utils.py is one example for this.
254 no_color: Do not emit color codes in output.
255
256 The following options determine which downloader is picked:
257 external_downloader: Executable of the external downloader to call.
258 None or unset for standard (built-in) downloader.
259 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv.
260
261 The following parameters are not used by YoutubeDL itself, they are used by
262 the downloader (see youtube_dl/downloader/common.py):
263 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
264 noresizebuffer, retries, continuedl, noprogress, consoletitle,
265 xattr_set_filesize, external_downloader_args.
266
267 The following options are used by the post processors:
268 prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available,
269 otherwise prefer avconv.
270 postprocessor_args: A list of additional command-line arguments for the
271 postprocessor.
272 """
273
274 params = None
275 _ies = []
276 _pps = []
277 _download_retcode = None
278 _num_downloads = None
279 _screen_file = None
280
281 def __init__(self, params=None, auto_init=True):
282 """Create a FileDownloader object with the given options."""
283 if params is None:
284 params = {}
285 self._ies = []
286 self._ies_instances = {}
287 self._pps = []
288 self._progress_hooks = []
289 self._download_retcode = 0
290 self._num_downloads = 0
291 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
292 self._err_file = sys.stderr
293 self.params = {
294 # Default parameters
295 'nocheckcertificate': False,
296 }
297 self.params.update(params)
298 self.cache = Cache(self)
299
300 if params.get('bidi_workaround', False):
301 try:
302 import pty
303 master, slave = pty.openpty()
304 width = compat_get_terminal_size().columns
305 if width is None:
306 width_args = []
307 else:
308 width_args = ['-w', str(width)]
309 sp_kwargs = dict(
310 stdin=subprocess.PIPE,
311 stdout=slave,
312 stderr=self._err_file)
313 try:
314 self._output_process = subprocess.Popen(
315 ['bidiv'] + width_args, **sp_kwargs
316 )
317 except OSError:
318 self._output_process = subprocess.Popen(
319 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
320 self._output_channel = os.fdopen(master, 'rb')
321 except OSError as ose:
322 if ose.errno == 2:
323 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 $PATH.')
324 else:
325 raise
326
327 if (sys.version_info >= (3,) and sys.platform != 'win32' and
328 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
329 not params.get('restrictfilenames', False)):
330 # On Python 3, the Unicode filesystem API will throw errors (#1474)
331 self.report_warning(
332 'Assuming --restrict-filenames since file system encoding '
333 'cannot encode all characters. '
334 'Set the LC_ALL environment variable to fix this.')
335 self.params['restrictfilenames'] = True
336
337 if isinstance(params.get('outtmpl'), bytes):
338 self.report_warning(
339 'Parameter outtmpl is bytes, but should be a unicode string. '
340 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
341
342 self._setup_opener()
343
344 if auto_init:
345 self.print_debug_header()
346 self.add_default_info_extractors()
347
348 for pp_def_raw in self.params.get('postprocessors', []):
349 pp_class = get_postprocessor(pp_def_raw['key'])
350 pp_def = dict(pp_def_raw)
351 del pp_def['key']
352 pp = pp_class(self, **compat_kwargs(pp_def))
353 self.add_post_processor(pp)
354
355 for ph in self.params.get('progress_hooks', []):
356 self.add_progress_hook(ph)
357
358 def warn_if_short_id(self, argv):
359 # short YouTube ID starting with dash?
360 idxs = [
361 i for i, a in enumerate(argv)
362 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
363 if idxs:
364 correct_argv = (
365 ['youtube-dl'] +
366 [a for i, a in enumerate(argv) if i not in idxs] +
367 ['--'] + [argv[i] for i in idxs]
368 )
369 self.report_warning(
370 'Long argument string detected. '
371 'Use -- to separate parameters and URLs, like this:\n%s\n' %
372 args_to_str(correct_argv))
373
374 def add_info_extractor(self, ie):
375 """Add an InfoExtractor object to the end of the list."""
376 self._ies.append(ie)
377 self._ies_instances[ie.ie_key()] = ie
378 ie.set_downloader(self)
379
380 def get_info_extractor(self, ie_key):
381 """
382 Get an instance of an IE with name ie_key, it will try to get one from
383 the _ies list, if there's no instance it will create a new one and add
384 it to the extractor list.
385 """
386 ie = self._ies_instances.get(ie_key)
387 if ie is None:
388 ie = get_info_extractor(ie_key)()
389 self.add_info_extractor(ie)
390 return ie
391
392 def add_default_info_extractors(self):
393 """
394 Add the InfoExtractors returned by gen_extractors to the end of the list
395 """
396 for ie in gen_extractors():
397 self.add_info_extractor(ie)
398
399 def add_post_processor(self, pp):
400 """Add a PostProcessor object to the end of the chain."""
401 self._pps.append(pp)
402 pp.set_downloader(self)
403
404 def add_progress_hook(self, ph):
405 """Add the progress hook (currently only for the file downloader)"""
406 self._progress_hooks.append(ph)
407
408 def _bidi_workaround(self, message):
409 if not hasattr(self, '_output_channel'):
410 return message
411
412 assert hasattr(self, '_output_process')
413 assert isinstance(message, compat_str)
414 line_count = message.count('\n') + 1
415 self._output_process.stdin.write((message + '\n').encode('utf-8'))
416 self._output_process.stdin.flush()
417 res = ''.join(self._output_channel.readline().decode('utf-8')
418 for _ in range(line_count))
419 return res[:-len('\n')]
420
421 def to_screen(self, message, skip_eol=False):
422 """Print message to stdout if not in quiet mode."""
423 return self.to_stdout(message, skip_eol, check_quiet=True)
424
425 def _write_string(self, s, out=None):
426 write_string(s, out=out, encoding=self.params.get('encoding'))
427
428 def to_stdout(self, message, skip_eol=False, check_quiet=False):
429 """Print message to stdout if not in quiet mode."""
430 if self.params.get('logger'):
431 self.params['logger'].debug(message)
432 elif not check_quiet or not self.params.get('quiet', False):
433 message = self._bidi_workaround(message)
434 terminator = ['\n', ''][skip_eol]
435 output = message + terminator
436
437 self._write_string(output, self._screen_file)
438
439 def to_stderr(self, message):
440 """Print message to stderr."""
441 assert isinstance(message, compat_str)
442 if self.params.get('logger'):
443 self.params['logger'].error(message)
444 else:
445 message = self._bidi_workaround(message)
446 output = message + '\n'
447 self._write_string(output, self._err_file)
448
449 def to_console_title(self, message):
450 if not self.params.get('consoletitle', False):
451 return
452 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
453 # c_wchar_p() might not be necessary if `message` is
454 # already of type unicode()
455 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
456 elif 'TERM' in os.environ:
457 self._write_string('\033]0;%s\007' % message, self._screen_file)
458
459 def save_console_title(self):
460 if not self.params.get('consoletitle', False):
461 return
462 if 'TERM' in os.environ:
463 # Save the title on stack
464 self._write_string('\033[22;0t', self._screen_file)
465
466 def restore_console_title(self):
467 if not self.params.get('consoletitle', False):
468 return
469 if 'TERM' in os.environ:
470 # Restore the title from stack
471 self._write_string('\033[23;0t', self._screen_file)
472
473 def __enter__(self):
474 self.save_console_title()
475 return self
476
477 def __exit__(self, *args):
478 self.restore_console_title()
479
480 if self.params.get('cookiefile') is not None:
481 self.cookiejar.save()
482
483 def trouble(self, message=None, tb=None):
484 """Determine action to take when a download problem appears.
485
486 Depending on if the downloader has been configured to ignore
487 download errors or not, this method may throw an exception or
488 not when errors are found, after printing the message.
489
490 tb, if given, is additional traceback information.
