]>
Commit | Line | Data |
---|---|---|
1 | #!/usr/bin/env python3 | |
2 | # coding: utf-8 | |
3 | ||
4 | from __future__ import absolute_import, unicode_literals | |
5 | ||
6 | import collections | |
7 | import contextlib | |
8 | import datetime | |
9 | import errno | |
10 | import fileinput | |
11 | import functools | |
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 sys | |
23 | import tempfile | |
24 | import time | |
25 | import tokenize | |
26 | import traceback | |
27 | import random | |
28 | import unicodedata | |
29 | ||
30 | from enum import Enum | |
31 | from string import ascii_letters | |
32 | ||
33 | from .compat import ( | |
34 | compat_basestring, | |
35 | compat_brotli, | |
36 | compat_get_terminal_size, | |
37 | compat_kwargs, | |
38 | compat_numeric_types, | |
39 | compat_os_name, | |
40 | compat_pycrypto_AES, | |
41 | compat_shlex_quote, | |
42 | compat_str, | |
43 | compat_tokenize_tokenize, | |
44 | compat_urllib_error, | |
45 | compat_urllib_request, | |
46 | compat_urllib_request_DataHandler, | |
47 | windows_enable_vt_mode, | |
48 | ) | |
49 | from .cookies import load_cookies | |
50 | from .utils import ( | |
51 | age_restricted, | |
52 | args_to_str, | |
53 | ContentTooShortError, | |
54 | date_from_str, | |
55 | DateRange, | |
56 | DEFAULT_OUTTMPL, | |
57 | determine_ext, | |
58 | determine_protocol, | |
59 | DownloadCancelled, | |
60 | DownloadError, | |
61 | encode_compat_str, | |
62 | encodeFilename, | |
63 | EntryNotInPlaylist, | |
64 | error_to_compat_str, | |
65 | ExistingVideoReached, | |
66 | expand_path, | |
67 | ExtractorError, | |
68 | float_or_none, | |
69 | format_bytes, | |
70 | format_field, | |
71 | format_decimal_suffix, | |
72 | formatSeconds, | |
73 | GeoRestrictedError, | |
74 | get_domain, | |
75 | HEADRequest, | |
76 | InAdvancePagedList, | |
77 | int_or_none, | |
78 | iri_to_uri, | |
79 | ISO3166Utils, | |
80 | join_nonempty, | |
81 | LazyList, | |
82 | LINK_TEMPLATES, | |
83 | locked_file, | |
84 | make_dir, | |
85 | make_HTTPS_handler, | |
86 | MaxDownloadsReached, | |
87 | merge_headers, | |
88 | network_exceptions, | |
89 | number_of_digits, | |
90 | orderedSet, | |
91 | OUTTMPL_TYPES, | |
92 | PagedList, | |
93 | parse_filesize, | |
94 | PerRequestProxyHandler, | |
95 | platform_name, | |
96 | Popen, | |
97 | POSTPROCESS_WHEN, | |
98 | PostProcessingError, | |
99 | preferredencoding, | |
100 | prepend_extension, | |
101 | ReExtractInfo, | |
102 | register_socks_protocols, | |
103 | RejectedVideoReached, | |
104 | remove_terminal_sequences, | |
105 | render_table, | |
106 | replace_extension, | |
107 | SameFileError, | |
108 | sanitize_filename, | |
109 | sanitize_path, | |
110 | sanitize_url, | |
111 | sanitized_Request, | |
112 | std_headers, | |
113 | STR_FORMAT_RE_TMPL, | |
114 | STR_FORMAT_TYPES, | |
115 | str_or_none, | |
116 | strftime_or_none, | |
117 | subtitles_filename, | |
118 | supports_terminal_sequences, | |
119 | timetuple_from_msec, | |
120 | to_high_limit_path, | |
121 | traverse_obj, | |
122 | try_get, | |
123 | UnavailableVideoError, | |
124 | url_basename, | |
125 | variadic, | |
126 | version_tuple, | |
127 | write_json_file, | |
128 | write_string, | |
129 | YoutubeDLCookieProcessor, | |
130 | YoutubeDLHandler, | |
131 | YoutubeDLRedirectHandler, | |
132 | ) | |
133 | from .cache import Cache | |
134 | from .minicurses import format_text | |
135 | from .extractor import ( | |
136 | gen_extractor_classes, | |
137 | get_info_extractor, | |
138 | _LAZY_LOADER, | |
139 | _PLUGIN_CLASSES as plugin_extractors | |
140 | ) | |
141 | from .extractor.openload import PhantomJSwrapper | |
142 | from .downloader import ( | |
143 | FFmpegFD, | |
144 | get_suitable_downloader, | |
145 | shorten_protocol_name | |
146 | ) | |
147 | from .downloader.rtmp import rtmpdump_version | |
148 | from .postprocessor import ( | |
149 | get_postprocessor, | |
150 | EmbedThumbnailPP, | |
151 | FFmpegFixupDuplicateMoovPP, | |
152 | FFmpegFixupDurationPP, | |
153 | FFmpegFixupM3u8PP, | |
154 | FFmpegFixupM4aPP, | |
155 | FFmpegFixupStretchedPP, | |
156 | FFmpegFixupTimestampPP, | |
157 | FFmpegMergerPP, | |
158 | FFmpegPostProcessor, | |
159 | MoveFilesAfterDownloadPP, | |
160 | _PLUGIN_CLASSES as plugin_postprocessors | |
161 | ) | |
162 | from .update import detect_variant | |
163 | from .version import __version__, RELEASE_GIT_HEAD | |
164 | ||
165 | if compat_os_name == 'nt': | |
166 | import ctypes | |
167 | ||
168 | ||
169 | class YoutubeDL(object): | |
170 | """YoutubeDL class. | |
171 | ||
172 | YoutubeDL objects are the ones responsible of downloading the | |
173 | actual video file and writing it to disk if the user has requested | |
174 | it, among some other tasks. In most cases there should be one per | |
175 | program. As, given a video URL, the downloader doesn't know how to | |
176 | extract all the needed information, task that InfoExtractors do, it | |
177 | has to pass the URL to one of them. | |
178 | ||
179 | For this, YoutubeDL objects have a method that allows | |
180 | InfoExtractors to be registered in a given order. When it is passed | |
181 | a URL, the YoutubeDL object handles it to the first InfoExtractor it | |
182 | finds that reports being able to handle it. The InfoExtractor extracts | |
183 | all the information about the video or videos the URL refers to, and | |
184 | YoutubeDL process the extracted information, possibly using a File | |
185 | Downloader to download the video. | |
186 | ||
187 | YoutubeDL objects accept a lot of parameters. In order not to saturate | |
188 | the object constructor with arguments, it receives a dictionary of | |
189 | options instead. These options are available through the params | |
190 | attribute for the InfoExtractors to use. The YoutubeDL also | |
191 | registers itself as the downloader in charge for the InfoExtractors | |
192 | that are added to it, so this is a "mutual registration". | |
193 | ||
194 | Available options: | |
195 | ||
196 | username: Username for authentication purposes. | |
197 | password: Password for authentication purposes. | |
198 | videopassword: Password for accessing a video. | |
199 | ap_mso: Adobe Pass multiple-system operator identifier. | |
200 | ap_username: Multiple-system operator account username. | |
201 | ap_password: Multiple-system operator account password. | |
202 | usenetrc: Use netrc for authentication instead. | |
203 | verbose: Print additional info to stdout. | |
204 | quiet: Do not print messages to stdout. | |
205 | no_warnings: Do not print out anything for warnings. | |
206 | forceprint: A dict with keys WHEN mapped to a list of templates to | |
207 | print to stdout. The allowed keys are video or any of the | |
208 | items in utils.POSTPROCESS_WHEN. | |
209 | For compatibility, a single list is also accepted | |
210 | print_to_file: A dict with keys WHEN (same as forceprint) mapped to | |
211 | a list of tuples with (template, filename) | |
212 | forceurl: Force printing final URL. (Deprecated) | |
213 | forcetitle: Force printing title. (Deprecated) | |
214 | forceid: Force printing ID. (Deprecated) | |
215 | forcethumbnail: Force printing thumbnail URL. (Deprecated) | |
216 | forcedescription: Force printing description. (Deprecated) | |
217 | forcefilename: Force printing final filename. (Deprecated) | |
218 | forceduration: Force printing duration. (Deprecated) | |
219 | forcejson: Force printing info_dict as JSON. | |
220 | dump_single_json: Force printing the info_dict of the whole playlist | |
221 | (or video) as a single JSON line. | |
222 | force_write_download_archive: Force writing download archive regardless | |
223 | of 'skip_download' or 'simulate'. | |
224 | simulate: Do not download the video files. If unset (or None), | |
225 | simulate only if listsubtitles, listformats or list_thumbnails is used | |
226 | format: Video format code. see "FORMAT SELECTION" for more details. | |
227 | You can also pass a function. The function takes 'ctx' as | |
228 | argument and returns the formats to download. | |
229 | See "build_format_selector" for an implementation | |
230 | allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded. | |
231 | ignore_no_formats_error: Ignore "No video formats" error. Usefull for | |
232 | extracting metadata even if the video is not actually | |
233 | available for download (experimental) | |
234 | format_sort: A list of fields by which to sort the video formats. | |
235 | See "Sorting Formats" for more details. | |
236 | format_sort_force: Force the given format_sort. see "Sorting Formats" | |
237 | for more details. | |
238 | prefer_free_formats: Whether to prefer video formats with free containers | |
239 | over non-free ones of same quality. | |
240 | allow_multiple_video_streams: Allow multiple video streams to be merged | |
241 | into a single file | |
242 | allow_multiple_audio_streams: Allow multiple audio streams to be merged | |
243 | into a single file | |
244 | check_formats Whether to test if the formats are downloadable. | |
245 | Can be True (check all), False (check none), | |
246 | 'selected' (check selected formats), | |
247 | or None (check only if requested by extractor) | |
248 | paths: Dictionary of output paths. The allowed keys are 'home' | |
249 | 'temp' and the keys of OUTTMPL_TYPES (in utils.py) | |
250 | outtmpl: Dictionary of templates for output names. Allowed keys | |
251 | are 'default' and the keys of OUTTMPL_TYPES (in utils.py). | |
252 | For compatibility with youtube-dl, a single string can also be used | |
253 | outtmpl_na_placeholder: Placeholder for unavailable meta fields. | |
254 | restrictfilenames: Do not allow "&" and spaces in file names | |
255 | trim_file_name: Limit length of filename (extension excluded) | |
256 | windowsfilenames: Force the filenames to be windows compatible | |
257 | ignoreerrors: Do not stop on download/postprocessing errors. | |
258 | Can be 'only_download' to ignore only download errors. | |
259 | Default is 'only_download' for CLI, but False for API | |
260 | skip_playlist_after_errors: Number of allowed failures until the rest of | |
261 | the playlist is skipped | |
262 | force_generic_extractor: Force downloader to use the generic extractor | |
263 | overwrites: Overwrite all video and metadata files if True, | |
264 | overwrite only non-video files if None | |
265 | and don't overwrite any file if False | |
266 | For compatibility with youtube-dl, | |
267 | "nooverwrites" may also be used instead | |
268 | playliststart: Playlist item to start at. | |
269 | playlistend: Playlist item to end at. | |
270 | playlist_items: Specific indices of playlist to download. | |
271 | playlistreverse: Download playlist items in reverse order. | |
272 | playlistrandom: Download playlist items in random order. | |
273 | matchtitle: Download only matching titles. | |
274 | rejecttitle: Reject downloads for matching titles. | |
275 | logger: Log messages to a logging.Logger instance. | |
276 | logtostderr: Log messages to stderr instead of stdout. | |
277 | consoletitle: Display progress in console window's titlebar. | |
278 | writedescription: Write the video description to a .description file | |
279 | writeinfojson: Write the video description to a .info.json file | |
280 | clean_infojson: Remove private fields from the infojson | |
281 | getcomments: Extract video comments. This will not be written to disk | |
282 | unless writeinfojson is also given | |
283 | writeannotations: Write the video annotations to a .annotations.xml file | |
284 | writethumbnail: Write the thumbnail image to a file | |
285 | allow_playlist_files: Whether to write playlists' description, infojson etc | |
286 | also to disk when using the 'write*' options | |
287 | write_all_thumbnails: Write all thumbnail formats to files | |
288 | writelink: Write an internet shortcut file, depending on the | |
289 | current platform (.url/.webloc/.desktop) | |
290 | writeurllink: Write a Windows internet shortcut file (.url) | |
291 | writewebloclink: Write a macOS internet shortcut file (.webloc) | |
292 | writedesktoplink: Write a Linux internet shortcut file (.desktop) | |
293 | writesubtitles: Write the video subtitles to a file | |
294 | writeautomaticsub: Write the automatically generated subtitles to a file | |
295 | allsubtitles: Deprecated - Use subtitleslangs = ['all'] | |
296 | Downloads all the subtitles of the video | |
297 | (requires writesubtitles or writeautomaticsub) | |
298 | listsubtitles: Lists all available subtitles for the video | |
299 | subtitlesformat: The format code for subtitles | |
300 | subtitleslangs: List of languages of the subtitles to download (can be regex). | |
301 | The list may contain "all" to refer to all the available | |
302 | subtitles. The language can be prefixed with a "-" to | |
303 | exclude it from the requested languages. Eg: ['all', '-live_chat'] | |
304 | keepvideo: Keep the video file after post-processing | |
305 | daterange: A DateRange object, download only if the upload_date is in the range. | |
306 | skip_download: Skip the actual download of the video file | |
307 | cachedir: Location of the cache files in the filesystem. | |
308 | False to disable filesystem cache. | |
309 | noplaylist: Download single video instead of a playlist if in doubt. | |
310 | age_limit: An integer representing the user's age in years. | |
311 | Unsuitable videos for the given age are skipped. | |
312 | min_views: An integer representing the minimum view count the video | |
313 | must have in order to not be skipped. | |
314 | Videos without view count information are always | |
315 | downloaded. None for no limit. | |
316 | max_views: An integer representing the maximum view count. | |
317 | Videos that are more popular than that are not | |
318 | downloaded. | |
319 | Videos without view count information are always | |
320 | downloaded. None for no limit. | |
321 | download_archive: File name of a file where all downloads are recorded. | |
322 | Videos already present in the file are not downloaded | |
323 | again. | |
324 | break_on_existing: Stop the download process after attempting to download a | |
325 | file that is in the archive. | |
326 | break_on_reject: Stop the download process when encountering a video that | |
327 | has been filtered out. | |
328 | break_per_url: Whether break_on_reject and break_on_existing | |
329 | should act on each input URL as opposed to for the entire queue | |
330 | cookiefile: File name where cookies should be read from and dumped to | |
331 | cookiesfrombrowser: A tuple containing the name of the browser, the profile | |
332 | name/pathfrom where cookies are loaded, and the name of the | |
333 | keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT') | |
334 | legacyserverconnect: Explicitly allow HTTPS connection to servers that do not | |
335 | support RFC 5746 secure renegotiation | |
336 | nocheckcertificate: Do not verify SSL certificates | |
337 | prefer_insecure: Use HTTP instead of HTTPS to retrieve information. | |
338 | At the moment, this is only supported by YouTube. | |
339 | http_headers: A dictionary of custom headers to be used for all requests | |
340 | proxy: URL of the proxy server to use | |
341 | geo_verification_proxy: URL of the proxy to use for IP address verification | |
342 | on geo-restricted sites. | |
343 | socket_timeout: Time to wait for unresponsive hosts, in seconds | |
344 | bidi_workaround: Work around buggy terminals without bidirectional text | |
345 | support, using fridibi | |
346 | debug_printtraffic:Print out sent and received HTTP traffic | |
347 | include_ads: Download ads as well (deprecated) | |
348 | default_search: Prepend this string if an input url is not valid. | |
349 | 'auto' for elaborate guessing | |
350 | encoding: Use this encoding instead of the system-specified. | |
351 | extract_flat: Do not resolve URLs, return the immediate result. | |
352 | Pass in 'in_playlist' to only show this behavior for | |
353 | playlist items. | |
354 | wait_for_video: If given, wait for scheduled streams to become available. | |
355 | The value should be a tuple containing the range | |
356 | (min_secs, max_secs) to wait between retries | |
357 | postprocessors: A list of dictionaries, each with an entry | |
358 | * key: The name of the postprocessor. See | |
359 | yt_dlp/postprocessor/__init__.py for a list. | |
360 | * when: When to run the postprocessor. Allowed values are | |
361 | the entries of utils.POSTPROCESS_WHEN | |
362 | Assumed to be 'post_process' if not given | |
363 | post_hooks: Deprecated - Register a custom postprocessor instead | |
364 | A list of functions that get called as the final step | |
365 | for each video file, after all postprocessors have been | |
366 | called. The filename will be passed as the only argument. | |
367 | progress_hooks: A list of functions that get called on download | |
368 | progress, with a dictionary with the entries | |
369 | * status: One of "downloading", "error", or "finished". | |
370 | Check this first and ignore unknown values. | |
371 | * info_dict: The extracted info_dict | |
372 | ||
373 | If status is one of "downloading", or "finished", the | |
374 | following properties may also be present: | |
375 | * filename: The final filename (always present) | |
376 | * tmpfilename: The filename we're currently writing to | |
377 | * downloaded_bytes: Bytes on disk | |
378 | * total_bytes: Size of the whole file, None if unknown | |
379 | * total_bytes_estimate: Guess of the eventual file size, | |
