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