491 """
492 if message is not None:
493 self.to_stderr(message)
494 if self.params.get('verbose'):
495 if tb is None:
496 if sys.exc_info()[0]: # if .trouble has been called from an except block
497 tb = ''
498 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
499 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
500 tb += encode_compat_str(traceback.format_exc())
501 else:
502 tb_data = traceback.format_list(traceback.extract_stack())
503 tb = ''.join(tb_data)
504 self.to_stderr(tb)
505 if not self.params.get('ignoreerrors', False):
506 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
507 exc_info = sys.exc_info()[1].exc_info
508 else:
509 exc_info = sys.exc_info()
510 raise DownloadError(message, exc_info)
511 self._download_retcode = 1
512
513 def report_warning(self, message):
514 '''
515 Print the message to stderr, it will be prefixed with 'WARNING:'
516 If stderr is a tty file the 'WARNING:' will be colored
517 '''
518 if self.params.get('logger') is not None:
519 self.params['logger'].warning(message)
520 else:
521 if self.params.get('no_warnings'):
522 return
523 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
524 _msg_header = '\033[0;33mWARNING:\033[0m'
525 else:
526 _msg_header = 'WARNING:'
527 warning_message = '%s %s' % (_msg_header, message)
528 self.to_stderr(warning_message)
529
530 def report_error(self, message, tb=None):
531 '''
532 Do the same as trouble, but prefixes the message with 'ERROR:', colored
533 in red if stderr is a tty file.
534 '''
535 if not self.params.get('no_color') and self._err_file.isatty() and os.name != 'nt':
536 _msg_header = '\033[0;31mERROR:\033[0m'
537 else:
538 _msg_header = 'ERROR:'
539 error_message = '%s %s' % (_msg_header, message)
540 self.trouble(error_message, tb)
541
542 def report_file_already_downloaded(self, file_name):
543 """Report file has already been fully downloaded."""
544 try:
545 self.to_screen('[download] %s has already been downloaded' % file_name)
546 except UnicodeEncodeError:
547 self.to_screen('[download] The file has already been downloaded')
548
549 def prepare_filename(self, info_dict):
550 """Generate the output filename."""
551 try:
552 template_dict = dict(info_dict)
553
554 template_dict['epoch'] = int(time.time())
555 autonumber_size = self.params.get('autonumber_size')
556 if autonumber_size is None:
557 autonumber_size = 5
558 autonumber_templ = '%0' + str(autonumber_size) + 'd'
559 template_dict['autonumber'] = autonumber_templ % self._num_downloads
560 if template_dict.get('playlist_index') is not None:
561 template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
562 if template_dict.get('resolution') is None:
563 if template_dict.get('width') and template_dict.get('height'):
564 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
565 elif template_dict.get('height'):
566 template_dict['resolution'] = '%sp' % template_dict['height']
567 elif template_dict.get('width'):
568 template_dict['resolution'] = '?x%d' % template_dict['width']
569
570 sanitize = lambda k, v: sanitize_filename(
571 compat_str(v),
572 restricted=self.params.get('restrictfilenames'),
573 is_id=(k == 'id'))
574 template_dict = dict((k, sanitize(k, v))
575 for k, v in template_dict.items()
576 if v is not None)
577 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
578
579 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
580 tmpl = compat_expanduser(outtmpl)
581 filename = tmpl % template_dict
582 # Temporary fix for #4787
583 # 'Treat' all problem characters by passing filename through preferredencoding
584 # to workaround encoding issues with subprocess on python2 @ Windows
585 if sys.version_info < (3, 0) and sys.platform == 'win32':
586 filename = encodeFilename(filename, True).decode(preferredencoding())
587 return sanitize_path(filename)
588 except ValueError as err:
589 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
590 return None
591
592 def _match_entry(self, info_dict, incomplete):
593 """ Returns None iff the file should be downloaded """
594
595 video_title = info_dict.get('title', info_dict.get('id', 'video'))
596 if 'title' in info_dict:
597 # This can happen when we're just evaluating the playlist
598 title = info_dict['title']
599 matchtitle = self.params.get('matchtitle', False)
600 if matchtitle:
601 if not re.search(matchtitle, title, re.IGNORECASE):
602 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
603 rejecttitle = self.params.get('rejecttitle', False)
604 if rejecttitle:
605 if re.search(rejecttitle, title, re.IGNORECASE):
606 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
607 date = info_dict.get('upload_date', None)
608 if date is not None:
609 dateRange = self.params.get('daterange', DateRange())
610 if date not in dateRange:
611 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
612 view_count = info_dict.get('view_count', None)
613 if view_count is not None:
614 min_views = self.params.get('min_views')
615 if min_views is not None and view_count < min_views:
616 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
617 max_views = self.params.get('max_views')
618 if max_views is not None and view_count > max_views:
619 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
620 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
621 return 'Skipping "%s" because it is age restricted' % video_title
622 if self.in_download_archive(info_dict):
623 return '%s has already been recorded in archive' % video_title
624
625 if not incomplete:
626 match_filter = self.params.get('match_filter')
627 if match_filter is not None:
628 ret = match_filter(info_dict)
629 if ret is not None:
630 return ret
631
632 return None
633
634 @staticmethod
635 def add_extra_info(info_dict, extra_info):
636 '''Set the keys from extra_info in info dict if they are missing'''
637 for key, value in extra_info.items():
638 info_dict.setdefault(key, value)
639
640 def extract_info(self, url, download=True, ie_key=None, extra_info={},
641 process=True, force_generic_extractor=False):
642 '''
643 Returns a list with a dictionary for each video we find.
644 If 'download', also downloads the videos.
645 extra_info is a dict containing the extra values to add to each result
646 '''
647
648 if not ie_key and force_generic_extractor:
649 ie_key = 'Generic'
650
651 if ie_key:
652 ies = [self.get_info_extractor(ie_key)]
653 else:
654 ies = self._ies
655
656 for ie in ies:
657 if not ie.suitable(url):
658 continue
659
660 if not ie.working():
661 self.report_warning('The program functionality for this site has been marked as broken, '
662 'and will probably not work.')
663
664 try:
665 ie_result = ie.extract(url)
666 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
667 break
668 if isinstance(ie_result, list):
669 # Backwards compatibility: old IE result format
670 ie_result = {
671 '_type': 'compat_list',
672 'entries': ie_result,
673 }
674 self.add_default_extra_info(ie_result, ie, url)
675 if process:
676 return self.process_ie_result(ie_result, download, extra_info)
677 else:
678 return ie_result
679 except ExtractorError as e: # An error we somewhat expected
680 self.report_error(compat_str(e), e.format_traceback())
681 break
682 except MaxDownloadsReached:
683 raise
684 except Exception as e:
685 if self.params.get('ignoreerrors', False):
686 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
687 break
688 else:
689 raise
690 else:
691 self.report_error('no suitable InfoExtractor for URL %s' % url)
692
693 def add_default_extra_info(self, ie_result, ie, url):
694 self.add_extra_info(ie_result, {
695 'extractor': ie.IE_NAME,
696 'webpage_url': url,
697 'webpage_url_basename': url_basename(url),
698 'extractor_key': ie.ie_key(),
699 })
700
701 def process_ie_result(self, ie_result, download=True, extra_info={}):
702 """
703 Take the result of the ie(may be modified) and resolve all unresolved
704 references (URLs, playlist items).