380 | None if unavailable. | |
381 | * elapsed: The number of seconds since download started. | |
382 | * eta: The estimated time in seconds, None if unknown | |
383 | * speed: The download speed in bytes/second, None if | |
384 | unknown | |
385 | * fragment_index: The counter of the currently | |
386 | downloaded video fragment. | |
387 | * fragment_count: The number of fragments (= individual | |
388 | files that will be merged) | |
389 | ||
390 | Progress hooks are guaranteed to be called at least once | |
391 | (with status "finished") if the download is successful. | |
392 | postprocessor_hooks: A list of functions that get called on postprocessing | |
393 | progress, with a dictionary with the entries | |
394 | * status: One of "started", "processing", or "finished". | |
395 | Check this first and ignore unknown values. | |
396 | * postprocessor: Name of the postprocessor | |
397 | * info_dict: The extracted info_dict | |
398 | ||
399 | Progress hooks are guaranteed to be called at least twice | |
400 | (with status "started" and "finished") if the processing is successful. | |
401 | merge_output_format: Extension to use when merging formats. | |
402 | final_ext: Expected final extension; used to detect when the file was | |
403 | already downloaded and converted | |
404 | fixup: Automatically correct known faults of the file. | |
405 | One of: | |
406 | - "never": do nothing | |
407 | - "warn": only emit a warning | |
408 | - "detect_or_warn": check whether we can do anything | |
409 | about it, warn otherwise (default) | |
410 | source_address: Client-side IP address to bind to. | |
411 | call_home: Boolean, true iff we are allowed to contact the | |
412 | yt-dlp servers for debugging. (BROKEN) | |
413 | sleep_interval_requests: Number of seconds to sleep between requests | |
414 | during extraction | |
415 | sleep_interval: Number of seconds to sleep before each download when | |
416 | used alone or a lower bound of a range for randomized | |
417 | sleep before each download (minimum possible number | |
418 | of seconds to sleep) when used along with | |
419 | max_sleep_interval. | |
420 | max_sleep_interval:Upper bound of a range for randomized sleep before each | |
421 | download (maximum possible number of seconds to sleep). | |
422 | Must only be used along with sleep_interval. | |
423 | Actual sleep time will be a random float from range | |
424 | [sleep_interval; max_sleep_interval]. | |
425 | sleep_interval_subtitles: Number of seconds to sleep before each subtitle download | |
426 | listformats: Print an overview of available video formats and exit. | |
427 | list_thumbnails: Print a table of all thumbnails and exit. | |
428 | match_filter: A function that gets called with the info_dict of | |
429 | every video. | |
430 | If it returns a message, the video is ignored. | |
431 | If it returns None, the video is downloaded. | |
432 | match_filter_func in utils.py is one example for this. | |
433 | no_color: Do not emit color codes in output. | |
434 | geo_bypass: Bypass geographic restriction via faking X-Forwarded-For | |
435 | HTTP header | |
436 | geo_bypass_country: | |
437 | Two-letter ISO 3166-2 country code that will be used for | |
438 | explicit geographic restriction bypassing via faking | |
439 | X-Forwarded-For HTTP header | |
440 | geo_bypass_ip_block: | |
441 | IP range in CIDR notation that will be used similarly to | |
442 | geo_bypass_country | |
443 | ||
444 | The following options determine which downloader is picked: | |
445 | external_downloader: A dictionary of protocol keys and the executable of the | |
446 | external downloader to use for it. The allowed protocols | |
447 | are default|http|ftp|m3u8|dash|rtsp|rtmp|mms. | |
448 | Set the value to 'native' to use the native downloader | |
449 | hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'} | |
450 | or {'m3u8': 'ffmpeg'} instead. | |
451 | Use the native HLS downloader instead of ffmpeg/avconv | |
452 | if True, otherwise use ffmpeg/avconv if False, otherwise | |
453 | use downloader suggested by extractor if None. | |
454 | compat_opts: Compatibility options. See "Differences in default behavior". | |
455 | The following options do not work when used through the API: | |
456 | filename, abort-on-error, multistreams, no-live-chat, format-sort | |
457 | no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json. | |
458 | Refer __init__.py for their implementation | |
459 | progress_template: Dictionary of templates for progress outputs. | |
460 | Allowed keys are 'download', 'postprocess', | |
461 | 'download-title' (console title) and 'postprocess-title'. | |
462 | The template is mapped on a dictionary with keys 'progress' and 'info' | |
463 | ||
464 | The following parameters are not used by YoutubeDL itself, they are used by | |
465 | the downloader (see yt_dlp/downloader/common.py): | |
466 | nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize, | |
467 | max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries, | |
468 | continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size, | |
469 | external_downloader_args, concurrent_fragment_downloads. | |
470 | ||
471 | The following options are used by the post processors: | |
472 | prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available, | |
473 | otherwise prefer ffmpeg. (avconv support is deprecated) | |
474 | ffmpeg_location: Location of the ffmpeg/avconv binary; either the path | |
475 | to the binary or its containing directory. | |
476 | postprocessor_args: A dictionary of postprocessor/executable keys (in lower case) | |
477 | and a list of additional command-line arguments for the | |
478 | postprocessor/executable. The dict can also have "PP+EXE" keys | |
479 | which are used when the given exe is used by the given PP. | |
480 | Use 'default' as the name for arguments to passed to all PP | |
481 | For compatibility with youtube-dl, a single list of args | |
482 | can also be used | |
483 | ||
484 | The following options are used by the extractors: | |
485 | extractor_retries: Number of times to retry for known errors | |
486 | dynamic_mpd: Whether to process dynamic DASH manifests (default: True) | |
487 | hls_split_discontinuity: Split HLS playlists to different formats at | |
488 | discontinuities such as ad breaks (default: False) | |
489 | extractor_args: A dictionary of arguments to be passed to the extractors. | |
490 | See "EXTRACTOR ARGUMENTS" for details. | |
491 | Eg: {'youtube': {'skip': ['dash', 'hls']}} | |
492 | mark_watched: Mark videos watched (even with --simulate). Only for YouTube | |
493 | youtube_include_dash_manifest: Deprecated - Use extractor_args instead. | |
494 | If True (default), DASH manifests and related | |
495 | data will be downloaded and processed by extractor. | |
496 | You can reduce network I/O by disabling it if you don't | |
497 | care about DASH. (only for youtube) | |
498 | youtube_include_hls_manifest: Deprecated - Use extractor_args instead. | |
499 | If True (default), HLS manifests and related | |
500 | data will be downloaded and processed by extractor. | |
501 | You can reduce network I/O by disabling it if you don't | |
502 | care about HLS. (only for youtube) | |
503 | """ | |
504 | ||
505 | _NUMERIC_FIELDS = set(( | |
506 | 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx', | |
507 | 'timestamp', 'release_timestamp', | |
508 | 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count', | |
509 | 'average_rating', 'comment_count', 'age_limit', | |
510 | 'start_time', 'end_time', | |
511 | 'chapter_number', 'season_number', 'episode_number', | |
512 | 'track_number', 'disc_number', 'release_year', | |
513 | )) | |
514 | ||
515 | _format_selection_exts = { | |
516 | 'audio': {'m4a', 'mp3', 'ogg', 'aac'}, | |
517 | 'video': {'mp4', 'flv', 'webm', '3gp'}, | |
518 | 'storyboards': {'mhtml'}, | |
519 | } | |
520 | ||
521 | params = None | |
522 | _ies = {} | |
523 | _pps = {k: [] for k in POSTPROCESS_WHEN} | |
524 | _printed_messages = set() | |
525 | _first_webpage_request = True | |
526 | _download_retcode = None | |
527 | _num_downloads = None | |
528 | _playlist_level = 0 | |
529 | _playlist_urls = set() | |
530 | _screen_file = None | |
531 | ||
532 | def __init__(self, params=None, auto_init=True): | |
533 | """Create a FileDownloader object with the given options. | |
534 | @param auto_init Whether to load the default extractors and print header (if verbose). | |
535 | Set to 'no_verbose_header' to not print the header | |
536 | """ | |
537 | if params is None: | |
538 | params = {} | |
539 | self._ies = {} | |
540 | self._ies_instances = {} | |
541 | self._pps = {k: [] for k in POSTPROCESS_WHEN} | |
542 | self._printed_messages = set() | |
543 | self._first_webpage_request = True | |
544 | self._post_hooks = [] | |
545 | self._progress_hooks = [] | |
546 | self._postprocessor_hooks = [] | |
547 | self._download_retcode = 0 | |
548 | self._num_downloads = 0 | |
549 | self._num_videos = 0 | |
550 | self.params = params | |
551 | self.cache = Cache(self) | |
552 | ||
553 | windows_enable_vt_mode() | |
554 | self._out_files = { | |
555 | 'error': sys.stderr, | |
556 | 'print': sys.stderr if self.params.get('logtostderr') else sys.stdout, | |
557 | 'console': None if compat_os_name == 'nt' else next( | |
558 | filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None) | |
559 | } | |
560 | self._out_files['screen'] = sys.stderr if self.params.get('quiet') else self._out_files['print'] | |
561 | self._allow_colors = { | |
562 | type_: not self.params.get('no_color') and supports_terminal_sequences(self._out_files[type_]) | |
563 | for type_ in ('screen', 'error') | |
564 | } | |
565 | ||
566 | if sys.version_info < (3, 6): | |
567 | self.report_warning( | |
568 | 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2]) | |
569 | ||
570 | if self.params.get('allow_unplayable_formats'): | |
571 | self.report_warning( | |
572 | f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. ' | |
573 | 'This is a developer option intended for debugging. \n' | |
574 | ' If you experience any issues while using this option, ' | |
575 | f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report') | |
576 | ||
577 | def check_deprecated(param, option, suggestion): | |
578 | if self.params.get(param) is not None: | |
579 | self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion)) | |
580 | return True | |
581 | return False | |
582 | ||
583 | if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'): | |
584 | if self.params.get('geo_verification_proxy') is None: | |
585 | self.params['geo_verification_proxy'] = self.params['cn_verification_proxy'] | |
586 | ||
587 | check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"') | |
588 | check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"') | |
589 | check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"') | |
590 | ||
591 | for msg in self.params.get('_warnings', []): | |
592 | self.report_warning(msg) | |
593 | for msg in self.params.get('_deprecation_warnings', []): | |
594 | self.deprecation_warning(msg) | |
595 | ||
596 | if 'list-formats' in self.params.get('compat_opts', []): | |
597 | self.params['listformats_table'] = False | |
598 | ||
599 | if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None: | |
600 | # nooverwrites was unnecessarily changed to overwrites | |
601 | # in 0c3d0f51778b153f65c21906031c2e091fcfb641 | |
602 | # This ensures compatibility with both keys | |
603 | self.params['overwrites'] = not self.params['nooverwrites'] | |
604 | elif self.params.get('overwrites') is None: | |
605 | self.params.pop('overwrites', None) | |
606 | else: | |
607 | self.params['nooverwrites'] = not self.params['overwrites'] | |
608 | ||
609 | self.params.setdefault('forceprint', {}) | |
610 | self.params.setdefault('print_to_file', {}) | |
611 | ||
612 | # Compatibility with older syntax | |
613 | if not isinstance(params['forceprint'], dict): | |
614 | self.params['forceprint'] = {'video': params['forceprint']} | |
615 | ||
616 | if self.params.get('bidi_workaround', False): | |
617 | try: | |
618 | import pty | |
619 | master, slave = pty.openpty() | |
620 | width = compat_get_terminal_size().columns | |
621 | if width is None: | |
622 | width_args = [] | |
623 | else: | |
624 | width_args = ['-w', str(width)] | |
625 | sp_kwargs = dict( | |
626 | stdin=subprocess.PIPE, | |
627 | stdout=slave, | |
628 | stderr=self._out_files['error']) | |
629 | try: | |
630 | self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs) | |
631 | except OSError: | |
632 | self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs) | |
633 | self._output_channel = os.fdopen(master, 'rb') | |
634 | except OSError as ose: | |
635 | if ose.errno == errno.ENOENT: | |
636 | self.report_warning( | |
637 | 'Could not find fribidi executable, ignoring --bidi-workaround. ' | |
638 | 'Make sure that fribidi is an executable file in one of the directories in your $PATH.') | |
639 | else: | |
640 | raise | |
641 | ||
642 | if (sys.platform != 'win32' | |
643 | and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] | |
644 | and not self.params.get('restrictfilenames', False)): | |
645 | # Unicode filesystem API will throw errors (#1474, #13027) | |
646 | self.report_warning( | |
647 | 'Assuming --restrict-filenames since file system encoding ' | |
648 | 'cannot encode all characters. ' | |
649 | 'Set the LC_ALL environment variable to fix this.') | |
650 | self.params['restrictfilenames'] = True | |
651 | ||
652 | self.outtmpl_dict = self.parse_outtmpl() | |
653 | ||
654 | # Creating format selector here allows us to catch syntax errors before the extraction | |
655 | self.format_selector = ( | |
656 | self.params.get('format') if self.params.get('format') in (None, '-') | |
657 | else self.params['format'] if callable(self.params['format']) | |
658 | else self.build_format_selector(self.params['format'])) | |
659 | ||
660 | # Set http_headers defaults according to std_headers | |
661 | self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {})) | |
662 | ||
663 | self._setup_opener() | |
664 | ||
665 | if auto_init: | |
666 | if auto_init != 'no_verbose_header': | |
667 | self.print_debug_header() | |
668 | self.add_default_info_extractors() | |
669 | ||
670 | hooks = { | |
671 | 'post_hooks': self.add_post_hook, | |
672 | 'progress_hooks': self.add_progress_hook, | |
673 | 'postprocessor_hooks': self.add_postprocessor_hook, | |
674 | } | |
675 | for opt, fn in hooks.items(): | |
676 | for ph in self.params.get(opt, []): | |
677 | fn(ph) | |
678 | ||
679 | for pp_def_raw in self.params.get('postprocessors', []): | |
680 | pp_def = dict(pp_def_raw) | |
681 | when = pp_def.pop('when', 'post_process') | |
682 | self.add_post_processor( | |
683 | get_postprocessor(pp_def.pop('key'))(self, **compat_kwargs(pp_def)), | |
684 | when=when) | |
685 | ||
686 | register_socks_protocols() | |
687 | ||
688 | def preload_download_archive(fn): | |
689 | """Preload the archive, if any is specified""" | |
690 | if fn is None: | |
691 | return False | |
692 | self.write_debug(f'Loading archive file {fn!r}') | |
693 | try: | |
694 | with locked_file(fn, 'r', encoding='utf-8') as archive_file: | |
695 | for line in archive_file: | |
696 | self.archive.add(line.strip()) | |
697 | except IOError as ioe: | |
698 | if ioe.errno != errno.ENOENT: | |
699 | raise | |
700 | return False | |
701 | return True | |
702 | ||
703 | self.archive = set() | |
704 | preload_download_archive(self.params.get('download_archive')) | |
705 | ||
706 | def warn_if_short_id(self, argv): | |
707 | # short YouTube ID starting with dash? | |
708 | idxs = [ | |
709 | i for i, a in enumerate(argv) | |
710 | if re.match(r'^-[0-9A-Za-z_-]{10}$', a)] | |
711 | if idxs: | |
712 | correct_argv = ( | |
713 | ['yt-dlp'] | |
714 | + [a for i, a in enumerate(argv) if i not in idxs] | |
715 | + ['--'] + [argv[i] for i in idxs] | |
716 | ) | |
717 | self.report_warning( | |
718 | 'Long argument string detected. ' | |
719 | 'Use -- to separate parameters and URLs, like this:\n%s' % | |
720 | args_to_str(correct_argv)) | |
721 | ||
722 | def add_info_extractor(self, ie): | |
723 | """Add an InfoExtractor object to the end of the list.""" | |
724 | ie_key = ie.ie_key() | |
725 | self._ies[ie_key] = ie | |
726 | if not isinstance(ie, type): | |
727 | self._ies_instances[ie_key] = ie | |
728 | ie.set_downloader(self) | |
729 | ||
730 | def _get_info_extractor_class(self, ie_key): | |
731 | ie = self._ies.get(ie_key) | |
732 | if ie is None: | |
733 | ie = get_info_extractor(ie_key) | |
734 | self.add_info_extractor(ie) | |
735 | return ie | |
736 | ||
737 | def get_info_extractor(self, ie_key): | |
738 | """ | |
739 | Get an instance of an IE with name ie_key, it will try to get one from | |
740 | the _ies list, if there's no instance it will create a new one and add | |
741 | it to the extractor list. | |
742 | """ | |
743 | ie = self._ies_instances.get(ie_key) | |
744 | if ie is None: | |
745 | ie = get_info_extractor(ie_key)() | |
746 | self.add_info_extractor(ie) | |
747 | return ie | |
748 | ||
749 | def add_default_info_extractors(self): | |
750 | """ | |
751 | Add the InfoExtractors returned by gen_extractors to the end of the list | |
752 | """ | |
753 | for ie in gen_extractor_classes(): | |
754 | self.add_info_extractor(ie) | |
755 | ||
756 | def add_post_processor(self, pp, when='post_process'): | |
757 | """Add a PostProcessor object to the end of the chain.""" | |