705
706 It will also download the videos if 'download'.
707 Returns the resolved ie_result.
708 """
709
710 result_type = ie_result.get('_type', 'video')
711
712 if result_type in ('url', 'url_transparent'):
713 extract_flat = self.params.get('extract_flat', False)
714 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
715 extract_flat is True):
716 if self.params.get('forcejson', False):
717 self.to_stdout(json.dumps(ie_result))
718 return ie_result
719
720 if result_type == 'video':
721 self.add_extra_info(ie_result, extra_info)
722 return self.process_video_result(ie_result, download=download)
723 elif result_type == 'url':
724 # We have to add extra_info to the results because it may be
725 # contained in a playlist
726 return self.extract_info(ie_result['url'],
727 download,
728 ie_key=ie_result.get('ie_key'),
729 extra_info=extra_info)
730 elif result_type == 'url_transparent':
731 # Use the information from the embedding page
732 info = self.extract_info(
733 ie_result['url'], ie_key=ie_result.get('ie_key'),
734 extra_info=extra_info, download=False, process=False)
735
736 force_properties = dict(
737 (k, v) for k, v in ie_result.items() if v is not None)
738 for f in ('_type', 'url'):
739 if f in force_properties:
740 del force_properties[f]
741 new_result = info.copy()
742 new_result.update(force_properties)
743
744 assert new_result.get('_type') != 'url_transparent'
745
746 return self.process_ie_result(
747 new_result, download=download, extra_info=extra_info)
748 elif result_type == 'playlist' or result_type == 'multi_video':
749 # We process each entry in the playlist
750 playlist = ie_result.get('title', None) or ie_result.get('id', None)
751 self.to_screen('[download] Downloading playlist: %s' % playlist)
752
753 playlist_results = []
754
755 playliststart = self.params.get('playliststart', 1) - 1
756 playlistend = self.params.get('playlistend', None)
757 # For backwards compatibility, interpret -1 as whole list
758 if playlistend == -1:
759 playlistend = None
760
761 playlistitems_str = self.params.get('playlist_items', None)
762 playlistitems = None
763 if playlistitems_str is not None:
764 def iter_playlistitems(format):
765 for string_segment in format.split(','):
766 if '-' in string_segment:
767 start, end = string_segment.split('-')
768 for item in range(int(start), int(end) + 1):
769 yield int(item)
770 else:
771 yield int(string_segment)
772 playlistitems = iter_playlistitems(playlistitems_str)
773
774 ie_entries = ie_result['entries']
775 if isinstance(ie_entries, list):
776 n_all_entries = len(ie_entries)
777 if playlistitems:
778 entries = [
779 ie_entries[i - 1] for i in playlistitems
780 if -n_all_entries <= i - 1 < n_all_entries]
781 else:
782 entries = ie_entries[playliststart:playlistend]
783 n_entries = len(entries)
784 self.to_screen(
785 "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
786 (ie_result['extractor'], playlist, n_all_entries, n_entries))
787 elif isinstance(ie_entries, PagedList):
788 if playlistitems:
789 entries = []
790 for item in playlistitems:
791 entries.extend(ie_entries.getslice(
792 item - 1, item
793 ))
794 else:
795 entries = ie_entries.getslice(
796 playliststart, playlistend)
797 n_entries = len(entries)
798 self.to_screen(
799 "[%s] playlist %s: Downloading %d videos" %
800 (ie_result['extractor'], playlist, n_entries))
801 else: # iterable
802 if playlistitems:
803 entry_list = list(ie_entries)
804 entries = [entry_list[i - 1] for i in playlistitems]
805 else:
806 entries = list(itertools.islice(
807 ie_entries, playliststart, playlistend))
808 n_entries = len(entries)
809 self.to_screen(
810 "[%s] playlist %s: Downloading %d videos" %
811 (ie_result['extractor'], playlist, n_entries))
812
813 if self.params.get('playlistreverse', False):
814 entries = entries[::-1]
815
816 for i, entry in enumerate(entries, 1):
817 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
818 extra = {
819 'n_entries': n_entries,
820 'playlist': playlist,
821 'playlist_id': ie_result.get('id'),
822 'playlist_title': ie_result.get('title'),
823 'playlist_index': i + playliststart,
824 'extractor': ie_result['extractor'],
825 'webpage_url': ie_result['webpage_url'],
826 'webpage_url_basename': url_basename(ie_result['webpage_url']),
827 'extractor_key': ie_result['extractor_key'],
828 }
829
830 reason = self._match_entry(entry, incomplete=True)
831 if reason is not None:
832 self.to_screen('[download] ' + reason)
833 continue
834
835 entry_result = self.process_ie_result(entry,
836 download=download,
837 extra_info=extra)
838 playlist_results.append(entry_result)
839 ie_result['entries'] = playlist_results
840 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
841 return ie_result
842 elif result_type == 'compat_list':
843 self.report_warning(
844 'Extractor %s returned a compat_list result. '
845 'It needs to be updated.' % ie_result.get('extractor'))
846
847 def _fixup(r):
848 self.add_extra_info(
849 r,
850 {
851 'extractor': ie_result['extractor'],
852 'webpage_url': ie_result['webpage_url'],
853 'webpage_url_basename': url_basename(ie_result['webpage_url']),
854 'extractor_key': ie_result['extractor_key'],
855 }
856 )
857 return r
858 ie_result['entries'] = [
859 self.process_ie_result(_fixup(r), download, extra_info)
860 for r in ie_result['entries']
861 ]
862 return ie_result
863 else:
864 raise Exception('Invalid result type: %s' % result_type)
865
866 def _build_format_filter(self, filter_spec):
867 " Returns a function to filter the formats according to the filter_spec "
868
869 OPERATORS = {
870 '<': operator.lt,
871 '<=': operator.le,
872 '>': operator.gt,
873 '>=': operator.ge,
874 '=': operator.eq,
875 '!=': operator.ne,
876 }
877 operator_rex = re.compile(r'''(?x)\s*
878 (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps)
879 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
880 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
881 $
882 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
883 m = operator_rex.search(filter_spec)
884 if m:
885 try:
886 comparison_value = int(m.group('value'))
887 except ValueError:
888 comparison_value = parse_filesize(m.group('value'))
889 if comparison_value is None:
890 comparison_value = parse_filesize(m.group('value') + 'B')
891 if comparison_value is None:
892 raise ValueError(
893 'Invalid value %r in format specification %r' % (
894 m.group('value'), filter_spec))
895 op = OPERATORS[m.group('op')]
896
897 if not m:
898 STR_OPERATORS = {
899 '=': operator.eq,
900 '!=': operator.ne,
901 '^=': lambda attr, value: attr.startswith(value),
902 '$=': lambda attr, value: attr.endswith(value),
903 '*=': lambda attr, value: value in attr,
904 }
905 str_operator_rex = re.compile(r'''(?x)
906 \s*(?P<key>ext|acodec|vcodec|container|protocol)
907 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?
908 \s*(?P<value>[a-zA-Z0-9_-]+)
909 \s*$
910 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
911 m = str_operator_rex.search(filter_spec)
912 if m:
913 comparison_value = m.group('value')
914 op = STR_OPERATORS[m.group('op')]
915
916 if not m:
917 raise ValueError('Invalid filter specification %r' % filter_spec)
918
919 def _filter(f):
920 actual_value = f.get(m.group('key'))
921 if actual_value is None:
922 return m.group('none_inclusive')
923 return op(actual_value, comparison_value)
924 return _filter
925
926 def build_format_selector(self, format_spec):
927 def syntax_error(note, start):
928 message = (
929 'Invalid format specification: '
930 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
931 return SyntaxError(message)
932
933 PICKFIRST = 'PICKFIRST'
934 MERGE = 'MERGE'
935 SINGLE = 'SINGLE'
936 GROUP = 'GROUP'
937 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
938
939 def _parse_filter(tokens):
940 filter_parts = []
941 for type, string, start, _, _ in tokens:
942 if type == tokenize.OP and string == ']':
943 return ''.join(filter_parts)
944 else:
945 filter_parts.append(string)
946
947 def _remove_unused_ops(tokens):
948 # Remove operators that we don't use and join them with the surrounding strings
949 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
950 ALLOWED_OPS = ('/', '+', ',', '(', ')')
951 last_string, last_start, last_end, last_line = None, None, None, None
952 for type, string, start, end, line in tokens:
953 if type == tokenize.OP and string == '[':
954 if last_string:
955 yield tokenize.NAME, last_string, last_start, last_end, last_line
956 last_string = None
957 yield type, string, start, end, line
958 # everything inside brackets will be handled by _parse_filter
959 for type, string, start, end, line in tokens:
960 yield type, string, start, end, line
961 if type == tokenize.OP and string == ']':
962 break
963 elif type == tokenize.OP and string in ALLOWED_OPS:
964 if last_string:
965 yield tokenize.NAME, last_string, last_start, last_end, last_line
966 last_string = None
967 yield type, string, start, end, line
968 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
969 if not last_string:
970 last_string = string
971 last_start = start
972 last_end = end
973 else:
974 last_string += string
975 if last_string:
976 yield tokenize.NAME, last_string, last_start, last_end, last_line
977
978 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
979 selectors = []
980 current_selector = None
981 for type, string, start, _, _ in tokens:
982 # ENCODING is only defined in python 3.x
983 if type == getattr(tokenize, 'ENCODING', None):
984 continue
985 elif type in [tokenize.NAME, tokenize.NUMBER]:
986 current_selector = FormatSelector(SINGLE, string, [])
987 elif type == tokenize.OP:
988 if string == ')':
989 if not inside_group:
990 # ')' will be handled by the parentheses group
991 tokens.restore_last_token()
992 break
993 elif inside_merge and string in ['/', ',']:
994 tokens.restore_last_token()
995 break
996 elif inside_choice and string == ',':
997 tokens.restore_last_token()
998 break
999 elif string == ',':
1000 if not current_selector:
1001 raise syntax_error('"," must follow a format selector', start)
1002 selectors.append(current_selector)
1003 current_selector = None
1004 elif string == '/':
1005 if not current_selector:
1006 raise syntax_error('"/" must follow a format selector', start)
1007 first_choice = current_selector
1008 second_choice = _parse_format_selection(tokens, inside_choice=True)
1009 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1010 elif string == '[':
1011 if not current_selector:
1012 current_selector = FormatSelector(SINGLE, 'best', [])
1013 format_filter = _parse_filter(tokens)
1014 current_selector.filters.append(format_filter)
1015 elif string == '(':
1016 if current_selector:
1017 raise syntax_error('Unexpected "("', start)
1018 group = _parse_format_selection(tokens, inside_group=True)
1019 current_selector = FormatSelector(GROUP, group, [])
1020 elif string == '+':
1021 video_selector = current_selector
1022 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1023 if not video_selector or not audio_selector:
1024 raise syntax_error('"+" must be between two format selectors', start)
1025 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1026 else:
1027 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1028 elif type == tokenize.ENDMARKER:
1029 break
1030 if current_selector:
1031 selectors.append(current_selector)
1032 return selectors
1033
1034 def _build_selector_function(selector):
1035 if isinstance(selector, list):
1036 fs = [_build_selector_function(s) for s in selector]
1037
1038 def selector_function(formats):
1039 for f in fs:
1040 for format in f(formats):
1041 yield format
1042 return selector_function
1043 elif selector.type == GROUP:
1044 selector_function = _build_selector_function(selector.selector)
1045 elif selector.type == PICKFIRST:
1046 fs = [_build_selector_function(s) for s in selector.selector]
1047
1048 def selector_function(formats):
1049 for f in fs:
1050 picked_formats = list(f(formats))
1051 if picked_formats:
1052 return picked_formats
1053 return []
1054 elif selector.type == SINGLE:
1055 format_spec = selector.selector
1056
1057 def selector_function(formats):
1058 formats = list(formats)
1059 if not formats:
1060 return
1061 if format_spec == 'all':
1062 for f in formats:
1063 yield f
1064 elif format_spec in ['best', 'worst', None]:
1065 format_idx = 0 if format_spec == 'worst' else -1
1066 audiovideo_formats = [
1067 f for f in formats
1068 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1069 if audiovideo_formats:
1070 yield audiovideo_formats[format_idx]
1071 # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
1072 elif (all(f.get('acodec') != 'none' for f in formats) or
1073 all(f.get('vcodec') != 'none' for f in formats)):
1074 yield formats[format_idx]
1075 elif format_spec == 'bestaudio':
1076 audio_formats = [
1077 f for f in formats
1078 if f.get('vcodec') == 'none']
1079 if audio_formats:
1080 yield audio_formats[-1]
1081 elif format_spec == 'worstaudio':
1082 audio_formats = [
1083 f for f in formats
1084 if f.get('vcodec') == 'none']
1085 if audio_formats:
1086 yield audio_formats[0]
1087 elif format_spec == 'bestvideo':
1088 video_formats = [
1089 f for f in formats
1090 if f.get('acodec') == 'none']
1091 if video_formats:
1092 yield video_formats[-1]
1093 elif format_spec == 'worstvideo':
1094 video_formats = [
1095 f for f in formats
1096 if f.get('acodec') == 'none']
1097 if video_formats:
1098 yield video_formats[0]
1099 else:
1100 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1101 if format_spec in extensions:
1102 filter_f = lambda f: f['ext'] == format_spec
1103 else:
1104 filter_f = lambda f: f['format_id'] == format_spec
1105 matches = list(filter(filter_f, formats))
1106 if matches:
1107 yield matches[-1]
1108 elif selector.type == MERGE:
1109 def _merge(formats_info):
1110 format_1, format_2 = [f['format_id'] for f in formats_info]
1111 # The first format must contain the video and the
1112 # second the audio
1113 if formats_info[0].get('vcodec') == 'none':
1114 self.report_error('The first format must '
1115 'contain the video, try using '
1116 '"-f %s+%s"' % (format_2, format_1))
1117 return
1118 # Formats must be opposite (video+audio)
1119 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1120 self.report_error(
1121 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1122 % (format_1, format_2))
1123 return
1124 output_ext = (
1125 formats_info[0]['ext']
1126 if self.params.get('merge_output_format') is None
1127 else self.params['merge_output_format'])
1128 return {
1129 'requested_formats': formats_info,
1130 'format': '%s+%s' % (formats_info[0].get('format'),
1131 formats_info[1].get('format')),
1132 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1133 formats_info[1].get('format_id')),
1134 'width': formats_info[0].get('width'),
1135 'height': formats_info[0].get('height'),
1136 'resolution': formats_info[0].get('resolution'),
1137 'fps': formats_info[0].get('fps'),
1138 'vcodec': formats_info[0].get('vcodec'),
1139 'vbr': formats_info[0].get('vbr'),
1140 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1141 'acodec': formats_info[1].get('acodec'),
1142 'abr': formats_info[1].get('abr'),
1143 'ext': output_ext,
1144 }
1145 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1146
1147 def selector_function(formats):
1148 formats = list(formats)
1149 for pair in itertools.product(video_selector(formats), audio_selector(formats)):
1150 yield _merge(pair)
1151
1152 filters = [self._build_format_filter(f) for f in selector.filters]
1153
1154 def final_selector(formats):
1155 for _filter in filters:
1156 formats = list(filter(_filter, formats))
1157 return selector_function(formats)
1158 return final_selector
1159
1160 stream = io.BytesIO(format_spec.encode('utf-8'))