758 | self._pps[when].append(pp) | |
759 | pp.set_downloader(self) | |
760 | ||
761 | def add_post_hook(self, ph): | |
762 | """Add the post hook""" | |
763 | self._post_hooks.append(ph) | |
764 | ||
765 | def add_progress_hook(self, ph): | |
766 | """Add the download progress hook""" | |
767 | self._progress_hooks.append(ph) | |
768 | ||
769 | def add_postprocessor_hook(self, ph): | |
770 | """Add the postprocessing progress hook""" | |
771 | self._postprocessor_hooks.append(ph) | |
772 | for pps in self._pps.values(): | |
773 | for pp in pps: | |
774 | pp.add_progress_hook(ph) | |
775 | ||
776 | def _bidi_workaround(self, message): | |
777 | if not hasattr(self, '_output_channel'): | |
778 | return message | |
779 | ||
780 | assert hasattr(self, '_output_process') | |
781 | assert isinstance(message, compat_str) | |
782 | line_count = message.count('\n') + 1 | |
783 | self._output_process.stdin.write((message + '\n').encode('utf-8')) | |
784 | self._output_process.stdin.flush() | |
785 | res = ''.join(self._output_channel.readline().decode('utf-8') | |
786 | for _ in range(line_count)) | |
787 | return res[:-len('\n')] | |
788 | ||
789 | def _write_string(self, message, out=None, only_once=False): | |
790 | if only_once: | |
791 | if message in self._printed_messages: | |
792 | return | |
793 | self._printed_messages.add(message) | |
794 | write_string(message, out=out, encoding=self.params.get('encoding')) | |
795 | ||
796 | def to_stdout(self, message, skip_eol=False, quiet=None): | |
797 | """Print message to stdout""" | |
798 | if quiet is not None: | |
799 | self.deprecation_warning('"ydl.to_stdout" no longer accepts the argument quiet. Use "ydl.to_screen" instead') | |
800 | self._write_string( | |
801 | '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')), | |
802 | self._out_files['print']) | |
803 | ||
804 | def to_screen(self, message, skip_eol=False, quiet=None): | |
805 | """Print message to screen if not in quiet mode""" | |
806 | if self.params.get('logger'): | |
807 | self.params['logger'].debug(message) | |
808 | return | |
809 | if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'): | |
810 | return | |
811 | self._write_string( | |
812 | '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')), | |
813 | self._out_files['screen']) | |
814 | ||
815 | def to_stderr(self, message, only_once=False): | |
816 | """Print message to stderr""" | |
817 | assert isinstance(message, compat_str) | |
818 | if self.params.get('logger'): | |
819 | self.params['logger'].error(message) | |
820 | else: | |
821 | self._write_string('%s\n' % self._bidi_workaround(message), self._out_files['error'], only_once=only_once) | |
822 | ||
823 | def _send_console_code(self, code): | |
824 | if compat_os_name == 'nt' or not self._out_files['console']: | |
825 | return | |
826 | self._write_string(code, self._out_files['console']) | |
827 | ||
828 | def to_console_title(self, message): | |
829 | if not self.params.get('consoletitle', False): | |
830 | return | |
831 | message = remove_terminal_sequences(message) | |
832 | if compat_os_name == 'nt': | |
833 | if ctypes.windll.kernel32.GetConsoleWindow(): | |
834 | # c_wchar_p() might not be necessary if `message` is | |
835 | # already of type unicode() | |
836 | ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message)) | |
837 | else: | |
838 | self._send_console_code(f'\033]0;{message}\007') | |
839 | ||
840 | def save_console_title(self): | |
841 | if not self.params.get('consoletitle') or self.params.get('simulate'): | |
842 | return | |
843 | self._send_console_code('\033[22;0t') # Save the title on stack | |
844 | ||
845 | def restore_console_title(self): | |
846 | if not self.params.get('consoletitle') or self.params.get('simulate'): | |
847 | return | |
848 | self._send_console_code('\033[23;0t') # Restore the title from stack | |
849 | ||
850 | def __enter__(self): | |
851 | self.save_console_title() | |
852 | return self | |
853 | ||
854 | def __exit__(self, *args): | |
855 | self.restore_console_title() | |
856 | ||
857 | if self.params.get('cookiefile') is not None: | |
858 | self.cookiejar.save(ignore_discard=True, ignore_expires=True) | |
859 | ||
860 | def trouble(self, message=None, tb=None, is_error=True): | |
861 | """Determine action to take when a download problem appears. | |
862 | ||
863 | Depending on if the downloader has been configured to ignore | |
864 | download errors or not, this method may throw an exception or | |
865 | not when errors are found, after printing the message. | |
866 | ||
867 | @param tb If given, is additional traceback information | |
868 | @param is_error Whether to raise error according to ignorerrors | |
869 | """ | |
870 | if message is not None: | |
871 | self.to_stderr(message) | |
872 | if self.params.get('verbose'): | |
873 | if tb is None: | |
874 | if sys.exc_info()[0]: # if .trouble has been called from an except block | |
875 | tb = '' | |
876 | if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]: | |
877 | tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info)) | |
878 | tb += encode_compat_str(traceback.format_exc()) | |
879 | else: | |
880 | tb_data = traceback.format_list(traceback.extract_stack()) | |
881 | tb = ''.join(tb_data) | |
882 | if tb: | |
883 | self.to_stderr(tb) | |
884 | if not is_error: | |
885 | return | |
886 | if not self.params.get('ignoreerrors'): | |
887 | if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]: | |
888 | exc_info = sys.exc_info()[1].exc_info | |
889 | else: | |
890 | exc_info = sys.exc_info() | |
891 | raise DownloadError(message, exc_info) | |
892 | self._download_retcode = 1 | |
893 | ||
894 | class Styles(Enum): | |
895 | HEADERS = 'yellow' | |
896 | EMPHASIS = 'light blue' | |
897 | ID = 'green' | |
898 | DELIM = 'blue' | |
899 | ERROR = 'red' | |
900 | WARNING = 'yellow' | |
901 | SUPPRESS = 'light black' | |
902 | ||
903 | def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False): | |
904 | if test_encoding: | |
905 | original_text = text | |
906 | # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711 | |
907 | encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii' | |
908 | text = text.encode(encoding, 'ignore').decode(encoding) | |
909 | if fallback is not None and text != original_text: | |
910 | text = fallback | |
911 | if isinstance(f, self.Styles): | |
912 | f = f.value | |
913 | return format_text(text, f) if allow_colors else text if fallback is None else fallback | |
914 | ||
915 | def _format_screen(self, *args, **kwargs): | |
916 | return self._format_text( | |
917 | self._out_files['screen'], self._allow_colors['screen'], *args, **kwargs) | |
918 | ||
919 | def _format_err(self, *args, **kwargs): | |
920 | return self._format_text( | |
921 | self._out_files['error'], self._allow_colors['error'], *args, **kwargs) | |
922 | ||
923 | def report_warning(self, message, only_once=False): | |
924 | ''' | |
925 | Print the message to stderr, it will be prefixed with 'WARNING:' | |
926 | If stderr is a tty file the 'WARNING:' will be colored | |
927 | ''' | |
928 | if self.params.get('logger') is not None: | |
929 | self.params['logger'].warning(message) | |
930 | else: | |
931 | if self.params.get('no_warnings'): | |
932 | return | |
933 | self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once) | |
934 | ||
935 | def deprecation_warning(self, message): | |
936 | if self.params.get('logger') is not None: | |
937 | self.params['logger'].warning('DeprecationWarning: {message}') | |
938 | else: | |
939 | self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True) | |
940 | ||
941 | def report_error(self, message, *args, **kwargs): | |
942 | ''' | |
943 | Do the same as trouble, but prefixes the message with 'ERROR:', colored | |
944 | in red if stderr is a tty file. | |
945 | ''' | |
946 | self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs) | |
947 | ||
948 | def write_debug(self, message, only_once=False): | |
949 | '''Log debug message or Print message to stderr''' | |
950 | if not self.params.get('verbose', False): | |
951 | return | |
952 | message = '[debug] %s' % message | |
953 | if self.params.get('logger'): | |
954 | self.params['logger'].debug(message) | |
955 | else: | |
956 | self.to_stderr(message, only_once) | |
957 | ||
958 | def report_file_already_downloaded(self, file_name): | |
959 | """Report file has already been fully downloaded.""" | |
960 | try: | |
961 | self.to_screen('[download] %s has already been downloaded' % file_name) | |
962 | except UnicodeEncodeError: | |
963 | self.to_screen('[download] The file has already been downloaded') | |
964 | ||
965 | def report_file_delete(self, file_name): | |
966 | """Report that existing file will be deleted.""" | |
967 | try: | |
968 | self.to_screen('Deleting existing file %s' % file_name) | |
969 | except UnicodeEncodeError: | |
970 | self.to_screen('Deleting existing file') | |
971 | ||
972 | def raise_no_formats(self, info, forced=False, *, msg=None): | |
973 | has_drm = info.get('__has_drm') | |
974 | ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg) | |
975 | msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!' | |
976 | if forced or not ignored: | |
977 | raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'], | |
978 | expected=has_drm or ignored or expected) | |
979 | else: | |
980 | self.report_warning(msg) | |
981 | ||
982 | def parse_outtmpl(self): | |
983 | outtmpl_dict = self.params.get('outtmpl', {}) | |
984 | if not isinstance(outtmpl_dict, dict): | |
985 | outtmpl_dict = {'default': outtmpl_dict} | |
986 | # Remove spaces in the default template | |
987 | if self.params.get('restrictfilenames'): | |
988 | sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-') | |
989 | else: | |
990 | sanitize = lambda x: x | |
991 | outtmpl_dict.update({ | |
992 | k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() | |
993 | if outtmpl_dict.get(k) is None}) | |
994 | for key, val in outtmpl_dict.items(): | |
995 | if isinstance(val, bytes): | |
996 | self.report_warning( | |
997 | 'Parameter outtmpl is bytes, but should be a unicode string. ' | |
998 | 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.') | |
999 | return outtmpl_dict | |
1000 | ||
1001 | def get_output_path(self, dir_type='', filename=None): | |
1002 | paths = self.params.get('paths', {}) | |
1003 | assert isinstance(paths, dict) | |
1004 | path = os.path.join( | |
1005 | expand_path(paths.get('home', '').strip()), | |
1006 | expand_path(paths.get(dir_type, '').strip()) if dir_type else '', | |
1007 | filename or '') | |
1008 | ||
1009 | # Temporary fix for #4787 | |
1010 | # 'Treat' all problem characters by passing filename through preferredencoding | |
1011 | # to workaround encoding issues with subprocess on python2 @ Windows | |
1012 | if sys.version_info < (3, 0) and sys.platform == 'win32': | |
1013 | path = encodeFilename(path, True).decode(preferredencoding()) | |
1014 | return sanitize_path(path, force=self.params.get('windowsfilenames')) | |
1015 | ||
1016 | @staticmethod | |
1017 | def _outtmpl_expandpath(outtmpl): | |
1018 | # expand_path translates '%%' into '%' and '$$' into '$' | |
1019 | # correspondingly that is not what we want since we need to keep | |
1020 | # '%%' intact for template dict substitution step. Working around | |
1021 | # with boundary-alike separator hack. | |
1022 | sep = ''.join([random.choice(ascii_letters) for _ in range(32)]) | |
1023 | outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep)) | |
1024 | ||
1025 | # outtmpl should be expand_path'ed before template dict substitution | |
1026 | # because meta fields may contain env variables we don't want to | |
1027 | # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and | |
1028 | # title "Hello $PATH", we don't want `$PATH` to be expanded. | |
1029 | return expand_path(outtmpl).replace(sep, '') | |
1030 | ||
1031 | @staticmethod | |
1032 | def escape_outtmpl(outtmpl): | |
1033 | ''' Escape any remaining strings like %s, %abc% etc. ''' | |
1034 | return re.sub( | |
1035 | STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'), | |
1036 | lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0), | |
1037 | outtmpl) | |
1038 | ||
1039 | @classmethod | |
1040 | def validate_outtmpl(cls, outtmpl): | |
1041 | ''' @return None or Exception object ''' | |
1042 | outtmpl = re.sub( | |
1043 | STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'), | |
1044 | lambda mobj: f'{mobj.group(0)[:-1]}s', | |
1045 | cls._outtmpl_expandpath(outtmpl)) | |
1046 | try: | |
1047 | cls.escape_outtmpl(outtmpl) % collections.defaultdict(int) | |
1048 | return None | |
1049 | except ValueError as err: | |
1050 | return err | |
1051 | ||
1052 | @staticmethod | |
1053 | def _copy_infodict(info_dict): | |
1054 | info_dict = dict(info_dict) | |
1055 | info_dict.pop('__postprocessors', None) | |
1056 | return info_dict | |
1057 | ||
1058 | def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False): | |
1059 | """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict | |
1060 | @param sanitize Whether to sanitize the output as a filename. | |
1061 | For backward compatibility, a function can also be passed | |
1062 | """ | |
1063 | ||
1064 | info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set | |
1065 | ||
1066 | info_dict = self._copy_infodict(info_dict) | |
1067 | info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs | |
1068 | formatSeconds(info_dict['duration'], '-' if sanitize else ':') | |
1069 | if info_dict.get('duration', None) is not None | |
1070 | else None) | |
1071 | info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads | |
1072 | info_dict['video_autonumber'] = self._num_videos | |
1073 | if info_dict.get('resolution') is None: | |
1074 | info_dict['resolution'] = self.format_resolution(info_dict, default=None) | |
1075 | ||
1076 | # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences | |
1077 | # of %(field)s to %(field)0Nd for backward compatibility | |
1078 | field_size_compat_map = { | |
1079 | 'playlist_index': number_of_digits(info_dict.get('_last_playlist_index') or 0), | |
1080 | 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0), | |
1081 | 'autonumber': self.params.get('autonumber_size') or 5, | |
1082 | } | |
1083 | ||
1084 | TMPL_DICT = {} | |
1085 | EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]')) | |
1086 | MATH_FUNCTIONS = { | |
1087 | '+': float.__add__, | |
1088 | '-': float.__sub__, | |
1089 | } | |
1090 | # Field is of the form key1.key2... | |
1091 | # where keys (except first) can be string, int or slice | |
1092 | FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)') | |
1093 | MATH_FIELD_RE = r'''(?:{field}|{num})'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?') | |
1094 | MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys())) | |
1095 | INTERNAL_FORMAT_RE = re.compile(r'''(?x) | |
1096 | (?P<negate>-)? | |
1097 | (?P<fields>{field}) | |
1098 | (?P<maths>(?:{math_op}{math_field})*) | |
1099 | (?:>(?P<strf_format>.+?))? | |
1100 | (?P<alternate>(?<!\\),[^|&)]+)? | |
1101 | (?:&(?P<replacement>.*?))? | |
1102 | (?:\|(?P<default>.*?))? | |
1103 | $'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE)) | |
1104 | ||
1105 | def _traverse_infodict(k): | |
1106 | k = k.split('.') | |
1107 | if k[0] == '': | |
1108 | k.pop(0) | |
1109 | return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True) | |
1110 | ||
1111 | def get_value(mdict): | |
1112 | # Object traversal | |
1113 | value = _traverse_infodict(mdict['fields']) | |
1114 | # Negative | |
1115 | if mdict['negate']: | |
1116 | value = float_or_none(value) | |
1117 | if value is not None: | |
1118 | value *= -1 | |
1119 | # Do maths | |
1120 | offset_key = mdict['maths'] | |
1121 | if offset_key: | |
1122 | value = float_or_none(value) | |
1123 | operator = None | |
1124 | while offset_key: | |
1125 | item = re.match( | |
1126 | MATH_FIELD_RE if operator else MATH_OPERATORS_RE, | |
1127 | offset_key).group(0) | |
1128 | offset_key = offset_key[len(item):] | |
1129 | if operator is None: | |
1130 | operator = MATH_FUNCTIONS[item] | |
1131 | continue | |
1132 | item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1) | |
1133 | offset = float_or_none(item) | |
1134 | if offset is None: | |
1135 | offset = float_or_none(_traverse_infodict(item)) | |
1136 | try: | |
1137 | value = operator(value, multiplier * offset) | |
1138 | except (TypeError, ZeroDivisionError): | |
1139 | return None | |
1140 | operator = None | |
1141 | # Datetime formatting | |
1142 | if mdict['strf_format']: | |
1143 | value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ',')) | |
1144 | ||
1145 | return value | |
1146 | ||
1147 | na = self.params.get('outtmpl_na_placeholder', 'NA') | |
1148 | ||
1149 | def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')): | |
1150 | return sanitize_filename(str(value), restricted=restricted, | |
1151 | is_id=re.search(r'(^|[_.])id(\.|$)', key)) | |
1152 | ||
1153 | sanitizer = sanitize if callable(sanitize) else filename_sanitizer | |
1154 | sanitize = bool(sanitize) | |
1155 | ||
1156 | def _dumpjson_default(obj): | |
1157 | if isinstance(obj, (set, LazyList)): | |
1158 | return list(obj) | |
1159 | return repr(obj) | |
1160 | ||
1161 | def create_key(outer_mobj): | |
1162 | if not outer_mobj.group('has_key'): | |
1163 | return outer_mobj.group(0) | |
1164 | key = outer_mobj.group('key') | |
1165 | mobj = re.match(INTERNAL_FORMAT_RE, key) | |