1161 try:
1162 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1163 except tokenize.TokenError:
1164 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1165
1166 class TokenIterator(object):
1167 def __init__(self, tokens):
1168 self.tokens = tokens
1169 self.counter = 0
1170
1171 def __iter__(self):
1172 return self
1173
1174 def __next__(self):
1175 if self.counter >= len(self.tokens):
1176 raise StopIteration()
1177 value = self.tokens[self.counter]
1178 self.counter += 1
1179 return value
1180
1181 next = __next__
1182
1183 def restore_last_token(self):
1184 self.counter -= 1
1185
1186 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1187 return _build_selector_function(parsed_selector)
1188
1189 def _calc_headers(self, info_dict):
1190 res = std_headers.copy()
1191
1192 add_headers = info_dict.get('http_headers')
1193 if add_headers:
1194 res.update(add_headers)
1195
1196 cookies = self._calc_cookies(info_dict)
1197 if cookies:
1198 res['Cookie'] = cookies
1199
1200 return res
1201
1202 def _calc_cookies(self, info_dict):
1203 pr = sanitized_Request(info_dict['url'])
1204 self.cookiejar.add_cookie_header(pr)
1205 return pr.get_header('Cookie')
1206
1207 def process_video_result(self, info_dict, download=True):
1208 assert info_dict.get('_type', 'video') == 'video'
1209
1210 if 'id' not in info_dict:
1211 raise ExtractorError('Missing "id" field in extractor result')
1212 if 'title' not in info_dict:
1213 raise ExtractorError('Missing "title" field in extractor result')
1214
1215 if 'playlist' not in info_dict:
1216 # It isn't part of a playlist
1217 info_dict['playlist'] = None
1218 info_dict['playlist_index'] = None
1219
1220 thumbnails = info_dict.get('thumbnails')
1221 if thumbnails is None:
1222 thumbnail = info_dict.get('thumbnail')
1223 if thumbnail:
1224 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1225 if thumbnails:
1226 thumbnails.sort(key=lambda t: (
1227 t.get('preference'), t.get('width'), t.get('height'),
1228 t.get('id'), t.get('url')))
1229 for i, t in enumerate(thumbnails):
1230 if t.get('width') and t.get('height'):
1231 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1232 if t.get('id') is None:
1233 t['id'] = '%d' % i
1234
1235 if thumbnails and 'thumbnail' not in info_dict:
1236 info_dict['thumbnail'] = thumbnails[-1]['url']
1237
1238 if 'display_id' not in info_dict and 'id' in info_dict:
1239 info_dict['display_id'] = info_dict['id']
1240
1241 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1242 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1243 # see http://bugs.python.org/issue1646728)
1244 try:
1245 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1246 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1247 except (ValueError, OverflowError, OSError):
1248 pass
1249
1250 # Auto generate title fields corresponding to the *_number fields when missing
1251 # in order to always have clean titles. This is very common for TV series.
1252 for field in ('chapter', 'season', 'episode'):
1253 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1254 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1255
1256 subtitles = info_dict.get('subtitles')
1257 if subtitles:
1258 for _, subtitle in subtitles.items():
1259 for subtitle_format in subtitle:
1260 if 'ext' not in subtitle_format:
1261 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1262
1263 if self.params.get('listsubtitles', False):
1264 if 'automatic_captions' in info_dict:
1265 self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions')
1266 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1267 return
1268 info_dict['requested_subtitles'] = self.process_subtitles(
1269 info_dict['id'], subtitles,
1270 info_dict.get('automatic_captions'))
1271
1272 # We now pick which formats have to be downloaded
1273 if info_dict.get('formats') is None:
1274 # There's only one format available
1275 formats = [info_dict]
1276 else:
1277 formats = info_dict['formats']
1278
1279 if not formats:
1280 raise ExtractorError('No video formats found!')
1281
1282 formats_dict = {}
1283
1284 # We check that all the formats have the format and format_id fields
1285 for i, format in enumerate(formats):
1286 if 'url' not in format:
1287 raise ExtractorError('Missing "url" key in result (index %d)' % i)
1288
1289 if format.get('format_id') is None:
1290 format['format_id'] = compat_str(i)
1291 format_id = format['format_id']
1292 if format_id not in formats_dict:
1293 formats_dict[format_id] = []
1294 formats_dict[format_id].append(format)
1295
1296 # Make sure all formats have unique format_id
1297 for format_id, ambiguous_formats in formats_dict.items():
1298 if len(ambiguous_formats) > 1:
1299 for i, format in enumerate(ambiguous_formats):
1300 format['format_id'] = '%s-%d' % (format_id, i)
1301
1302 for i, format in enumerate(formats):
1303 if format.get('format') is None:
1304 format['format'] = '{id} - {res}{note}'.format(
1305 id=format['format_id'],
1306 res=self.format_resolution(format),
1307 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1308 )
1309 # Automatically determine file extension if missing
1310 if 'ext' not in format:
1311 format['ext'] = determine_ext(format['url']).lower()
1312 # Add HTTP headers, so that external programs can use them from the
1313 # json output
1314 full_format_info = info_dict.copy()
1315 full_format_info.update(format)
1316 format['http_headers'] = self._calc_headers(full_format_info)
1317
1318 # TODO Central sorting goes here
1319
1320 if formats[0] is not info_dict:
1321 # only set the 'formats' fields if the original info_dict list them
1322 # otherwise we end up with a circular reference, the first (and unique)
1323 # element in the 'formats' field in info_dict is info_dict itself,
1324 # which can't be exported to json
1325 info_dict['formats'] = formats
1326 if self.params.get('listformats'):
1327 self.list_formats(info_dict)
1328 return
1329 if self.params.get('list_thumbnails'):
1330 self.list_thumbnails(info_dict)
1331 return
1332
1333 req_format = self.params.get('format')
1334 if req_format is None:
1335 req_format_list = []
1336 if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
1337 info_dict['extractor'] in ['youtube', 'ted'] and
1338 not info_dict.get('is_live')):
1339 merger = FFmpegMergerPP(self)
1340 if merger.available and merger.can_merge():
1341 req_format_list.append('bestvideo+bestaudio')
1342 req_format_list.append('best')
1343 req_format = '/'.join(req_format_list)
1344 format_selector = self.build_format_selector(req_format)
1345 formats_to_download = list(format_selector(formats))
1346 if not formats_to_download:
1347 raise ExtractorError('requested format not available',
1348 expected=True)
1349
1350 if download:
1351 if len(formats_to_download) > 1:
1352 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1353 for format in formats_to_download:
1354 new_info = dict(info_dict)
1355 new_info.update(format)
1356 self.process_info(new_info)
1357 # We update the info dict with the best quality format (backwards compatibility)
1358 info_dict.update(formats_to_download[-1])
1359 return info_dict
1360
1361 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1362 """Select the requested subtitles and their format"""
1363 available_subs = {}
1364 if normal_subtitles and self.params.get('writesubtitles'):
1365 available_subs.update(normal_subtitles)
1366 if automatic_captions and self.params.get('writeautomaticsub'):
1367 for lang, cap_info in automatic_captions.items():
1368 if lang not in available_subs:
1369 available_subs[lang] = cap_info
1370
1371 if (not self.params.get('writesubtitles') and not
1372 self.params.get('writeautomaticsub') or not
1373 available_subs):
1374 return None
1375
1376 if self.params.get('allsubtitles', False):
1377 requested_langs = available_subs.keys()
1378 else:
1379 if self.params.get('subtitleslangs', False):
1380 requested_langs = self.params.get('subtitleslangs')
1381 elif 'en' in available_subs:
1382 requested_langs = ['en']
1383 else:
1384 requested_langs = [list(available_subs.keys())[0]]
1385
1386 formats_query = self.params.get('subtitlesformat', 'best')
1387 formats_preference = formats_query.split('/') if formats_query else []
1388 subs = {}
1389 for lang in requested_langs:
1390 formats = available_subs.get(lang)
1391 if formats is None:
1392 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1393 continue
1394 for ext in formats_preference:
1395 if ext == 'best':
1396 f = formats[-1]
1397 break
1398 matches = list(filter(lambda f: f['ext'] == ext, formats))
1399 if matches:
1400 f = matches[-1]
1401 break
1402 else:
1403 f = formats[-1]
1404 self.report_warning(
1405 'No subtitle format found matching "%s" for language %s, '
1406 'using %s' % (formats_query, lang, f['ext']))
1407 subs[lang] = f
1408 return subs
1409
1410 def process_info(self, info_dict):
1411 """Process a single resolved IE result."""