1166 | initial_field = mobj.group('fields') if mobj else '' | |
1167 | value, replacement, default = None, None, na | |
1168 | while mobj: | |
1169 | mobj = mobj.groupdict() | |
1170 | default = mobj['default'] if mobj['default'] is not None else default | |
1171 | value = get_value(mobj) | |
1172 | replacement = mobj['replacement'] | |
1173 | if value is None and mobj['alternate']: | |
1174 | mobj = re.match(INTERNAL_FORMAT_RE, mobj['alternate'][1:]) | |
1175 | else: | |
1176 | break | |
1177 | ||
1178 | fmt = outer_mobj.group('format') | |
1179 | if fmt == 's' and value is not None and key in field_size_compat_map.keys(): | |
1180 | fmt = '0{:d}d'.format(field_size_compat_map[key]) | |
1181 | ||
1182 | value = default if value is None else value if replacement is None else replacement | |
1183 | ||
1184 | flags = outer_mobj.group('conversion') or '' | |
1185 | str_fmt = f'{fmt[:-1]}s' | |
1186 | if fmt[-1] == 'l': # list | |
1187 | delim = '\n' if '#' in flags else ', ' | |
1188 | value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt | |
1189 | elif fmt[-1] == 'j': # json | |
1190 | value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt | |
1191 | elif fmt[-1] == 'q': # quoted | |
1192 | value = map(str, variadic(value) if '#' in flags else [value]) | |
1193 | value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt | |
1194 | elif fmt[-1] == 'B': # bytes | |
1195 | value = f'%{str_fmt}'.encode('utf-8') % str(value).encode('utf-8') | |
1196 | value, fmt = value.decode('utf-8', 'ignore'), 's' | |
1197 | elif fmt[-1] == 'U': # unicode normalized | |
1198 | value, fmt = unicodedata.normalize( | |
1199 | # "+" = compatibility equivalence, "#" = NFD | |
1200 | 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'), | |
1201 | value), str_fmt | |
1202 | elif fmt[-1] == 'D': # decimal suffix | |
1203 | num_fmt, fmt = fmt[:-1].replace('#', ''), 's' | |
1204 | value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s', | |
1205 | factor=1024 if '#' in flags else 1000) | |
1206 | elif fmt[-1] == 'S': # filename sanitization | |
1207 | value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt | |
1208 | elif fmt[-1] == 'c': | |
1209 | if value: | |
1210 | value = str(value)[0] | |
1211 | else: | |
1212 | fmt = str_fmt | |
1213 | elif fmt[-1] not in 'rs': # numeric | |
1214 | value = float_or_none(value) | |
1215 | if value is None: | |
1216 | value, fmt = default, 's' | |
1217 | ||
1218 | if sanitize: | |
1219 | if fmt[-1] == 'r': | |
1220 | # If value is an object, sanitize might convert it to a string | |
1221 | # So we convert it to repr first | |
1222 | value, fmt = repr(value), str_fmt | |
1223 | if fmt[-1] in 'csr': | |
1224 | value = sanitizer(initial_field, value) | |
1225 | ||
1226 | key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format')) | |
1227 | TMPL_DICT[key] = value | |
1228 | return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix')) | |
1229 | ||
1230 | return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT | |
1231 | ||
1232 | def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs): | |
1233 | outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs) | |
1234 | return self.escape_outtmpl(outtmpl) % info_dict | |
1235 | ||
1236 | def _prepare_filename(self, info_dict, tmpl_type='default'): | |
1237 | try: | |
1238 | outtmpl = self._outtmpl_expandpath(self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])) | |
1239 | filename = self.evaluate_outtmpl(outtmpl, info_dict, True) | |
1240 | if not filename: | |
1241 | return None | |
1242 | ||
1243 | if tmpl_type in ('default', 'temp'): | |
1244 | final_ext, ext = self.params.get('final_ext'), info_dict.get('ext') | |
1245 | if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'): | |
1246 | filename = replace_extension(filename, ext, final_ext) | |
1247 | else: | |
1248 | force_ext = OUTTMPL_TYPES[tmpl_type] | |
1249 | if force_ext: | |
1250 | filename = replace_extension(filename, force_ext, info_dict.get('ext')) | |
1251 | ||
1252 | # https://github.com/blackjack4494/youtube-dlc/issues/85 | |
1253 | trim_file_name = self.params.get('trim_file_name', False) | |
1254 | if trim_file_name: | |
1255 | no_ext, *ext = filename.rsplit('.', 2) | |
1256 | filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.') | |
1257 | ||
1258 | return filename | |
1259 | except ValueError as err: | |
1260 | self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')') | |
1261 | return None | |
1262 | ||
1263 | def prepare_filename(self, info_dict, dir_type='', warn=False): | |
1264 | """Generate the output filename.""" | |
1265 | ||
1266 | filename = self._prepare_filename(info_dict, dir_type or 'default') | |
1267 | if not filename and dir_type not in ('', 'temp'): | |
1268 | return '' | |
1269 | ||
1270 | if warn: | |
1271 | if not self.params.get('paths'): | |
1272 | pass | |
1273 | elif filename == '-': | |
1274 | self.report_warning('--paths is ignored when an outputting to stdout', only_once=True) | |
1275 | elif os.path.isabs(filename): | |
1276 | self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True) | |
1277 | if filename == '-' or not filename: | |
1278 | return filename | |
1279 | ||
1280 | return self.get_output_path(dir_type, filename) | |
1281 | ||
1282 | def _match_entry(self, info_dict, incomplete=False, silent=False): | |
1283 | """ Returns None if the file should be downloaded """ | |
1284 | ||
1285 | video_title = info_dict.get('title', info_dict.get('id', 'video')) | |
1286 | ||
1287 | def check_filter(): | |
1288 | if 'title' in info_dict: | |
1289 | # This can happen when we're just evaluating the playlist | |
1290 | title = info_dict['title'] | |
1291 | matchtitle = self.params.get('matchtitle', False) | |
1292 | if matchtitle: | |
1293 | if not re.search(matchtitle, title, re.IGNORECASE): | |
1294 | return '"' + title + '" title did not match pattern "' + matchtitle + '"' | |
1295 | rejecttitle = self.params.get('rejecttitle', False) | |
1296 | if rejecttitle: | |
1297 | if re.search(rejecttitle, title, re.IGNORECASE): | |
1298 | return '"' + title + '" title matched reject pattern "' + rejecttitle + '"' | |
1299 | date = info_dict.get('upload_date') | |
1300 | if date is not None: | |
1301 | dateRange = self.params.get('daterange', DateRange()) | |
1302 | if date not in dateRange: | |
1303 | return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange) | |
1304 | view_count = info_dict.get('view_count') | |
1305 | if view_count is not None: | |
1306 | min_views = self.params.get('min_views') | |
1307 | if min_views is not None and view_count < min_views: | |
1308 | return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views) | |
1309 | max_views = self.params.get('max_views') | |
1310 | if max_views is not None and view_count > max_views: | |
1311 | return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views) | |
1312 | if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')): | |
1313 | return 'Skipping "%s" because it is age restricted' % video_title | |
1314 | ||
1315 | match_filter = self.params.get('match_filter') | |
1316 | if match_filter is not None: | |
1317 | try: | |
1318 | ret = match_filter(info_dict, incomplete=incomplete) | |
1319 | except TypeError: | |
1320 | # For backward compatibility | |
1321 | ret = None if incomplete else match_filter(info_dict) | |
1322 | if ret is not None: | |
1323 | return ret | |
1324 | return None | |
1325 | ||
1326 | if self.in_download_archive(info_dict): | |
1327 | reason = '%s has already been recorded in the archive' % video_title | |
1328 | break_opt, break_err = 'break_on_existing', ExistingVideoReached | |
1329 | else: | |
1330 | reason = check_filter() | |
1331 | break_opt, break_err = 'break_on_reject', RejectedVideoReached | |
1332 | if reason is not None: | |
1333 | if not silent: | |
1334 | self.to_screen('[download] ' + reason) | |
1335 | if self.params.get(break_opt, False): | |
1336 | raise break_err() | |
1337 | return reason | |
1338 | ||
1339 | @staticmethod | |
1340 | def add_extra_info(info_dict, extra_info): | |
1341 | '''Set the keys from extra_info in info dict if they are missing''' | |
1342 | for key, value in extra_info.items(): | |
1343 | info_dict.setdefault(key, value) | |
1344 | ||
1345 | def extract_info(self, url, download=True, ie_key=None, extra_info=None, | |
1346 | process=True, force_generic_extractor=False): | |
1347 | """ | |
1348 | Return a list with a dictionary for each video extracted. | |
1349 | ||
1350 | Arguments: | |
1351 | url -- URL to extract | |
1352 | ||
1353 | Keyword arguments: | |
1354 | download -- whether to download videos during extraction | |
1355 | ie_key -- extractor key hint | |
1356 | extra_info -- dictionary containing the extra values to add to each result | |
1357 | process -- whether to resolve all unresolved references (URLs, playlist items), | |
1358 | must be True for download to work. | |
1359 | force_generic_extractor -- force using the generic extractor | |
1360 | """ | |
1361 | ||
1362 | if extra_info is None: | |
1363 | extra_info = {} | |
1364 | ||
1365 | if not ie_key and force_generic_extractor: | |
1366 | ie_key = 'Generic' | |
1367 | ||
1368 | if ie_key: | |
1369 | ies = {ie_key: self._get_info_extractor_class(ie_key)} | |
1370 | else: | |
1371 | ies = self._ies | |
1372 | ||
1373 | for ie_key, ie in ies.items(): | |
1374 | if not ie.suitable(url): | |
1375 | continue | |
1376 | ||
1377 | if not ie.working(): | |
1378 | self.report_warning('The program functionality for this site has been marked as broken, ' | |
1379 | 'and will probably not work.') | |
1380 | ||
1381 | temp_id = ie.get_temp_id(url) | |
1382 | if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}): | |
1383 | self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive') | |
1384 | if self.params.get('break_on_existing', False): | |
1385 | raise ExistingVideoReached() | |
1386 | break | |
1387 | return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process) | |
1388 | else: | |
1389 | self.report_error('no suitable InfoExtractor for URL %s' % url) | |
1390 | ||
1391 | def __handle_extraction_exceptions(func): | |
1392 | @functools.wraps(func) | |
1393 | def wrapper(self, *args, **kwargs): | |
1394 | while True: | |
1395 | try: | |
1396 | return func(self, *args, **kwargs) | |
1397 | except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError): | |
1398 | raise | |
1399 | except ReExtractInfo as e: | |
1400 | if e.expected: | |
1401 | self.to_screen(f'{e}; Re-extracting data') | |
1402 | else: | |
1403 | self.to_stderr('\r') | |
1404 | self.report_warning(f'{e}; Re-extracting data') | |
1405 | continue | |
1406 | except GeoRestrictedError as e: | |
1407 | msg = e.msg | |
1408 | if e.countries: | |
1409 | msg += '\nThis video is available in %s.' % ', '.join( | |
1410 | map(ISO3166Utils.short2full, e.countries)) | |
1411 | msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.' | |
1412 | self.report_error(msg) | |
1413 | except ExtractorError as e: # An error we somewhat expected | |
1414 | self.report_error(str(e), e.format_traceback()) | |
1415 | except Exception as e: | |
1416 | if self.params.get('ignoreerrors'): | |
1417 | self.report_error(str(e), tb=encode_compat_str(traceback.format_exc())) | |
1418 | else: | |
1419 | raise | |
1420 | break | |
1421 | return wrapper | |
1422 | ||
1423 | def _wait_for_video(self, ie_result): | |
1424 | if (not self.params.get('wait_for_video') | |
1425 | or ie_result.get('_type', 'video') != 'video' | |
1426 | or ie_result.get('formats') or ie_result.get('url')): | |
1427 | return | |
1428 | ||
1429 | format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1] | |
1430 | last_msg = '' | |
1431 | ||
1432 | def progress(msg): | |
1433 | nonlocal last_msg | |
1434 | self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True) | |
1435 | last_msg = msg | |
1436 | ||
1437 | min_wait, max_wait = self.params.get('wait_for_video') | |
1438 | diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time()) | |
1439 | if diff is None and ie_result.get('live_status') == 'is_upcoming': | |
1440 | diff = random.randrange(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait) | |
1441 | self.report_warning('Release time of video is not known') | |
1442 | elif (diff or 0) <= 0: | |
1443 | self.report_warning('Video should already be available according to extracted info') | |
1444 | diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf')) | |
1445 | self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now') | |
1446 | ||
1447 | wait_till = time.time() + diff | |
1448 | try: | |
1449 | while True: | |
1450 | diff = wait_till - time.time() | |
1451 | if diff <= 0: | |
1452 | progress('') | |
1453 | raise ReExtractInfo('[wait] Wait period ended', expected=True) | |
1454 | progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}') | |
1455 | time.sleep(1) | |
1456 | except KeyboardInterrupt: | |
1457 | progress('') | |
1458 | raise ReExtractInfo('[wait] Interrupted by user', expected=True) | |
1459 | except BaseException as e: | |
1460 | if not isinstance(e, ReExtractInfo): | |
1461 | self.to_screen('') | |
1462 | raise | |
1463 | ||
1464 | @__handle_extraction_exceptions | |
1465 | def __extract_info(self, url, ie, download, extra_info, process): | |
1466 | ie_result = ie.extract(url) | |
1467 | if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here) | |
1468 | return | |
1469 | if isinstance(ie_result, list): | |
1470 | # Backwards compatibility: old IE result format | |
1471 | ie_result = { | |
1472 | '_type': 'compat_list', | |
1473 | 'entries': ie_result, | |
1474 | } | |
1475 | if extra_info.get('original_url'): | |
1476 | ie_result.setdefault('original_url', extra_info['original_url']) | |
1477 | self.add_default_extra_info(ie_result, ie, url) | |
1478 | if process: | |
1479 | self._wait_for_video(ie_result) | |
1480 | return self.process_ie_result(ie_result, download, extra_info) | |
1481 | else: | |
1482 | return ie_result | |
1483 | ||
1484 | def add_default_extra_info(self, ie_result, ie, url): | |
1485 | if url is not None: | |
1486 | self.add_extra_info(ie_result, { | |
1487 | 'webpage_url': url, | |
1488 | 'original_url': url, | |
1489 | }) | |
1490 | webpage_url = ie_result.get('webpage_url') | |
1491 | if webpage_url: | |
1492 | self.add_extra_info(ie_result, { | |
1493 | 'webpage_url_basename': url_basename(webpage_url), | |
1494 | 'webpage_url_domain': get_domain(webpage_url), | |
1495 | }) | |
1496 | if ie is not None: | |
1497 | self.add_extra_info(ie_result, { | |
1498 | 'extractor': ie.IE_NAME, | |
1499 | 'extractor_key': ie.ie_key(), | |
1500 | }) | |
1501 | ||
1502 | def process_ie_result(self, ie_result, download=True, extra_info=None): | |
1503 | """ | |
1504 | Take the result of the ie(may be modified) and resolve all unresolved | |
1505 | references (URLs, playlist items). | |
1506 | ||
1507 | It will also download the videos if 'download'. | |
1508 | Returns the resolved ie_result. | |
1509 | """ | |
1510 | if extra_info is None: | |
1511 | extra_info = {} | |
1512 | result_type = ie_result.get('_type', 'video') | |
1513 | ||
1514 | if result_type in ('url', 'url_transparent'): | |
1515 | ie_result['url'] = sanitize_url(ie_result['url']) | |
1516 | if ie_result.get('original_url'): | |
1517 | extra_info.setdefault('original_url', ie_result['original_url']) | |
1518 | ||
1519 | extract_flat = self.params.get('extract_flat', False) | |
1520 | if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) | |
1521 | or extract_flat is True): | |
1522 | info_copy = ie_result.copy() | |
1523 | ie = try_get(ie_result.get('ie_key'), self.get_info_extractor) | |
1524 | if ie and not ie_result.get('id'): | |
1525 | info_copy['id'] = ie.get_temp_id(ie_result['url']) | |
1526 | self.add_default_extra_info(info_copy, ie, ie_result['url']) | |
1527 | self.add_extra_info(info_copy, extra_info) | |
1528 | info_copy, _ = self.pre_process(info_copy) | |
1529 | self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True) | |
1530 | if self.params.get('force_write_download_archive', False): | |
1531 | self.record_download_archive(info_copy) | |
1532 | return ie_result | |
1533 | ||
1534 | if result_type == 'video': | |
1535 | self.add_extra_info(ie_result, extra_info) | |
1536 | ie_result = self.process_video_result(ie_result, download=download) | |
1537 | additional_urls = (ie_result or {}).get('additional_urls') | |
1538 | if additional_urls: | |
1539 | # TODO: Improve MetadataParserPP to allow setting a list | |
1540 | if isinstance(additional_urls, compat_str): | |
1541 | additional_urls = [additional_urls] | |
1542 | self.to_screen( | |
1543 | '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls))) | |
1544 | self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls)) | |
1545 | ie_result['additional_entries'] = [ | |
1546 | self.extract_info( | |
1547 | url, download, extra_info=extra_info, | |
1548 | force_generic_extractor=self.params.get('force_generic_extractor')) | |
1549 | for url in additional_urls | |
1550 | ] | |
1551 | return ie_result | |
1552 | elif result_type == 'url': | |
1553 | # We have to add extra_info to the results because it may be | |
1554 | # contained in a playlist | |
1555 | return self.extract_info( | |
1556 | ie_result['url'], download, | |
1557 | ie_key=ie_result.get('ie_key'), | |
1558 | extra_info=extra_info) | |
1559 | elif result_type == 'url_transparent': | |
1560 | # Use the information from the embedding page | |
1561 | info = self.extract_info( | |