1412
1413 assert info_dict.get('_type', 'video') == 'video'
1414
1415 max_downloads = self.params.get('max_downloads')
1416 if max_downloads is not None:
1417 if self._num_downloads >= int(max_downloads):
1418 raise MaxDownloadsReached()
1419
1420 info_dict['fulltitle'] = info_dict['title']
1421 if len(info_dict['title']) > 200:
1422 info_dict['title'] = info_dict['title'][:197] + '...'
1423
1424 if 'format' not in info_dict:
1425 info_dict['format'] = info_dict['ext']
1426
1427 reason = self._match_entry(info_dict, incomplete=False)
1428 if reason is not None:
1429 self.to_screen('[download] ' + reason)
1430 return
1431
1432 self._num_downloads += 1
1433
1434 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1435
1436 # Forced printings
1437 if self.params.get('forcetitle', False):
1438 self.to_stdout(info_dict['fulltitle'])
1439 if self.params.get('forceid', False):
1440 self.to_stdout(info_dict['id'])
1441 if self.params.get('forceurl', False):
1442 if info_dict.get('requested_formats') is not None:
1443 for f in info_dict['requested_formats']:
1444 self.to_stdout(f['url'] + f.get('play_path', ''))
1445 else:
1446 # For RTMP URLs, also include the playpath
1447 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1448 if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
1449 self.to_stdout(info_dict['thumbnail'])
1450 if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
1451 self.to_stdout(info_dict['description'])
1452 if self.params.get('forcefilename', False) and filename is not None:
1453 self.to_stdout(filename)
1454 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1455 self.to_stdout(formatSeconds(info_dict['duration']))
1456 if self.params.get('forceformat', False):
1457 self.to_stdout(info_dict['format'])
1458 if self.params.get('forcejson', False):
1459 self.to_stdout(json.dumps(info_dict))
1460
1461 # Do nothing else if in simulate mode
1462 if self.params.get('simulate', False):
1463 return
1464
1465 if filename is None:
1466 return
1467
1468 try:
1469 dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
1470 if dn and not os.path.exists(dn):
1471 os.makedirs(dn)
1472 except (OSError, IOError) as err:
1473 self.report_error('unable to create directory ' + error_to_compat_str(err))
1474 return
1475
1476 if self.params.get('writedescription', False):
1477 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1478 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1479 self.to_screen('[info] Video description is already present')
1480 elif info_dict.get('description') is None:
1481 self.report_warning('There\'s no description to write.')
1482 else:
1483 try:
1484 self.to_screen('[info] Writing video description to: ' + descfn)
1485 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1486 descfile.write(info_dict['description'])
1487 except (OSError, IOError):
1488 self.report_error('Cannot write description file ' + descfn)
1489 return
1490
1491 if self.params.get('writeannotations', False):
1492 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1493 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1494 self.to_screen('[info] Video annotations are already present')
1495 else:
1496 try:
1497 self.to_screen('[info] Writing video annotations to: ' + annofn)
1498 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1499 annofile.write(info_dict['annotations'])
1500 except (KeyError, TypeError):
1501 self.report_warning('There are no annotations to write.')
1502 except (OSError, IOError):
1503 self.report_error('Cannot write annotations file: ' + annofn)
1504 return
1505
1506 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1507 self.params.get('writeautomaticsub')])
1508
1509 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1510 # subtitles download errors are already managed as troubles in relevant IE
1511 # that way it will silently go on when used with unsupporting IE
1512 subtitles = info_dict['requested_subtitles']
1513 ie = self.get_info_extractor(info_dict['extractor_key'])
1514 for sub_lang, sub_info in subtitles.items():
1515 sub_format = sub_info['ext']
1516 if sub_info.get('data') is not None:
1517 sub_data = sub_info['data']
1518 else:
1519 try:
1520 sub_data = ie._download_webpage(
1521 sub_info['url'], info_dict['id'], note=False)
1522 except ExtractorError as err:
1523 self.report_warning('Unable to download subtitle for "%s": %s' %
1524 (sub_lang, error_to_compat_str(err.cause)))
1525 continue
1526 try:
1527 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
1528 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1529 self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
1530 else:
1531 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1532 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
1533 subfile.write(sub_data)
1534 except (OSError, IOError):
1535 self.report_error('Cannot write subtitles file ' + sub_filename)
1536 return
1537
1538 if self.params.get('writeinfojson', False):
1539 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1540 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1541 self.to_screen('[info] Video description metadata is already present')
1542 else:
1543 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1544 try:
1545 write_json_file(self.filter_requested_info(info_dict), infofn)
1546 except (OSError, IOError):
1547 self.report_error('Cannot write metadata to JSON file ' + infofn)
1548 return
1549
1550 self._write_thumbnails(info_dict, filename)
1551
1552 if not self.params.get('skip_download', False):
1553 try:
1554 def dl(name, info):
1555 fd = get_suitable_downloader(info, self.params)(self, self.params)
1556 for ph in self._progress_hooks:
1557 fd.add_progress_hook(ph)
1558 if self.params.get('verbose'):
1559 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1560 return fd.download(name, info)
1561
1562 if info_dict.get('requested_formats') is not None:
1563 downloaded = []
1564 success = True
1565 merger = FFmpegMergerPP(self)
1566 if not merger.available:
1567 postprocessors = []
1568 self.report_warning('You have requested multiple '
1569 'formats but ffmpeg or avconv are not installed.'
1570 ' The formats won\'t be merged.')
1571 else:
1572 postprocessors = [merger]
1573
1574 def compatible_formats(formats):
1575 video, audio = formats
1576 # Check extension
1577 video_ext, audio_ext = audio.get('ext'), video.get('ext')
1578 if video_ext and audio_ext:
1579 COMPATIBLE_EXTS = (
1580 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
1581 ('webm')
1582 )
1583 for exts in COMPATIBLE_EXTS:
1584 if video_ext in exts and audio_ext in exts:
1585 return True
1586 # TODO: Check acodec/vcodec
1587 return False
1588
1589 filename_real_ext = os.path.splitext(filename)[1][1:]
1590 filename_wo_ext = (
1591 os.path.splitext(filename)[0]
1592 if filename_real_ext == info_dict['ext']
1593 else filename)
1594 requested_formats = info_dict['requested_formats']
1595 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1596 info_dict['ext'] = 'mkv'
1597 self.report_warning(
1598 'Requested formats are incompatible for merge and will be merged into mkv.')