1562 | ie_result['url'], ie_key=ie_result.get('ie_key'), | |
1563 | extra_info=extra_info, download=False, process=False) | |
1564 | ||
1565 | # extract_info may return None when ignoreerrors is enabled and | |
1566 | # extraction failed with an error, don't crash and return early | |
1567 | # in this case | |
1568 | if not info: | |
1569 | return info | |
1570 | ||
1571 | force_properties = dict( | |
1572 | (k, v) for k, v in ie_result.items() if v is not None) | |
1573 | for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'): | |
1574 | if f in force_properties: | |
1575 | del force_properties[f] | |
1576 | new_result = info.copy() | |
1577 | new_result.update(force_properties) | |
1578 | ||
1579 | # Extracted info may not be a video result (i.e. | |
1580 | # info.get('_type', 'video') != video) but rather an url or | |
1581 | # url_transparent. In such cases outer metadata (from ie_result) | |
1582 | # should be propagated to inner one (info). For this to happen | |
1583 | # _type of info should be overridden with url_transparent. This | |
1584 | # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163. | |
1585 | if new_result.get('_type') == 'url': | |
1586 | new_result['_type'] = 'url_transparent' | |
1587 | ||
1588 | return self.process_ie_result( | |
1589 | new_result, download=download, extra_info=extra_info) | |
1590 | elif result_type in ('playlist', 'multi_video'): | |
1591 | # Protect from infinite recursion due to recursively nested playlists | |
1592 | # (see https://github.com/ytdl-org/youtube-dl/issues/27833) | |
1593 | webpage_url = ie_result['webpage_url'] | |
1594 | if webpage_url in self._playlist_urls: | |
1595 | self.to_screen( | |
1596 | '[download] Skipping already downloaded playlist: %s' | |
1597 | % ie_result.get('title') or ie_result.get('id')) | |
1598 | return | |
1599 | ||
1600 | self._playlist_level += 1 | |
1601 | self._playlist_urls.add(webpage_url) | |
1602 | self._fill_common_fields(ie_result, False) | |
1603 | self._sanitize_thumbnails(ie_result) | |
1604 | try: | |
1605 | return self.__process_playlist(ie_result, download) | |
1606 | finally: | |
1607 | self._playlist_level -= 1 | |
1608 | if not self._playlist_level: | |
1609 | self._playlist_urls.clear() | |
1610 | elif result_type == 'compat_list': | |
1611 | self.report_warning( | |
1612 | 'Extractor %s returned a compat_list result. ' | |
1613 | 'It needs to be updated.' % ie_result.get('extractor')) | |
1614 | ||
1615 | def _fixup(r): | |
1616 | self.add_extra_info(r, { | |
1617 | 'extractor': ie_result['extractor'], | |
1618 | 'webpage_url': ie_result['webpage_url'], | |
1619 | 'webpage_url_basename': url_basename(ie_result['webpage_url']), | |
1620 | 'webpage_url_domain': get_domain(ie_result['webpage_url']), | |
1621 | 'extractor_key': ie_result['extractor_key'], | |
1622 | }) | |
1623 | return r | |
1624 | ie_result['entries'] = [ | |
1625 | self.process_ie_result(_fixup(r), download, extra_info) | |
1626 | for r in ie_result['entries'] | |
1627 | ] | |
1628 | return ie_result | |
1629 | else: | |
1630 | raise Exception('Invalid result type: %s' % result_type) | |
1631 | ||
1632 | def _ensure_dir_exists(self, path): | |
1633 | return make_dir(path, self.report_error) | |
1634 | ||
1635 | @staticmethod | |
1636 | def _playlist_infodict(ie_result, **kwargs): | |
1637 | return { | |
1638 | **ie_result, | |
1639 | 'playlist': ie_result.get('title') or ie_result.get('id'), | |
1640 | 'playlist_id': ie_result.get('id'), | |
1641 | 'playlist_title': ie_result.get('title'), | |
1642 | 'playlist_uploader': ie_result.get('uploader'), | |
1643 | 'playlist_uploader_id': ie_result.get('uploader_id'), | |
1644 | 'playlist_index': 0, | |
1645 | **kwargs, | |
1646 | } | |
1647 | ||
1648 | def __process_playlist(self, ie_result, download): | |
1649 | # We process each entry in the playlist | |
1650 | playlist = ie_result.get('title') or ie_result.get('id') | |
1651 | self.to_screen('[download] Downloading playlist: %s' % playlist) | |
1652 | ||
1653 | if 'entries' not in ie_result: | |
1654 | raise EntryNotInPlaylist('There are no entries') | |
1655 | ||
1656 | MissingEntry = object() | |
1657 | incomplete_entries = bool(ie_result.get('requested_entries')) | |
1658 | if incomplete_entries: | |
1659 | def fill_missing_entries(entries, indices): | |
1660 | ret = [MissingEntry] * max(indices) | |
1661 | for i, entry in zip(indices, entries): | |
1662 | ret[i - 1] = entry | |
1663 | return ret | |
1664 | ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries']) | |
1665 | ||
1666 | playlist_results = [] | |
1667 | ||
1668 | playliststart = self.params.get('playliststart', 1) | |
1669 | playlistend = self.params.get('playlistend') | |
1670 | # For backwards compatibility, interpret -1 as whole list | |
1671 | if playlistend == -1: | |
1672 | playlistend = None | |
1673 | ||
1674 | playlistitems_str = self.params.get('playlist_items') | |
1675 | playlistitems = None | |
1676 | if playlistitems_str is not None: | |
1677 | def iter_playlistitems(format): | |
1678 | for string_segment in format.split(','): | |
1679 | if '-' in string_segment: | |
1680 | start, end = string_segment.split('-') | |
1681 | for item in range(int(start), int(end) + 1): | |
1682 | yield int(item) | |
1683 | else: | |
1684 | yield int(string_segment) | |
1685 | playlistitems = orderedSet(iter_playlistitems(playlistitems_str)) | |
1686 | ||
1687 | ie_entries = ie_result['entries'] | |
1688 | if isinstance(ie_entries, list): | |
1689 | playlist_count = len(ie_entries) | |
1690 | msg = f'Collected {playlist_count} videos; downloading %d of them' | |
1691 | ie_result['playlist_count'] = ie_result.get('playlist_count') or playlist_count | |
1692 | ||
1693 | def get_entry(i): | |
1694 | return ie_entries[i - 1] | |
1695 | else: | |
1696 | msg = 'Downloading %d videos' | |
1697 | if not isinstance(ie_entries, (PagedList, LazyList)): | |
1698 | ie_entries = LazyList(ie_entries) | |
1699 | elif isinstance(ie_entries, InAdvancePagedList): | |
1700 | if ie_entries._pagesize == 1: | |
1701 | playlist_count = ie_entries._pagecount | |
1702 | ||
1703 | def get_entry(i): | |
1704 | return YoutubeDL.__handle_extraction_exceptions( | |
1705 | lambda self, i: ie_entries[i - 1] | |
1706 | )(self, i) | |
1707 | ||
1708 | entries, broken = [], False | |
1709 | items = playlistitems if playlistitems is not None else itertools.count(playliststart) | |
1710 | for i in items: | |
1711 | if i == 0: | |
1712 | continue | |
1713 | if playlistitems is None and playlistend is not None and playlistend < i: | |
1714 | break | |
1715 | entry = None | |
1716 | try: | |
1717 | entry = get_entry(i) | |
1718 | if entry is MissingEntry: | |
1719 | raise EntryNotInPlaylist() | |
1720 | except (IndexError, EntryNotInPlaylist): | |
1721 | if incomplete_entries: | |
1722 | raise EntryNotInPlaylist(f'Entry {i} cannot be found') | |
1723 | elif not playlistitems: | |
1724 | break | |
1725 | entries.append(entry) | |
1726 | try: | |
1727 | if entry is not None: | |
1728 | self._match_entry(entry, incomplete=True, silent=True) | |
1729 | except (ExistingVideoReached, RejectedVideoReached): | |
1730 | broken = True | |
1731 | break | |
1732 | ie_result['entries'] = entries | |
1733 | ||
1734 | # Save playlist_index before re-ordering | |
1735 | entries = [ | |
1736 | ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry) | |
1737 | for i, entry in enumerate(entries, 1) | |
1738 | if entry is not None] | |
1739 | n_entries = len(entries) | |
1740 | ||
1741 | if not (ie_result.get('playlist_count') or broken or playlistitems or playlistend): | |
1742 | ie_result['playlist_count'] = n_entries | |
1743 | ||
1744 | if not playlistitems and (playliststart != 1 or playlistend): | |
1745 | playlistitems = list(range(playliststart, playliststart + n_entries)) | |
1746 | ie_result['requested_entries'] = playlistitems | |
1747 | ||
1748 | _infojson_written = False | |
1749 | write_playlist_files = self.params.get('allow_playlist_files', True) | |
1750 | if write_playlist_files and self.params.get('list_thumbnails'): | |
1751 | self.list_thumbnails(ie_result) | |
1752 | if write_playlist_files and not self.params.get('simulate'): | |
1753 | ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries) | |
1754 | _infojson_written = self._write_info_json( | |
1755 | 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson')) | |
1756 | if _infojson_written is None: | |
1757 | return | |
1758 | if self._write_description('playlist', ie_result, | |
1759 | self.prepare_filename(ie_copy, 'pl_description')) is None: | |
1760 | return | |
1761 | # TODO: This should be passed to ThumbnailsConvertor if necessary | |
1762 | self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail')) | |
1763 | ||
1764 | if self.params.get('playlistreverse', False): | |
1765 | entries = entries[::-1] | |
1766 | if self.params.get('playlistrandom', False): | |
1767 | random.shuffle(entries) | |
1768 | ||
1769 | x_forwarded_for = ie_result.get('__x_forwarded_for_ip') | |
1770 | ||
1771 | self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries)) | |
1772 | failures = 0 | |
1773 | max_failures = self.params.get('skip_playlist_after_errors') or float('inf') | |
1774 | for i, entry_tuple in enumerate(entries, 1): | |
1775 | playlist_index, entry = entry_tuple | |
1776 | if 'playlist-index' in self.params.get('compat_opts', []): | |
1777 | playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1 | |
1778 | self.to_screen('[download] Downloading video %s of %s' % (i, n_entries)) | |
1779 | # This __x_forwarded_for_ip thing is a bit ugly but requires | |
1780 | # minimal changes | |
1781 | if x_forwarded_for: | |
1782 | entry['__x_forwarded_for_ip'] = x_forwarded_for | |
1783 | extra = { | |
1784 | 'n_entries': n_entries, | |
1785 | '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries), | |
1786 | 'playlist_count': ie_result.get('playlist_count'), | |
1787 | 'playlist_index': playlist_index, | |
1788 | 'playlist_autonumber': i, | |
1789 | 'playlist': playlist, | |
1790 | 'playlist_id': ie_result.get('id'), | |
1791 | 'playlist_title': ie_result.get('title'), | |
1792 | 'playlist_uploader': ie_result.get('uploader'), | |
1793 | 'playlist_uploader_id': ie_result.get('uploader_id'), | |
1794 | 'extractor': ie_result['extractor'], | |
1795 | 'webpage_url': ie_result['webpage_url'], | |
1796 | 'webpage_url_basename': url_basename(ie_result['webpage_url']), | |
1797 | 'webpage_url_domain': get_domain(ie_result['webpage_url']), | |
1798 | 'extractor_key': ie_result['extractor_key'], | |
1799 | } | |
1800 | ||
1801 | if self._match_entry(entry, incomplete=True) is not None: | |
1802 | continue | |
1803 | ||
1804 | entry_result = self.__process_iterable_entry(entry, download, extra) | |
1805 | if not entry_result: | |
1806 | failures += 1 | |
1807 | if failures >= max_failures: | |
1808 | self.report_error( | |
1809 | 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures)) | |
1810 | break | |
1811 | playlist_results.append(entry_result) | |
1812 | ie_result['entries'] = playlist_results | |
1813 | ||
1814 | # Write the updated info to json | |
1815 | if _infojson_written and self._write_info_json( | |
1816 | 'updated playlist', ie_result, | |
1817 | self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None: | |
1818 | return | |
1819 | ||
1820 | ie_result = self.run_all_pps('playlist', ie_result) | |
1821 | self.to_screen(f'[download] Finished downloading playlist: {playlist}') | |
1822 | return ie_result | |
1823 | ||
1824 | @__handle_extraction_exceptions | |
1825 | def __process_iterable_entry(self, entry, download, extra_info): | |
1826 | return self.process_ie_result( | |
1827 | entry, download=download, extra_info=extra_info) | |
1828 | ||
1829 | def _build_format_filter(self, filter_spec): | |
1830 | " Returns a function to filter the formats according to the filter_spec " | |
1831 | ||
1832 | OPERATORS = { | |
1833 | '<': operator.lt, | |
1834 | '<=': operator.le, | |
1835 | '>': operator.gt, | |
1836 | '>=': operator.ge, | |
1837 | '=': operator.eq, | |
1838 | '!=': operator.ne, | |
1839 | } | |
1840 | operator_rex = re.compile(r'''(?x)\s* | |
1841 | (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s* | |
1842 | (?P<op>%s)(?P<none_inclusive>\s*\?)?\s* | |
1843 | (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s* | |
1844 | ''' % '|'.join(map(re.escape, OPERATORS.keys()))) | |
1845 | m = operator_rex.fullmatch(filter_spec) | |
1846 | if m: | |
1847 | try: | |
1848 | comparison_value = int(m.group('value')) | |
1849 | except ValueError: | |
1850 | comparison_value = parse_filesize(m.group('value')) | |
1851 | if comparison_value is None: | |
1852 | comparison_value = parse_filesize(m.group('value') + 'B') | |
1853 | if comparison_value is None: | |
1854 | raise ValueError( | |
1855 | 'Invalid value %r in format specification %r' % ( | |
1856 | m.group('value'), filter_spec)) | |
1857 | op = OPERATORS[m.group('op')] | |
1858 | ||
1859 | if not m: | |
1860 | STR_OPERATORS = { | |
1861 | '=': operator.eq, | |
1862 | '^=': lambda attr, value: attr.startswith(value), | |
1863 | '$=': lambda attr, value: attr.endswith(value), | |
1864 | '*=': lambda attr, value: value in attr, | |
1865 | '~=': lambda attr, value: value.search(attr) is not None | |
1866 | } | |
1867 | str_operator_rex = re.compile(r'''(?x)\s* | |
1868 | (?P<key>[a-zA-Z0-9._-]+)\s* | |
1869 | (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)? | |
1870 | (?P<quote>["'])? | |
1871 | (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+)) | |
1872 | (?(quote)(?P=quote))\s* | |
1873 | ''' % '|'.join(map(re.escape, STR_OPERATORS.keys()))) | |
1874 | m = str_operator_rex.fullmatch(filter_spec) | |
1875 | if m: | |
1876 | if m.group('op') == '~=': | |
1877 | comparison_value = re.compile(m.group('value')) | |
1878 | else: | |
1879 | comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value')) | |
1880 | str_op = STR_OPERATORS[m.group('op')] | |
1881 | if m.group('negation'): | |
1882 | op = lambda attr, value: not str_op(attr, value) | |
1883 | else: | |
1884 | op = str_op | |
1885 | ||
1886 | if not m: | |
1887 | raise SyntaxError('Invalid filter specification %r' % filter_spec) | |
1888 | ||
1889 | def _filter(f): | |
1890 | actual_value = f.get(m.group('key')) | |
1891 | if actual_value is None: | |
1892 | return m.group('none_inclusive') | |
1893 | return op(actual_value, comparison_value) | |
1894 | return _filter | |
1895 | ||
1896 | def _check_formats(self, formats): | |
1897 | for f in formats: | |
1898 | self.to_screen('[info] Testing format %s' % f['format_id']) | |
1899 | path = self.get_output_path('temp') | |
1900 | if not self._ensure_dir_exists(f'{path}/'): | |
1901 | continue | |
1902 | temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None) | |
1903 | temp_file.close() | |
1904 | try: | |
1905 | success, _ = self.dl(temp_file.name, f, test=True) | |
1906 | except (DownloadError, IOError, OSError, ValueError) + network_exceptions: | |
1907 | success = False | |
1908 | finally: | |
1909 | if os.path.exists(temp_file.name): | |
1910 | try: | |
1911 | os.remove(temp_file.name) | |
1912 | except OSError: | |
1913 | self.report_warning('Unable to delete temporary file "%s"' % temp_file.name) | |
1914 | if success: | |
1915 | yield f | |
1916 | else: | |
1917 | self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id']) | |
1918 | ||
1919 | def _default_format_spec(self, info_dict, download=True): | |
1920 | ||
1921 | def can_merge(): | |
1922 | merger = FFmpegMergerPP(self) | |
1923 | return merger.available and merger.can_merge() | |
1924 | ||
1925 | prefer_best = ( | |
1926 | not self.params.get('simulate') | |
1927 | and download | |
1928 | and ( | |
1929 | not can_merge() | |
1930 | or info_dict.get('is_live', False) | |
1931 | or self.outtmpl_dict['default'] == '-')) | |
1932 | compat = ( | |
1933 | prefer_best | |
1934 | or self.params.get('allow_multiple_audio_streams', False) | |
1935 | or 'format-spec' in self.params.get('compat_opts', [])) | |
1936 | ||
1937 | return ( | |
1938 | 'best/bestvideo+bestaudio' if prefer_best | |
1939 | else 'bestvideo*+bestaudio/best' if not compat | |
1940 | else 'bestvideo+bestaudio/best') | |
1941 | ||
1942 | def build_format_selector(self, format_spec): | |
1943 | def syntax_error(note, start): | |
1944 | message = ( | |
1945 | 'Invalid format specification: ' | |
1946 | '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1])) | |
1947 | return SyntaxError(message) | |
1948 | ||
1949 | PICKFIRST = 'PICKFIRST' | |
1950 | MERGE = 'MERGE' | |
1951 | SINGLE = 'SINGLE' | |
1952 | GROUP = 'GROUP' | |
1953 | FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters']) | |
1954 | ||
1955 | allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False), | |
1956 | 'video': self.params.get('allow_multiple_video_streams', False)} | |
1957 | ||
1958 | check_formats = self.params.get('check_formats') == 'selected' | |
1959 | ||
1960 | def _parse_filter(tokens): | |
1961 | filter_parts = [] | |
1962 | for type, string, start, _, _ in tokens: | |
1963 | if type == tokenize.OP and string == ']': | |
1964 | return ''.join(filter_parts) | |
1965 | else: | |
1966 | filter_parts.append(string) | |
1967 | ||
1968 | def _remove_unused_ops(tokens): | |
1969 | # Remove operators that we don't use and join them with the surrounding strings | |
1970 | # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9' | |
1971 | ALLOWED_OPS = ('/', '+', ',', '(', ')') | |
1972 | last_string, last_start, last_end, last_line = None, None, None, None | |
1973 | for type, string, start, end, line in tokens: | |
1974 | if type == tokenize.OP and string == '[': | |
1975 | if last_string: | |
1976 | yield tokenize.NAME, last_string, last_start, last_end, last_line | |
1977 | last_string = None | |
1978 | yield type, string, start, end, line | |
1979 | # everything inside brackets will be handled by _parse_filter | |
1980 | for type, string, start, end, line in tokens: | |
1981 | yield type, string, start, end, line | |
1982 | if type == tokenize.OP and string == ']': | |
1983 | break | |
1984 | elif type == tokenize.OP and string in ALLOWED_OPS: | |
1985 | if last_string: | |
1986 | yield tokenize.NAME, last_string, last_start, last_end, last_line | |
1987 | last_string = None | |
1988 | yield type, string, start, end, line | |
1989 | elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]: | |
1990 | if not last_string: | |
1991 | last_string = string | |
1992 | last_start = start | |
1993 | last_end = end | |
1994 | else: | |
1995 | last_string += string | |
1996 | if last_string: | |
1997 | yield tokenize.NAME, last_string, last_start, last_end, last_line | |
1998 | ||
1999 | def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False): | |
2000 | selectors = [] | |
2001 | current_selector = None | |
2002 | for type, string, start, _, _ in tokens: | |
2003 | # ENCODING is only defined in python 3.x | |
2004 | if type == getattr(tokenize, 'ENCODING', None): | |
2005 | continue | |
2006 | elif type in [tokenize.NAME, tokenize.NUMBER]: | |
2007 | current_selector = FormatSelector(SINGLE, string, []) | |
2008 | elif type == tokenize.OP: | |
2009 | if string == ')': | |
2010 | if not inside_group: | |
2011 | # ')' will be handled by the parentheses group | |
2012 | tokens.restore_last_token() | |
2013 | break | |
2014 | elif inside_merge and string in ['/', ',']: | |
2015 | tokens.restore_last_token() | |
2016 | break | |
2017 | elif inside_choice and string == ',': | |
2018 | tokens.restore_last_token() | |
2019 | break | |
2020 | elif string == ',': | |
2021 | if not current_selector: | |
2022 | raise syntax_error('"," must follow a format selector', start) | |
2023 | selectors.append(current_selector) | |
2024 | current_selector = None | |
2025 | elif string == '/': | |
2026 | if not current_selector: | |
2027 | raise syntax_error('"/" must follow a format selector', start) | |
2028 | first_choice = current_selector | |
2029 | second_choice = _parse_format_selection(tokens, inside_choice=True) | |
2030 | current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), []) | |
2031 | elif string == '[': | |
2032 | if not current_selector: | |
2033 | current_selector = FormatSelector(SINGLE, 'best', []) | |
2034 | format_filter = _parse_filter(tokens) | |
2035 | current_selector.filters.append(format_filter) | |
2036 | elif string == '(': | |
2037 | if current_selector: | |
2038 | raise syntax_error('Unexpected "("', start) | |
2039 | group = _parse_format_selection(tokens, inside_group=True) | |
2040 | current_selector = FormatSelector(GROUP, group, []) | |
2041 | elif string == '+': | |
2042 | if not current_selector: | |
2043 | raise syntax_error('Unexpected "+"', start) | |
2044 | selector_1 = current_selector | |
2045 | selector_2 = _parse_format_selection(tokens, inside_merge=True) | |
2046 | if not selector_2: | |
2047 | raise syntax_error('Expected a selector', start) | |
2048 | current_selector = FormatSelector(MERGE, (selector_1, selector_2), []) | |
2049 | else: | |
2050 | raise syntax_error('Operator not recognized: "{0}"'.format(string), start) | |
2051 | elif type == tokenize.ENDMARKER: | |
2052 | break | |
2053 | if current_selector: | |
2054 | selectors.append(current_selector) | |
2055 | return selectors | |
2056 | ||
2057 | def _merge(formats_pair): | |
2058 | format_1, format_2 = formats_pair | |
2059 | ||
2060 | formats_info = [] | |
2061 | formats_info.extend(format_1.get('requested_formats', (format_1,))) | |
2062 | formats_info.extend(format_2.get('requested_formats', (format_2,))) | |
2063 | ||
2064 | if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']: | |
2065 | get_no_more = {'video': False, 'audio': False} | |
2066 | for (i, fmt_info) in enumerate(formats_info): | |
2067 | if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none': | |
2068 | formats_info.pop(i) | |
2069 | continue | |
2070 | for aud_vid in ['audio', 'video']: | |
2071 | if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none': | |
2072 | if get_no_more[aud_vid]: | |
2073 | formats_info.pop(i) | |
2074 | break | |
2075 | get_no_more[aud_vid] = True | |
2076 | ||
2077 | if len(formats_info) == 1: | |
2078 | return formats_info[0] | |
2079 | ||
2080 | video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none'] | |
2081 | audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none'] | |
2082 | ||
2083 | the_only_video = video_fmts[0] if len(video_fmts) == 1 else None | |
2084 | the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None | |
2085 | ||
2086 | output_ext = self.params.get('merge_output_format') | |
2087 | if not output_ext: | |
2088 | if the_only_video: | |
2089 | output_ext = the_only_video['ext'] | |
2090 | elif the_only_audio and not video_fmts: | |
2091 | output_ext = the_only_audio['ext'] | |
2092 | else: | |
2093 | output_ext = 'mkv' | |
2094 | ||
2095 | filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info)) | |
2096 | ||
2097 | new_dict = { | |
2098 | 'requested_formats': formats_info, | |
2099 | 'format': '+'.join(filtered('format')), | |
2100 | 'format_id': '+'.join(filtered('format_id')), | |
2101 | 'ext': output_ext, | |
2102 | 'protocol': '+'.join(map(determine_protocol, formats_info)), | |
2103 | 'language': '+'.join(orderedSet(filtered('language'))) or None, | |
2104 | 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None, | |
2105 | 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None, | |
2106 | 'tbr': sum(filtered('tbr', 'vbr', 'abr')), | |
2107 | } | |
2108 | ||
2109 | if the_only_video: | |
2110 | new_dict.update({ | |
2111 | 'width': the_only_video.get('width'), | |
2112 | 'height': the_only_video.get('height'), | |
2113 | 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video), | |
2114 | 'fps': the_only_video.get('fps'), | |
2115 | 'dynamic_range': the_only_video.get('dynamic_range'), | |
2116 | 'vcodec': the_only_video.get('vcodec'), | |
2117 | 'vbr': the_only_video.get('vbr'), | |
2118 | 'stretched_ratio': the_only_video.get('stretched_ratio'), | |
2119 | }) | |
2120 | ||
2121 | if the_only_audio: | |
2122 | new_dict.update({ | |
2123 | 'acodec': the_only_audio.get('acodec'), | |
2124 | 'abr': the_only_audio.get('abr'), | |
2125 | 'asr': the_only_audio.get('asr'), | |
2126 | }) | |
2127 | ||
2128 | return new_dict | |
2129 | ||
2130 | def _check_formats(formats): | |
2131 | if not check_formats: | |
2132 | yield from formats | |
2133 | return | |
2134 | yield from self._check_formats(formats) | |
2135 | ||
2136 | def _build_selector_function(selector): | |
2137 | if isinstance(selector, list): # , | |
2138 | fs = [_build_selector_function(s) for s in selector] | |
2139 | ||
2140 | def selector_function(ctx): | |
2141 | for f in fs: | |
2142 | yield from f(ctx) | |
2143 | return selector_function | |
2144 | ||
2145 | elif selector.type == GROUP: # () | |
2146 | selector_function = _build_selector_function(selector.selector) | |
2147 | ||
2148 | elif selector.type == PICKFIRST: # / | |
2149 | fs = [_build_selector_function(s) for s in selector.selector] | |
2150 | ||
2151 | def selector_function(ctx): | |
2152 | for f in fs: | |
2153 | picked_formats = list(f(ctx)) | |
2154 | if picked_formats: | |
2155 | return picked_formats | |
2156 | return [] | |
2157 | ||
2158 | elif selector.type == MERGE: # + | |
2159 | selector_1, selector_2 = map(_build_selector_function, selector.selector) | |
2160 | ||
2161 | def selector_function(ctx): | |
2162 | for pair in itertools.product(selector_1(ctx), selector_2(ctx)): | |
2163 | yield _merge(pair) | |
2164 | ||
2165 | elif selector.type == SINGLE: # atom | |
2166 | format_spec = selector.selector or 'best' | |
2167 | ||
2168 | # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector | |
2169 | if format_spec == 'all': | |
2170 | def selector_function(ctx): | |
2171 | yield from _check_formats(ctx['formats'][::-1]) | |
2172 | elif format_spec == 'mergeall': | |
2173 | def selector_function(ctx): | |
2174 | formats = list(_check_formats(ctx['formats'])) | |
2175 | if not formats: | |
2176 | return | |
2177 | merged_format = formats[-1] | |
2178 | for f in formats[-2::-1]: | |
2179 | merged_format = _merge((merged_format, f)) | |
2180 | yield merged_format | |
2181 | ||
2182 | else: | |
2183 | format_fallback, format_reverse, format_idx = False, True, 1 | |
2184 | mobj = re.match( | |
2185 | r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$', | |
2186 | format_spec) | |
2187 | if mobj is not None: | |
2188 | format_idx = int_or_none(mobj.group('n'), default=1) | |
2189 | format_reverse = mobj.group('bw')[0] == 'b' | |
2190 | format_type = (mobj.group('type') or [None])[0] | |
2191 | not_format_type = {'v': 'a', 'a': 'v'}.get(format_type) | |
2192 | format_modified = mobj.group('mod') is not None | |
2193 | ||
2194 | format_fallback = not format_type and not format_modified # for b, w | |
2195 | _filter_f = ( | |
2196 | (lambda f: f.get('%scodec' % format_type) != 'none') | |
2197 | if format_type and format_modified # bv*, ba*, wv*, wa* | |
2198 | else (lambda f: f.get('%scodec' % not_format_type) == 'none') | |
2199 | if format_type # bv, ba, wv, wa | |
2200 | else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none') | |
2201 | if not format_modified # b, w | |
2202 | else lambda f: True) # b*, w* | |
2203 | filter_f = lambda f: _filter_f(f) and ( | |
2204 | f.get('vcodec') != 'none' or f.get('acodec') != 'none') | |
2205 | else: | |
2206 | if format_spec in self._format_selection_exts['audio']: | |
2207 | filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' | |
2208 | elif format_spec in self._format_selection_exts['video']: | |
2209 | filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none' | |
2210 | elif format_spec in self._format_selection_exts['storyboards']: | |
2211 | filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none' | |
2212 | else: | |
2213 | filter_f = lambda f: f.get('format_id') == format_spec # id | |
2214 | ||
2215 | def selector_function(ctx): | |
2216 | formats = list(ctx['formats']) | |
2217 | matches = list(filter(filter_f, formats)) if filter_f is not None else formats | |
2218 | if format_fallback and ctx['incomplete_formats'] and not matches: | |
2219 | # for extractors with incomplete formats (audio only (soundcloud) | |
2220 | # or video only (imgur)) best/worst will fallback to | |
2221 | # best/worst {video,audio}-only format | |
2222 | matches = formats | |
2223 | matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1])) | |
2224 | try: | |
2225 | yield matches[format_idx - 1] | |
2226 | except IndexError: | |
2227 | return | |
2228 | ||
2229 | filters = [self._build_format_filter(f) for f in selector.filters] | |
2230 | ||
2231 | def final_selector(ctx): | |
2232 | ctx_copy = dict(ctx) | |
2233 | for _filter in filters: | |
2234 | ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats'])) | |
2235 | return selector_function(ctx_copy) | |
2236 | return final_selector | |
2237 | ||
2238 | stream = io.BytesIO(format_spec.encode('utf-8')) | |
2239 | try: | |
2240 | tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline))) | |
2241 | except tokenize.TokenError: | |
2242 | raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec))) | |
2243 | ||
2244 | class TokenIterator(object): | |
2245 | def __init__(self, tokens): | |
2246 | self.tokens = tokens | |
2247 | self.counter = 0 | |
2248 | ||
2249 | def __iter__(self): | |
2250 | return self | |
2251 | ||
2252 | def __next__(self): | |
2253 | if self.counter >= len(self.tokens): | |
2254 | raise StopIteration() | |
2255 | value = self.tokens[self.counter] | |
2256 | self.counter += 1 | |
2257 | return value | |
2258 | ||
2259 | next = __next__ | |
2260 | ||
2261 | def restore_last_token(self): | |
2262 | self.counter -= 1 | |
2263 | ||
2264 | parsed_selector = _parse_format_selection(iter(TokenIterator(tokens))) | |
2265 | return _build_selector_function(parsed_selector) | |
2266 | ||
2267 | def _calc_headers(self, info_dict): | |
2268 | res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {}) | |
2269 | ||
2270 | cookies = self._calc_cookies(info_dict) | |
2271 | if cookies: | |
2272 | res['Cookie'] = cookies | |
2273 | ||
2274 | if 'X-Forwarded-For' not in res: | |
2275 | x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip') | |
2276 | if x_forwarded_for_ip: | |
2277 | res['X-Forwarded-For'] = x_forwarded_for_ip | |
2278 | ||
2279 | return res | |
2280 | ||
2281 | def _calc_cookies(self, info_dict): | |
2282 | pr = sanitized_Request(info_dict['url']) | |
2283 | self.cookiejar.add_cookie_header(pr) | |
2284 | return pr.get_header('Cookie') | |
2285 | ||
2286 | def _sort_thumbnails(self, thumbnails): | |
2287 | thumbnails.sort(key=lambda t: ( | |
2288 | t.get('preference') if t.get('preference') is not None else -1, | |
2289 | t.get('width') if t.get('width') is not None else -1, | |
2290 | t.get('height') if t.get('height') is not None else -1, | |
2291 | t.get('id') if t.get('id') is not None else '', | |
2292 | t.get('url'))) | |
2293 | ||
2294 | def _sanitize_thumbnails(self, info_dict): | |
2295 | thumbnails = info_dict.get('thumbnails') | |
2296 | if thumbnails is None: | |
2297 | thumbnail = info_dict.get('thumbnail') | |
2298 | if thumbnail: | |
2299 | info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}] | |
2300 | if not thumbnails: | |
2301 | return | |
2302 | ||
2303 | def check_thumbnails(thumbnails): | |
2304 | for t in thumbnails: | |
2305 | self.to_screen(f'[info] Testing thumbnail {t["id"]}') | |
2306 | try: | |
2307 | self.urlopen(HEADRequest(t['url'])) | |
2308 | except network_exceptions as err: | |
2309 | self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...') | |
2310 | continue | |
2311 | yield t | |
2312 | ||
2313 | self._sort_thumbnails(thumbnails) | |
2314 | for i, t in enumerate(thumbnails): | |
2315 | if t.get('id') is None: | |
2316 | t['id'] = '%d' % i | |
2317 | if t.get('width') and t.get('height'): | |
2318 | t['resolution'] = '%dx%d' % (t['width'], t['height']) | |
2319 | t['url'] = sanitize_url(t['url']) | |
2320 | ||
2321 | if self.params.get('check_formats') is True: | |
2322 | info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True) | |
2323 | else: | |
2324 | info_dict['thumbnails'] = thumbnails | |
2325 | ||
2326 | def _fill_common_fields(self, info_dict, is_video=True): | |
2327 | # TODO: move sanitization here | |
2328 | if is_video: | |
2329 | # playlists are allowed to lack "title" | |
2330 | info_dict['fulltitle'] = info_dict.get('title') | |
2331 | if 'title' not in info_dict: | |
2332 | raise ExtractorError('Missing "title" field in extractor result', | |
2333 | video_id=info_dict['id'], ie=info_dict['extractor']) | |
2334 | elif not info_dict.get('title'): | |
2335 | self.report_warning('Extractor failed to obtain "title". Creating a generic title instead') | |
2336 | info_dict['title'] = f'{info_dict["extractor"]} video #{info_dict["id"]}' | |
2337 | ||
2338 | if info_dict.get('duration') is not None: | |
2339 | info_dict['duration_string'] = formatSeconds(info_dict['duration']) | |
2340 | ||
2341 | for ts_key, date_key in ( | |
2342 | ('timestamp', 'upload_date'), | |
2343 | ('release_timestamp', 'release_date'), | |
2344 | ('modified_timestamp', 'modified_date'), | |
2345 | ): | |
2346 | if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None: | |
2347 | # Working around out-of-range timestamp values (e.g. negative ones on Windows, | |
2348 | # see http://bugs.python.org/issue1646728) | |
2349 | try: | |
2350 | upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key]) | |
2351 | info_dict[date_key] = upload_date.strftime('%Y%m%d') | |
2352 | except (ValueError, OverflowError, OSError): | |
2353 | pass | |
2354 | ||
2355 | live_keys = ('is_live', 'was_live') | |
2356 | live_status = info_dict.get('live_status') | |
2357 | if live_status is None: | |
2358 | for key in live_keys: | |
2359 | if info_dict.get(key) is False: | |
2360 | continue | |
2361 | if info_dict.get(key): | |
2362 | live_status = key | |
2363 | break | |
2364 | if all(info_dict.get(key) is False for key in live_keys): | |
2365 | live_status = 'not_live' | |
2366 | if live_status: | |
2367 | info_dict['live_status'] = live_status | |
2368 | for key in live_keys: | |
2369 | if info_dict.get(key) is None: | |
2370 | info_dict[key] = (live_status == key) | |
2371 | ||
2372 | # Auto generate title fields corresponding to the *_number fields when missing | |
2373 | # in order to always have clean titles. This is very common for TV series. | |
2374 | for field in ('chapter', 'season', 'episode'): | |
2375 | if info_dict.get('%s_number' % field) is not None and not info_dict.get(field): | |
2376 | info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field]) | |
2377 | ||
2378 | def process_video_result(self, info_dict, download=True): | |
2379 | assert info_dict.get('_type', 'video') == 'video' | |
2380 | self._num_videos += 1 | |
2381 | ||
2382 | if 'id' not in info_dict: | |
2383 | raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor']) | |