1599 # Ensure filename always has a correct extension for successful merge
1600 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1601 if os.path.exists(encodeFilename(filename)):
1602 self.to_screen(
1603 '[download] %s has already been downloaded and '
1604 'merged' % filename)
1605 else:
1606 for f in requested_formats:
1607 new_info = dict(info_dict)
1608 new_info.update(f)
1609 fname = self.prepare_filename(new_info)
1610 fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
1611 downloaded.append(fname)
1612 partial_success = dl(fname, new_info)
1613 success = success and partial_success
1614 info_dict['__postprocessors'] = postprocessors
1615 info_dict['__files_to_merge'] = downloaded
1616 else:
1617 # Just a single file
1618 success = dl(filename, info_dict)
1619 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1620 self.report_error('unable to download video data: %s' % str(err))
1621 return
1622 except (OSError, IOError) as err:
1623 raise UnavailableVideoError(err)
1624 except (ContentTooShortError, ) as err:
1625 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1626 return
1627
1628 if success:
1629 # Fixup content
1630 fixup_policy = self.params.get('fixup')
1631 if fixup_policy is None:
1632 fixup_policy = 'detect_or_warn'
1633
1634 stretched_ratio = info_dict.get('stretched_ratio')
1635 if stretched_ratio is not None and stretched_ratio != 1:
1636 if fixup_policy == 'warn':
1637 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1638 info_dict['id'], stretched_ratio))
1639 elif fixup_policy == 'detect_or_warn':
1640 stretched_pp = FFmpegFixupStretchedPP(self)
1641 if stretched_pp.available:
1642 info_dict.setdefault('__postprocessors', [])
1643 info_dict['__postprocessors'].append(stretched_pp)
1644 else:
1645 self.report_warning(
1646 '%s: Non-uniform pixel ratio (%s). Install ffmpeg or avconv to fix this automatically.' % (
1647 info_dict['id'], stretched_ratio))
1648 else:
1649 assert fixup_policy in ('ignore', 'never')
1650
1651 if info_dict.get('requested_formats') is None and info_dict.get('container') == 'm4a_dash':
1652 if fixup_policy == 'warn':
1653 self.report_warning('%s: writing DASH m4a. Only some players support this container.' % (
1654 info_dict['id']))
1655 elif fixup_policy == 'detect_or_warn':
1656 fixup_pp = FFmpegFixupM4aPP(self)
1657 if fixup_pp.available:
1658 info_dict.setdefault('__postprocessors', [])
1659 info_dict['__postprocessors'].append(fixup_pp)
1660 else:
1661 self.report_warning(
1662 '%s: writing DASH m4a. Only some players support this container. Install ffmpeg or avconv to fix this automatically.' % (
1663 info_dict['id']))
1664 else:
1665 assert fixup_policy in ('ignore', 'never')
1666
1667 try:
1668 self.post_process(filename, info_dict)
1669 except (PostProcessingError) as err:
1670 self.report_error('postprocessing: %s' % str(err))
1671 return
1672 self.record_download_archive(info_dict)
1673
1674 def download(self, url_list):
1675 """Download a given list of URLs."""
1676 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
1677 if (len(url_list) > 1 and
1678 '%' not in outtmpl and
1679 self.params.get('max_downloads') != 1):
1680 raise SameFileError(outtmpl)
1681
1682 for url in url_list:
1683 try:
1684 # It also downloads the videos
1685 res = self.extract_info(
1686 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
1687 except UnavailableVideoError:
1688 self.report_error('unable to download video')
1689 except MaxDownloadsReached:
1690 self.to_screen('[info] Maximum number of downloaded files reached.')
1691 raise
1692 else:
1693 if self.params.get('dump_single_json', False):
1694 self.to_stdout(json.dumps(res))
1695
1696 return self._download_retcode
1697
1698 def download_with_info_file(self, info_filename):
1699 with contextlib.closing(fileinput.FileInput(
1700 [info_filename], mode='r',
1701 openhook=fileinput.hook_encoded('utf-8'))) as f:
1702 # FileInput doesn't have a read method, we can't call json.load
1703 info = self.filter_requested_info(json.loads('\n'.join(f)))
1704 try:
1705 self.process_ie_result(info, download=True)
1706 except DownloadError:
1707 webpage_url = info.get('webpage_url')
1708 if webpage_url is not None:
1709 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
1710 return self.download([webpage_url])
1711 else:
1712 raise
1713 return self._download_retcode
1714
1715 @staticmethod
1716 def filter_requested_info(info_dict):
1717 return dict(
1718 (k, v) for k, v in info_dict.items()
1719 if k not in ['requested_formats', 'requested_subtitles'])
1720
1721 def post_process(self, filename, ie_info):
1722 """Run all the postprocessors on the given file."""
1723 info = dict(ie_info)
1724 info['filepath'] = filename
1725 pps_chain = []
1726 if ie_info.get('__postprocessors') is not None:
1727 pps_chain.extend(ie_info['__postprocessors'])
1728 pps_chain.extend(self._pps)
1729 for pp in pps_chain:
1730 files_to_delete = []
1731 try:
1732 files_to_delete, info = pp.run(info)
1733 except PostProcessingError as e:
1734 self.report_error(e.msg)
1735 if files_to_delete and not self.params.get('keepvideo', False):
1736 for old_filename in files_to_delete:
1737 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
1738 try:
1739 os.remove(encodeFilename(old_filename))
1740 except (IOError, OSError):
1741 self.report_warning('Unable to remove downloaded original file')
1742
1743 def _make_archive_id(self, info_dict):
1744 # Future-proof against any change in case
1745 # and backwards compatibility with prior versions
1746 extractor = info_dict.get('extractor_key')
1747 if extractor is None:
1748 if 'id' in info_dict:
1749 extractor = info_dict.get('ie_key') # key in a playlist
1750 if extractor is None:
1751 return None # Incomplete video information
1752 return extractor.lower() + ' ' + info_dict['id']
1753
1754 def in_download_archive(self, info_dict):
1755 fn = self.params.get('download_archive')
1756 if fn is None:
1757 return False
1758
1759 vid_id = self._make_archive_id(info_dict)
1760 if vid_id is None:
1761 return False # Incomplete video information
1762
1763 try:
1764 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
1765 for line in archive_file:
1766 if line.strip() == vid_id:
1767 return True
1768 except IOError as ioe:
1769 if ioe.errno != errno.ENOENT:
1770 raise
1771 return False
1772
1773 def record_download_archive(self, info_dict):
1774 fn = self.params.get('download_archive')
1775 if fn is None:
1776 return
1777 vid_id = self._make_archive_id(info_dict)
1778 assert vid_id
1779 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
1780 archive_file.write(vid_id + '\n')
1781
1782 @staticmethod
1783 def format_resolution(format, default='unknown'):
1784 if format.get('vcodec') == 'none':
1785 return 'audio only'
1786 if format.get('resolution') is not None:
1787 return format['resolution']
1788 if format.get('height') is not None:
1789 if format.get('width') is not None:
1790 res = '%sx%s' % (format['width'], format['height'])
1791 else:
1792 res = '%sp' % format['height']
1793 elif format.get('width') is not None:
1794 res = '?x%d' % format['width']
1795 else:
1796 res = default
1797 return res
1798
1799 def _format_note(self, fdict):
1800 res = ''
1801 if fdict.get('ext') in ['f4f', 'f4m']:
1802 res += '(unsupported) '
1803 if fdict.get('language'):
1804 if res:
1805 res += ' '
1806 res += '[%s]' % fdict['language']
1807 if fdict.get('format_note') is not None:
1808 res += fdict['format_note'] + ' '
1809 if fdict.get('tbr') is not None:
1810 res += '%4dk ' % fdict['tbr']
1811 if fdict.get('container') is not None:
1812 if res:
1813 res += ', '
1814 res += '%s container' % fdict['container']
1815 if (fdict.get('vcodec') is not None and
1816 fdict.get('vcodec') != 'none'):
1817 if res:
1818 res += ', '
1819 res += fdict['vcodec']
1820 if fdict.get('vbr') is not None:
1821 res += '@'
1822 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
1823 res += 'video@'
1824 if fdict.get('vbr') is not None:
1825 res += '%4dk' % fdict['vbr']
1826 if fdict.get('fps') is not None:
1827 res += ', %sfps' % fdict['fps']
1828 if fdict.get('acodec') is not None:
1829 if res:
1830 res += ', '
1831 if fdict['acodec'] == 'none':
1832 res += 'video only'
1833 else:
1834 res += '%-5s' % fdict['acodec']
1835 elif fdict.get('abr') is not None:
1836 if res:
1837 res += ', '