2384 | elif not info_dict.get('id'): | |
2385 | raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor']) | |
2386 | ||
2387 | def report_force_conversion(field, field_not, conversion): | |
2388 | self.report_warning( | |
2389 | '"%s" field is not %s - forcing %s conversion, there is an error in extractor' | |
2390 | % (field, field_not, conversion)) | |
2391 | ||
2392 | def sanitize_string_field(info, string_field): | |
2393 | field = info.get(string_field) | |
2394 | if field is None or isinstance(field, compat_str): | |
2395 | return | |
2396 | report_force_conversion(string_field, 'a string', 'string') | |
2397 | info[string_field] = compat_str(field) | |
2398 | ||
2399 | def sanitize_numeric_fields(info): | |
2400 | for numeric_field in self._NUMERIC_FIELDS: | |
2401 | field = info.get(numeric_field) | |
2402 | if field is None or isinstance(field, compat_numeric_types): | |
2403 | continue | |
2404 | report_force_conversion(numeric_field, 'numeric', 'int') | |
2405 | info[numeric_field] = int_or_none(field) | |
2406 | ||
2407 | sanitize_string_field(info_dict, 'id') | |
2408 | sanitize_numeric_fields(info_dict) | |
2409 | if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None): | |
2410 | self.report_warning('"duration" field is negative, there is an error in extractor') | |
2411 | ||
2412 | if 'playlist' not in info_dict: | |
2413 | # It isn't part of a playlist | |
2414 | info_dict['playlist'] = None | |
2415 | info_dict['playlist_index'] = None | |
2416 | ||
2417 | self._sanitize_thumbnails(info_dict) | |
2418 | ||
2419 | thumbnail = info_dict.get('thumbnail') | |
2420 | thumbnails = info_dict.get('thumbnails') | |
2421 | if thumbnail: | |
2422 | info_dict['thumbnail'] = sanitize_url(thumbnail) | |
2423 | elif thumbnails: | |
2424 | info_dict['thumbnail'] = thumbnails[-1]['url'] | |
2425 | ||
2426 | if info_dict.get('display_id') is None and 'id' in info_dict: | |
2427 | info_dict['display_id'] = info_dict['id'] | |
2428 | ||
2429 | self._fill_common_fields(info_dict) | |
2430 | ||
2431 | for cc_kind in ('subtitles', 'automatic_captions'): | |
2432 | cc = info_dict.get(cc_kind) | |
2433 | if cc: | |
2434 | for _, subtitle in cc.items(): | |
2435 | for subtitle_format in subtitle: | |
2436 | if subtitle_format.get('url'): | |
2437 | subtitle_format['url'] = sanitize_url(subtitle_format['url']) | |
2438 | if subtitle_format.get('ext') is None: | |
2439 | subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower() | |
2440 | ||
2441 | automatic_captions = info_dict.get('automatic_captions') | |
2442 | subtitles = info_dict.get('subtitles') | |
2443 | ||
2444 | info_dict['requested_subtitles'] = self.process_subtitles( | |
2445 | info_dict['id'], subtitles, automatic_captions) | |
2446 | ||
2447 | if info_dict.get('formats') is None: | |
2448 | # There's only one format available | |
2449 | formats = [info_dict] | |
2450 | else: | |
2451 | formats = info_dict['formats'] | |
2452 | ||
2453 | info_dict['__has_drm'] = any(f.get('has_drm') for f in formats) | |
2454 | if not self.params.get('allow_unplayable_formats'): | |
2455 | formats = [f for f in formats if not f.get('has_drm')] | |
2456 | ||
2457 | get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start')) | |
2458 | if not get_from_start: | |
2459 | info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M') | |
2460 | if info_dict.get('is_live') and formats: | |
2461 | formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start] | |
2462 | if get_from_start and not formats: | |
2463 | self.raise_no_formats(info_dict, msg='--live-from-start is passed, but there are no formats that can be downloaded from the start. ' | |
2464 | 'If you want to download from the current time, pass --no-live-from-start') | |
2465 | ||
2466 | if not formats: | |
2467 | self.raise_no_formats(info_dict) | |
2468 | ||
2469 | def is_wellformed(f): | |
2470 | url = f.get('url') | |
2471 | if not url: | |
2472 | self.report_warning( | |
2473 | '"url" field is missing or empty - skipping format, ' | |
2474 | 'there is an error in extractor') | |
2475 | return False | |
2476 | if isinstance(url, bytes): | |
2477 | sanitize_string_field(f, 'url') | |
2478 | return True | |
2479 | ||
2480 | # Filter out malformed formats for better extraction robustness | |
2481 | formats = list(filter(is_wellformed, formats)) | |
2482 | ||
2483 | formats_dict = {} | |
2484 | ||
2485 | # We check that all the formats have the format and format_id fields | |
2486 | for i, format in enumerate(formats): | |
2487 | sanitize_string_field(format, 'format_id') | |
2488 | sanitize_numeric_fields(format) | |
2489 | format['url'] = sanitize_url(format['url']) | |
2490 | if not format.get('format_id'): | |
2491 | format['format_id'] = compat_str(i) | |
2492 | else: | |
2493 | # Sanitize format_id from characters used in format selector expression | |
2494 | format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id']) | |
2495 | format_id = format['format_id'] | |
2496 | if format_id not in formats_dict: | |
2497 | formats_dict[format_id] = [] | |
2498 | formats_dict[format_id].append(format) | |
2499 | ||
2500 | # Make sure all formats have unique format_id | |
2501 | common_exts = set(itertools.chain(*self._format_selection_exts.values())) | |
2502 | for format_id, ambiguous_formats in formats_dict.items(): | |
2503 | ambigious_id = len(ambiguous_formats) > 1 | |
2504 | for i, format in enumerate(ambiguous_formats): | |
2505 | if ambigious_id: | |
2506 | format['format_id'] = '%s-%d' % (format_id, i) | |
2507 | if format.get('ext') is None: | |
2508 | format['ext'] = determine_ext(format['url']).lower() | |
2509 | # Ensure there is no conflict between id and ext in format selection | |
2510 | # See https://github.com/yt-dlp/yt-dlp/issues/1282 | |
2511 | if format['format_id'] != format['ext'] and format['format_id'] in common_exts: | |
2512 | format['format_id'] = 'f%s' % format['format_id'] | |
2513 | ||
2514 | for i, format in enumerate(formats): | |
2515 | if format.get('format') is None: | |
2516 | format['format'] = '{id} - {res}{note}'.format( | |
2517 | id=format['format_id'], | |
2518 | res=self.format_resolution(format), | |
2519 | note=format_field(format, 'format_note', ' (%s)'), | |
2520 | ) | |
2521 | if format.get('protocol') is None: | |
2522 | format['protocol'] = determine_protocol(format) | |
2523 | if format.get('resolution') is None: | |
2524 | format['resolution'] = self.format_resolution(format, default=None) | |
2525 | if format.get('dynamic_range') is None and format.get('vcodec') != 'none': | |
2526 | format['dynamic_range'] = 'SDR' | |
2527 | if (info_dict.get('duration') and format.get('tbr') | |
2528 | and not format.get('filesize') and not format.get('filesize_approx')): | |
2529 | format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8) | |
2530 | ||
2531 | # Add HTTP headers, so that external programs can use them from the | |
2532 | # json output | |
2533 | full_format_info = info_dict.copy() | |
2534 | full_format_info.update(format) | |
2535 | format['http_headers'] = self._calc_headers(full_format_info) | |
2536 | # Remove private housekeeping stuff | |
2537 | if '__x_forwarded_for_ip' in info_dict: | |
2538 | del info_dict['__x_forwarded_for_ip'] | |
2539 | ||
2540 | if self.params.get('check_formats') is True: | |
2541 | formats = LazyList(self._check_formats(formats[::-1]), reverse=True) | |
2542 | ||
2543 | if not formats or formats[0] is not info_dict: | |
2544 | # only set the 'formats' fields if the original info_dict list them | |
2545 | # otherwise we end up with a circular reference, the first (and unique) | |
2546 | # element in the 'formats' field in info_dict is info_dict itself, | |
2547 | # which can't be exported to json | |
2548 | info_dict['formats'] = formats | |
2549 | ||
2550 | info_dict, _ = self.pre_process(info_dict) | |
2551 | ||
2552 | if self._match_entry(info_dict) is not None: | |
2553 | return info_dict | |
2554 | ||
2555 | self.post_extract(info_dict) | |
2556 | info_dict, _ = self.pre_process(info_dict, 'after_filter') | |
2557 | ||
2558 | # The pre-processors may have modified the formats | |
2559 | formats = info_dict.get('formats', [info_dict]) | |
2560 | ||
2561 | list_only = self.params.get('simulate') is None and ( | |
2562 | self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles')) | |
2563 | interactive_format_selection = not list_only and self.format_selector == '-' | |
2564 | if self.params.get('list_thumbnails'): | |
2565 | self.list_thumbnails(info_dict) | |
2566 | if self.params.get('listsubtitles'): | |
2567 | if 'automatic_captions' in info_dict: | |
2568 | self.list_subtitles( | |
2569 | info_dict['id'], automatic_captions, 'automatic captions') | |
2570 | self.list_subtitles(info_dict['id'], subtitles, 'subtitles') | |
2571 | if self.params.get('listformats') or interactive_format_selection: | |
2572 | self.list_formats(info_dict) | |
2573 | if list_only: | |
2574 | # Without this printing, -F --print-json will not work | |
2575 | self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True) | |
2576 | return | |
2577 | ||
2578 | format_selector = self.format_selector | |
2579 | if format_selector is None: | |
2580 | req_format = self._default_format_spec(info_dict, download=download) | |
2581 | self.write_debug('Default format spec: %s' % req_format) | |
2582 | format_selector = self.build_format_selector(req_format) | |
2583 | ||
2584 | while True: | |
2585 | if interactive_format_selection: | |
2586 | req_format = input( | |
2587 | self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS)) | |
2588 | try: | |
2589 | format_selector = self.build_format_selector(req_format) | |
2590 | except SyntaxError as err: | |
2591 | self.report_error(err, tb=False, is_error=False) | |
2592 | continue | |
2593 | ||
2594 | # While in format selection we may need to have an access to the original | |
2595 | # format set in order to calculate some metrics or do some processing. | |
2596 | # For now we need to be able to guess whether original formats provided | |
2597 | # by extractor are incomplete or not (i.e. whether extractor provides only | |
2598 | # video-only or audio-only formats) for proper formats selection for | |
2599 | # extractors with such incomplete formats (see | |
2600 | # https://github.com/ytdl-org/youtube-dl/pull/5556). | |
2601 | # Since formats may be filtered during format selection and may not match | |
2602 | # the original formats the results may be incorrect. Thus original formats | |
2603 | # or pre-calculated metrics should be passed to format selection routines | |
2604 | # as well. | |
2605 | # We will pass a context object containing all necessary additional data | |
2606 | # instead of just formats. | |
2607 | # This fixes incorrect format selection issue (see | |
2608 | # https://github.com/ytdl-org/youtube-dl/issues/10083). | |
2609 | incomplete_formats = ( | |
2610 | # All formats are video-only or | |
2611 | all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) | |
2612 | # all formats are audio-only | |
2613 | or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)) | |
2614 | ||
2615 | ctx = { | |
2616 | 'formats': formats, | |
2617 | 'incomplete_formats': incomplete_formats, | |
2618 | } | |
2619 | ||
2620 | formats_to_download = list(format_selector(ctx)) | |
2621 | if interactive_format_selection and not formats_to_download: | |
2622 | self.report_error('Requested format is not available', tb=False, is_error=False) | |
2623 | continue | |
2624 | break | |
2625 | ||
2626 | if not formats_to_download: | |
2627 | if not self.params.get('ignore_no_formats_error'): | |
2628 | raise ExtractorError('Requested format is not available', expected=True, | |
2629 | video_id=info_dict['id'], ie=info_dict['extractor']) | |
2630 | self.report_warning('Requested format is not available') | |
2631 | # Process what we can, even without any available formats. | |
2632 | formats_to_download = [{}] | |
2633 | ||
2634 | best_format = formats_to_download[-1] | |
2635 | if download: | |
2636 | if best_format: | |
2637 | self.to_screen( | |
2638 | f'[info] {info_dict["id"]}: Downloading {len(formats_to_download)} format(s): ' | |
2639 | + ', '.join([f['format_id'] for f in formats_to_download])) | |
2640 | max_downloads_reached = False | |
2641 | for i, fmt in enumerate(formats_to_download): | |
2642 | formats_to_download[i] = new_info = self._copy_infodict(info_dict) | |
2643 | new_info.update(fmt) | |
2644 | try: | |
2645 | self.process_info(new_info) | |
2646 | except MaxDownloadsReached: | |
2647 | max_downloads_reached = True | |
2648 | # Remove copied info | |
2649 | for key, val in tuple(new_info.items()): | |
2650 | if info_dict.get(key) == val: | |
2651 | new_info.pop(key) | |
2652 | if max_downloads_reached: | |
2653 | break | |
2654 | ||
2655 | write_archive = set(f.get('__write_download_archive', False) for f in formats_to_download) | |
2656 | assert write_archive.issubset({True, False, 'ignore'}) | |
2657 | if True in write_archive and False not in write_archive: | |
2658 | self.record_download_archive(info_dict) | |
2659 | ||
2660 | info_dict['requested_downloads'] = formats_to_download | |
2661 | info_dict = self.run_all_pps('after_video', info_dict) | |
2662 | if max_downloads_reached: | |
2663 | raise MaxDownloadsReached() | |
2664 | ||
2665 | # We update the info dict with the selected best quality format (backwards compatibility) | |
2666 | info_dict.update(best_format) | |
2667 | return info_dict | |
2668 | ||
2669 | def process_subtitles(self, video_id, normal_subtitles, automatic_captions): | |
2670 | """Select the requested subtitles and their format""" | |
2671 | available_subs = {} | |
2672 | if normal_subtitles and self.params.get('writesubtitles'): | |
2673 | available_subs.update(normal_subtitles) | |
2674 | if automatic_captions and self.params.get('writeautomaticsub'): | |
2675 | for lang, cap_info in automatic_captions.items(): | |
2676 | if lang not in available_subs: | |
2677 | available_subs[lang] = cap_info | |
2678 | ||
2679 | if (not self.params.get('writesubtitles') and not | |
2680 | self.params.get('writeautomaticsub') or not | |
2681 | available_subs): | |
2682 | return None | |
2683 | ||
2684 | all_sub_langs = available_subs.keys() | |
2685 | if self.params.get('allsubtitles', False): | |
2686 | requested_langs = all_sub_langs | |
2687 | elif self.params.get('subtitleslangs', False): | |
2688 | # A list is used so that the order of languages will be the same as | |
2689 | # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041 | |
2690 | requested_langs = [] | |
2691 | for lang_re in self.params.get('subtitleslangs'): | |
2692 | discard = lang_re[0] == '-' | |
2693 | if discard: | |
2694 | lang_re = lang_re[1:] | |
2695 | if lang_re == 'all': | |
2696 | if discard: | |
2697 | requested_langs = [] | |
2698 | else: | |
2699 | requested_langs.extend(all_sub_langs) | |
2700 | continue | |
2701 | current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs) | |
2702 | if discard: | |
2703 | for lang in current_langs: | |
2704 | while lang in requested_langs: | |
2705 | requested_langs.remove(lang) | |
2706 | else: | |
2707 | requested_langs.extend(current_langs) | |
2708 | requested_langs = orderedSet(requested_langs) | |
2709 | elif 'en' in available_subs: | |
2710 | requested_langs = ['en'] | |
2711 | else: | |
2712 | requested_langs = [list(all_sub_langs)[0]] | |
2713 | if requested_langs: | |
2714 | self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs)) | |
2715 | ||
2716 | formats_query = self.params.get('subtitlesformat', 'best') | |
2717 | formats_preference = formats_query.split('/') if formats_query else [] | |
2718 | subs = {} | |
2719 | for lang in requested_langs: | |
2720 | formats = available_subs.get(lang) | |
2721 | if formats is None: | |
2722 | self.report_warning('%s subtitles not available for %s' % (lang, video_id)) | |
2723 | continue | |
2724 | for ext in formats_preference: | |
2725 | if ext == 'best': | |
2726 | f = formats[-1] | |
2727 | break | |
2728 | matches = list(filter(lambda f: f['ext'] == ext, formats)) | |
2729 | if matches: | |
2730 | f = matches[-1] | |
2731 | break | |
2732 | else: | |
2733 | f = formats[-1] | |
2734 | self.report_warning( | |
2735 | 'No subtitle format found matching "%s" for language %s, ' | |
2736 | 'using %s' % (formats_query, lang, f['ext'])) | |
2737 | subs[lang] = f | |
2738 | return subs | |
2739 | ||
2740 | def _forceprint(self, key, info_dict): | |
2741 | if info_dict is None: | |
2742 | return | |
2743 | info_copy = info_dict.copy() | |
2744 | info_copy['formats_table'] = self.render_formats_table(info_dict) | |
2745 | info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict) | |
2746 | info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles')) | |
2747 | info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions')) | |
2748 | ||
2749 | def format_tmpl(tmpl): | |
2750 | mobj = re.match(r'\w+(=?)$', tmpl) | |
2751 | if mobj and mobj.group(1): | |
2752 | return f'{tmpl[:-1]} = %({tmpl[:-1]})r' | |
2753 | elif mobj: | |
2754 | return f'%({tmpl})s' | |
2755 | return tmpl | |
2756 | ||
2757 | for tmpl in self.params['forceprint'].get(key, []): | |
2758 | self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy)) | |
2759 | ||
2760 | for tmpl, file_tmpl in self.params['print_to_file'].get(key, []): | |
2761 | filename = self.evaluate_outtmpl(file_tmpl, info_dict) | |
2762 | tmpl = format_tmpl(tmpl) | |