1838 res += 'audio'
1839 if fdict.get('abr') is not None:
1840 res += '@%3dk' % fdict['abr']
1841 if fdict.get('asr') is not None:
1842 res += ' (%5dHz)' % fdict['asr']
1843 if fdict.get('filesize') is not None:
1844 if res:
1845 res += ', '
1846 res += format_bytes(fdict['filesize'])
1847 elif fdict.get('filesize_approx') is not None:
1848 if res:
1849 res += ', '
1850 res += '~' + format_bytes(fdict['filesize_approx'])
1851 return res
1852
1853 def list_formats(self, info_dict):
1854 formats = info_dict.get('formats', [info_dict])
1855 table = [
1856 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
1857 for f in formats
1858 if f.get('preference') is None or f['preference'] >= -1000]
1859 if len(formats) > 1:
1860 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
1861
1862 header_line = ['format code', 'extension', 'resolution', 'note']
1863 self.to_screen(
1864 '[info] Available formats for %s:\n%s' %
1865 (info_dict['id'], render_table(header_line, table)))
1866
1867 def list_thumbnails(self, info_dict):
1868 thumbnails = info_dict.get('thumbnails')
1869 if not thumbnails:
1870 tn_url = info_dict.get('thumbnail')
1871 if tn_url:
1872 thumbnails = [{'id': '0', 'url': tn_url}]
1873 else:
1874 self.to_screen(
1875 '[info] No thumbnails present for %s' % info_dict['id'])
1876 return
1877
1878 self.to_screen(
1879 '[info] Thumbnails for %s:' % info_dict['id'])
1880 self.to_screen(render_table(
1881 ['ID', 'width', 'height', 'URL'],
1882 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
1883
1884 def list_subtitles(self, video_id, subtitles, name='subtitles'):
1885 if not subtitles:
1886 self.to_screen('%s has no %s' % (video_id, name))
1887 return
1888 self.to_screen(
1889 'Available %s for %s:' % (name, video_id))
1890 self.to_screen(render_table(
1891 ['Language', 'formats'],
1892 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
1893 for lang, formats in subtitles.items()]))
1894
1895 def urlopen(self, req):
1896 """ Start an HTTP download """
1897 if isinstance(req, compat_basestring):
1898 req = sanitized_Request(req)
1899 return self._opener.open(req, timeout=self._socket_timeout)
1900
1901 def print_debug_header(self):
1902 if not self.params.get('verbose'):
1903 return
1904
1905 if type('') is not compat_str:
1906 # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326)
1907 self.report_warning(
1908 'Your Python is broken! Update to a newer and supported version')
1909
1910 stdout_encoding = getattr(
1911 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
1912 encoding_str = (
1913 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
1914 locale.getpreferredencoding(),
1915 sys.getfilesystemencoding(),
1916 stdout_encoding,
1917 self.get_encoding()))
1918 write_string(encoding_str, encoding=None)
1919
1920 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
1921 try:
1922 sp = subprocess.Popen(
1923 ['git', 'rev-parse', '--short', 'HEAD'],
1924 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
1925 cwd=os.path.dirname(os.path.abspath(__file__)))
1926 out, err = sp.communicate()
1927 out = out.decode().strip()
1928 if re.match('[0-9a-f]+', out):
1929 self._write_string('[debug] Git HEAD: ' + out + '\n')
1930 except Exception:
1931 try:
1932 sys.exc_clear()
1933 except Exception:
1934 pass
1935 self._write_string('[debug] Python version %s - %s\n' % (
1936 platform.python_version(), platform_name()))
1937
1938 exe_versions = FFmpegPostProcessor.get_versions(self)
1939 exe_versions['rtmpdump'] = rtmpdump_version()
1940 exe_str = ', '.join(
1941 '%s %s' % (exe, v)
1942 for exe, v in sorted(exe_versions.items())
1943 if v
1944 )
1945 if not exe_str:
1946 exe_str = 'none'
1947 self._write_string('[debug] exe versions: %s\n' % exe_str)
1948
1949 proxy_map = {}
1950 for handler in self._opener.handlers:
1951 if hasattr(handler, 'proxies'):
1952 proxy_map.update(handler.proxies)
1953 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
1954
1955 if self.params.get('call_home', False):
1956 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
1957 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
1958 latest_version = self.urlopen(
1959 'https://yt-dl.org/latest/version').read().decode('utf-8')
1960 if version_tuple(latest_version) > version_tuple(__version__):
1961 self.report_warning(
1962 'You are using an outdated version (newest version: %s)! '
1963 'See https://yt-dl.org/update if you need help updating.' %
1964 latest_version)
1965
1966 def _setup_opener(self):
1967 timeout_val = self.params.get('socket_timeout')
1968 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
1969
1970 opts_cookiefile = self.params.get('cookiefile')
1971 opts_proxy = self.params.get('proxy')
1972
1973 if opts_cookiefile is None:
1974 self.cookiejar = compat_cookiejar.CookieJar()
1975 else:
1976 self.cookiejar = compat_cookiejar.MozillaCookieJar(
1977 opts_cookiefile)
1978 if os.access(opts_cookiefile, os.R_OK):
1979 self.cookiejar.load()
1980
1981 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
1982 if opts_proxy is not None:
1983 if opts_proxy == '':
1984 proxies = {}
1985 else:
1986 proxies = {'http': opts_proxy, 'https': opts_proxy}
1987 else:
1988 proxies = compat_urllib_request.getproxies()
1989 # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
1990 if 'http' in proxies and 'https' not in proxies:
1991 proxies['https'] = proxies['http']
1992 proxy_handler = PerRequestProxyHandler(proxies)
1993
1994 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
1995 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
1996 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
1997 data_handler = compat_urllib_request_DataHandler()
1998
1999 # When passing our own FileHandler instance, build_opener won't add the
2000 # default FileHandler and allows us to disable the file protocol, which
2001 # can be used for malicious purposes (see
2002 # https://github.com/rg3/youtube-dl/issues/8227)
2003 file_handler = compat_urllib_request.FileHandler()
2004
2005 def file_open(*args, **kwargs):
2006 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2007 file_handler.file_open = file_open
2008
2009 opener = compat_urllib_request.build_opener(
2010 proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
2011
2012 # Delete the default user-agent header, which would otherwise apply in
2013 # cases where our custom HTTP handler doesn't come into play
2014 # (See https://github.com/rg3/youtube-dl/issues/1309 for details)
2015 opener.addheaders = []
2016 self._opener = opener
2017
2018 def encode(self, s):
2019 if isinstance(s, bytes):
2020 return s # Already encoded
2021
2022 try:
2023 return s.encode(self.get_encoding())
2024 except UnicodeEncodeError as err:
2025 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2026 raise
2027
2028 def get_encoding(self):
2029 encoding = self.params.get('encoding')
2030 if encoding is None:
2031 encoding = preferredencoding()
2032 return encoding
2033
2034 def _write_thumbnails(self, info_dict, filename):
2035 if self.params.get('writethumbnail', False):
2036 thumbnails = info_dict.get('thumbnails')
2037 if thumbnails:
2038 thumbnails = [thumbnails[-1]]
2039 elif self.params.get('write_all_thumbnails', False):
2040 thumbnails = info_dict.get('thumbnails')
2041 else:
2042 return
2043
2044 if not thumbnails:
2045 # No thumbnails present, so return immediately
2046 return
2047
2048 for t in thumbnails:
2049 thumb_ext = determine_ext(t['url'], 'jpg')
2050 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2051 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2052 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2053
2054 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2055 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2056 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2057 else:
2058 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2059 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2060 try:
2061 uf = self.urlopen(t['url'])
2062 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2063 shutil.copyfileobj(uf, thumbf)
2064 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2065 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2066 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2067 self.report_warning('Unable to download thumbnail "%s": %s' %
2068 (t['url'], error_to_compat_str(err)))