2763 | self.to_screen(f'[info] Writing {tmpl!r} to: {filename}') | |
2764 | if self._ensure_dir_exists(filename): | |
2765 | with io.open(filename, 'a', encoding='utf-8') as f: | |
2766 | f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n') | |
2767 | ||
2768 | def __forced_printings(self, info_dict, filename, incomplete): | |
2769 | def print_mandatory(field, actual_field=None): | |
2770 | if actual_field is None: | |
2771 | actual_field = field | |
2772 | if (self.params.get('force%s' % field, False) | |
2773 | and (not incomplete or info_dict.get(actual_field) is not None)): | |
2774 | self.to_stdout(info_dict[actual_field]) | |
2775 | ||
2776 | def print_optional(field): | |
2777 | if (self.params.get('force%s' % field, False) | |
2778 | and info_dict.get(field) is not None): | |
2779 | self.to_stdout(info_dict[field]) | |
2780 | ||
2781 | info_dict = info_dict.copy() | |
2782 | if filename is not None: | |
2783 | info_dict['filename'] = filename | |
2784 | if info_dict.get('requested_formats') is not None: | |
2785 | # For RTMP URLs, also include the playpath | |
2786 | info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats']) | |
2787 | elif info_dict.get('url'): | |
2788 | info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '') | |
2789 | ||
2790 | if (self.params.get('forcejson') | |
2791 | or self.params['forceprint'].get('video') | |
2792 | or self.params['print_to_file'].get('video')): | |
2793 | self.post_extract(info_dict) | |
2794 | self._forceprint('video', info_dict) | |
2795 | ||
2796 | print_mandatory('title') | |
2797 | print_mandatory('id') | |
2798 | print_mandatory('url', 'urls') | |
2799 | print_optional('thumbnail') | |
2800 | print_optional('description') | |
2801 | print_optional('filename') | |
2802 | if self.params.get('forceduration') and info_dict.get('duration') is not None: | |
2803 | self.to_stdout(formatSeconds(info_dict['duration'])) | |
2804 | print_mandatory('format') | |
2805 | ||
2806 | if self.params.get('forcejson'): | |
2807 | self.to_stdout(json.dumps(self.sanitize_info(info_dict))) | |
2808 | ||
2809 | def dl(self, name, info, subtitle=False, test=False): | |
2810 | if not info.get('url'): | |
2811 | self.raise_no_formats(info, True) | |
2812 | ||
2813 | if test: | |
2814 | verbose = self.params.get('verbose') | |
2815 | params = { | |
2816 | 'test': True, | |
2817 | 'quiet': self.params.get('quiet') or not verbose, | |
2818 | 'verbose': verbose, | |
2819 | 'noprogress': not verbose, | |
2820 | 'nopart': True, | |
2821 | 'skip_unavailable_fragments': False, | |
2822 | 'keep_fragments': False, | |
2823 | 'overwrites': True, | |
2824 | '_no_ytdl_file': True, | |
2825 | } | |
2826 | else: | |
2827 | params = self.params | |
2828 | fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params) | |
2829 | if not test: | |
2830 | for ph in self._progress_hooks: | |
2831 | fd.add_progress_hook(ph) | |
2832 | urls = '", "'.join( | |
2833 | (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url']) | |
2834 | for f in info.get('requested_formats', []) or [info]) | |
2835 | self.write_debug('Invoking downloader on "%s"' % urls) | |
2836 | ||
2837 | # Note: Ideally info should be a deep-copied so that hooks cannot modify it. | |
2838 | # But it may contain objects that are not deep-copyable | |
2839 | new_info = self._copy_infodict(info) | |
2840 | if new_info.get('http_headers') is None: | |
2841 | new_info['http_headers'] = self._calc_headers(new_info) | |
2842 | return fd.download(name, new_info, subtitle) | |
2843 | ||
2844 | def existing_file(self, filepaths, *, default_overwrite=True): | |
2845 | existing_files = list(filter(os.path.exists, orderedSet(filepaths))) | |
2846 | if existing_files and not self.params.get('overwrites', default_overwrite): | |
2847 | return existing_files[0] | |
2848 | ||
2849 | for file in existing_files: | |
2850 | self.report_file_delete(file) | |
2851 | os.remove(file) | |
2852 | return None | |
2853 | ||
2854 | def process_info(self, info_dict): | |
2855 | """Process a single resolved IE result. (Modifies it in-place)""" | |
2856 | ||
2857 | assert info_dict.get('_type', 'video') == 'video' | |
2858 | original_infodict = info_dict | |
2859 | ||
2860 | if 'format' not in info_dict and 'ext' in info_dict: | |
2861 | info_dict['format'] = info_dict['ext'] | |
2862 | ||
2863 | # This is mostly just for backward compatibility of process_info | |
2864 | # As a side-effect, this allows for format-specific filters | |
2865 | if self._match_entry(info_dict) is not None: | |
2866 | info_dict['__write_download_archive'] = 'ignore' | |
2867 | return | |
2868 | ||
2869 | # Does nothing under normal operation - for backward compatibility of process_info | |
2870 | self.post_extract(info_dict) | |
2871 | ||
2872 | # info_dict['_filename'] needs to be set for backward compatibility | |
2873 | info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True) | |
2874 | temp_filename = self.prepare_filename(info_dict, 'temp') | |
2875 | files_to_move = {} | |
2876 | ||
2877 | self._num_downloads += 1 | |
2878 | ||
2879 | # Forced printings | |
2880 | self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict)) | |
2881 | ||
2882 | if self.params.get('simulate'): | |
2883 | info_dict['__write_download_archive'] = self.params.get('force_write_download_archive') | |
2884 | return | |
2885 | ||
2886 | if full_filename is None: | |
2887 | return | |
2888 | if not self._ensure_dir_exists(encodeFilename(full_filename)): | |
2889 | return | |
2890 | if not self._ensure_dir_exists(encodeFilename(temp_filename)): | |
2891 | return | |
2892 | ||
2893 | if self._write_description('video', info_dict, | |
2894 | self.prepare_filename(info_dict, 'description')) is None: | |
2895 | return | |
2896 | ||
2897 | sub_files = self._write_subtitles(info_dict, temp_filename) | |
2898 | if sub_files is None: | |
2899 | return | |
2900 | files_to_move.update(dict(sub_files)) | |
2901 | ||
2902 | thumb_files = self._write_thumbnails( | |
2903 | 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail')) | |
2904 | if thumb_files is None: | |
2905 | return | |
2906 | files_to_move.update(dict(thumb_files)) | |
2907 | ||
2908 | infofn = self.prepare_filename(info_dict, 'infojson') | |
2909 | _infojson_written = self._write_info_json('video', info_dict, infofn) | |
2910 | if _infojson_written: | |
2911 | info_dict['infojson_filename'] = infofn | |
2912 | # For backward compatibility, even though it was a private field | |
2913 | info_dict['__infojson_filename'] = infofn | |
2914 | elif _infojson_written is None: | |
2915 | return | |
2916 | ||
2917 | # Note: Annotations are deprecated | |
2918 | annofn = None | |
2919 | if self.params.get('writeannotations', False): | |
2920 | annofn = self.prepare_filename(info_dict, 'annotation') | |
2921 | if annofn: | |
2922 | if not self._ensure_dir_exists(encodeFilename(annofn)): | |
2923 | return | |
2924 | if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)): | |
2925 | self.to_screen('[info] Video annotations are already present') | |
2926 | elif not info_dict.get('annotations'): | |
2927 | self.report_warning('There are no annotations to write.') | |
2928 | else: | |
2929 | try: | |
2930 | self.to_screen('[info] Writing video annotations to: ' + annofn) | |
2931 | with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile: | |
2932 | annofile.write(info_dict['annotations']) | |
2933 | except (KeyError, TypeError): | |
2934 | self.report_warning('There are no annotations to write.') | |
2935 | except (OSError, IOError): | |
2936 | self.report_error('Cannot write annotations file: ' + annofn) | |
2937 | return | |
2938 | ||
2939 | # Write internet shortcut files | |
2940 | def _write_link_file(link_type): | |
2941 | url = try_get(info_dict['webpage_url'], iri_to_uri) | |
2942 | if not url: | |
2943 | self.report_warning( | |
2944 | f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown') | |
2945 | return True | |
2946 | linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext')) | |
2947 | if not self._ensure_dir_exists(encodeFilename(linkfn)): | |
2948 | return False | |
2949 | if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)): | |
2950 | self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present') | |
2951 | return True | |
2952 | try: | |
2953 | self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}') | |
2954 | with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', | |
2955 | newline='\r\n' if link_type == 'url' else '\n') as linkfile: | |
2956 | template_vars = {'url': url} | |
2957 | if link_type == 'desktop': | |
2958 | template_vars['filename'] = linkfn[:-(len(link_type) + 1)] | |
2959 | linkfile.write(LINK_TEMPLATES[link_type] % template_vars) | |
2960 | except (OSError, IOError): | |
2961 | self.report_error(f'Cannot write internet shortcut {linkfn}') | |
2962 | return False | |
2963 | return True | |
2964 | ||
2965 | write_links = { | |
2966 | 'url': self.params.get('writeurllink'), | |
2967 | 'webloc': self.params.get('writewebloclink'), | |
2968 | 'desktop': self.params.get('writedesktoplink'), | |
2969 | } | |
2970 | if self.params.get('writelink'): | |
2971 | link_type = ('webloc' if sys.platform == 'darwin' | |
2972 | else 'desktop' if sys.platform.startswith('linux') | |
2973 | else 'url') | |
2974 | write_links[link_type] = True | |
2975 | ||
2976 | if any(should_write and not _write_link_file(link_type) | |
2977 | for link_type, should_write in write_links.items()): | |
2978 | return | |
2979 | ||
2980 | def replace_info_dict(new_info): | |
2981 | nonlocal info_dict | |
2982 | if new_info == info_dict: | |
2983 | return | |
2984 | info_dict.clear() | |
2985 | info_dict.update(new_info) | |
2986 | ||
2987 | try: | |
2988 | new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move) | |
2989 | replace_info_dict(new_info) | |
2990 | except PostProcessingError as err: | |
2991 | self.report_error('Preprocessing: %s' % str(err)) | |
2992 | return | |
2993 | ||
2994 | if self.params.get('skip_download'): | |
2995 | info_dict['filepath'] = temp_filename | |
2996 | info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename))) | |
2997 | info_dict['__files_to_move'] = files_to_move | |
2998 | replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)) | |
2999 | info_dict['__write_download_archive'] = self.params.get('force_write_download_archive') | |
3000 | else: | |
3001 | # Download | |
3002 | info_dict.setdefault('__postprocessors', []) | |
3003 | try: | |
3004 | ||
3005 | def existing_video_file(*filepaths): | |
3006 | ext = info_dict.get('ext') | |
3007 | converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext) | |
3008 | file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)), | |
3009 | default_overwrite=False) | |
3010 | if file: | |
3011 | info_dict['ext'] = os.path.splitext(file)[1][1:] | |
3012 | return file | |
3013 | ||
3014 | success = True | |
3015 | if info_dict.get('requested_formats') is not None: | |
3016 | ||
3017 | def compatible_formats(formats): | |
3018 | # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them. | |
3019 | video_formats = [format for format in formats if format.get('vcodec') != 'none'] | |
3020 | audio_formats = [format for format in formats if format.get('acodec') != 'none'] | |
3021 | if len(video_formats) > 2 or len(audio_formats) > 2: | |
3022 | return False | |
3023 | ||
3024 | # Check extension | |
3025 | exts = set(format.get('ext') for format in formats) | |
3026 | COMPATIBLE_EXTS = ( | |
3027 | set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')), | |
3028 | set(('webm',)), | |
3029 | ) | |
3030 | for ext_sets in COMPATIBLE_EXTS: | |
3031 | if ext_sets.issuperset(exts): | |
3032 | return True | |
3033 | # TODO: Check acodec/vcodec | |
3034 | return False | |
3035 | ||
3036 | requested_formats = info_dict['requested_formats'] | |
3037 | old_ext = info_dict['ext'] | |
3038 | if self.params.get('merge_output_format') is None: | |
3039 | if not compatible_formats(requested_formats): | |
3040 | info_dict['ext'] = 'mkv' | |
3041 | self.report_warning( | |
3042 | 'Requested formats are incompatible for merge and will be merged into mkv') | |
3043 | if (info_dict['ext'] == 'webm' | |
3044 | and info_dict.get('thumbnails') | |
3045 | # check with type instead of pp_key, __name__, or isinstance | |
3046 | # since we dont want any custom PPs to trigger this | |
3047 | and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): | |
3048 | info_dict['ext'] = 'mkv' | |
3049 | self.report_warning( | |
3050 | 'webm doesn\'t support embedding a thumbnail, mkv will be used') | |
3051 | new_ext = info_dict['ext'] | |
3052 | ||
3053 | def correct_ext(filename, ext=new_ext): | |
3054 | if filename == '-': | |
3055 | return filename | |
3056 | filename_real_ext = os.path.splitext(filename)[1][1:] | |
3057 | filename_wo_ext = ( | |
3058 | os.path.splitext(filename)[0] | |
3059 | if filename_real_ext in (old_ext, new_ext) | |
3060 | else filename) | |
3061 | return '%s.%s' % (filename_wo_ext, ext) | |
3062 | ||
3063 | # Ensure filename always has a correct extension for successful merge | |
3064 | full_filename = correct_ext(full_filename) | |
3065 | temp_filename = correct_ext(temp_filename) | |
3066 | dl_filename = existing_video_file(full_filename, temp_filename) | |
3067 | info_dict['__real_download'] = False | |
3068 | ||
3069 | downloaded = [] | |
3070 | merger = FFmpegMergerPP(self) | |
3071 | ||
3072 | fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-') | |
3073 | if dl_filename is not None: | |
3074 | self.report_file_already_downloaded(dl_filename) | |
3075 | elif fd: | |
3076 | for f in requested_formats if fd != FFmpegFD else []: | |
3077 | f['filepath'] = fname = prepend_extension( | |
3078 | correct_ext(temp_filename, info_dict['ext']), | |
3079 | 'f%s' % f['format_id'], info_dict['ext']) | |
3080 | downloaded.append(fname) | |
3081 | info_dict['url'] = '\n'.join(f['url'] for f in requested_formats) | |
3082 | success, real_download = self.dl(temp_filename, info_dict) | |
3083 | info_dict['__real_download'] = real_download | |
3084 | else: | |
3085 | if self.params.get('allow_unplayable_formats'): | |
3086 | self.report_warning( | |
3087 | 'You have requested merging of multiple formats ' | |
3088 | 'while also allowing unplayable formats to be downloaded. ' | |
3089 | 'The formats won\'t be merged to prevent data corruption.') | |
3090 | elif not merger.available: | |
3091 | msg = 'You have requested merging of multiple formats but ffmpeg is not installed' | |
3092 | if not self.params.get('ignoreerrors'): | |
3093 | self.report_error(f'{msg}. Aborting due to --abort-on-error') | |
3094 | return | |
3095 | self.report_warning(f'{msg}. The formats won\'t be merged') | |
3096 | ||
3097 | if temp_filename == '-': | |
3098 | reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params) | |
3099 | else 'but the formats are incompatible for simultaneous download' if merger.available | |
3100 | else 'but ffmpeg is not installed') | |
3101 | self.report_warning( | |
3102 | f'You have requested downloading multiple formats to stdout {reason}. ' | |
3103 | 'The formats will be streamed one after the other') | |
3104 | fname = temp_filename | |
3105 | for f in requested_formats: | |
3106 | new_info = dict(info_dict) | |
3107 | del new_info['requested_formats'] | |
3108 | new_info.update(f) | |
3109 | if temp_filename != '-': | |
3110 | fname = prepend_extension( | |
3111 | correct_ext(temp_filename, new_info['ext']), | |
3112 | 'f%s' % f['format_id'], new_info['ext']) | |
3113 | if not self._ensure_dir_exists(fname): | |
3114 | return | |
3115 | f['filepath'] = fname | |
3116 | downloaded.append(fname) | |
3117 | partial_success, real_download = self.dl(fname, new_info) | |
3118 | info_dict['__real_download'] = info_dict['__real_download'] or real_download | |
3119 | success = success and partial_success | |
3120 | ||
3121 | if downloaded and merger.available and not self.params.get('allow_unplayable_formats'): | |
3122 | info_dict['__postprocessors'].append(merger) | |
3123 | info_dict['__files_to_merge'] = downloaded | |
3124 | # Even if there were no downloads, it is being merged only now | |
3125 | info_dict['__real_download'] = True | |
3126 | else: | |
3127 | for file in downloaded: | |
3128 | files_to_move[file] = None | |
3129 | else: | |
3130 | # Just a single file | |
3131 | dl_filename = existing_video_file(full_filename, temp_filename) | |
3132 | if dl_filename is None or dl_filename == temp_filename: | |
3133 | # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part. | |
3134 | # So we should try to resume the download | |
3135 | success, real_download = self.dl(temp_filename, info_dict) | |
3136 | info_dict['__real_download'] = real_download | |
3137 | else: | |
3138 | self.report_file_already_downloaded(dl_filename) | |
3139 | ||
3140 | dl_filename = dl_filename or temp_filename | |
3141 | info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename))) | |
3142 | ||
3143 | except network_exceptions as err: | |
3144 | self.report_error('unable to download video data: %s' % error_to_compat_str(err)) | |
3145 | return | |
3146 | except (OSError, IOError) as err: | |
3147 | raise UnavailableVideoError(err) | |
3148 | except (ContentTooShortError, ) as err: | |
3149 | self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded)) | |
3150 | return | |
3151 | ||
3152 | if success and full_filename != '-': | |
3153 | ||
3154 | def fixup(): | |
3155 | do_fixup = True | |
3156 | fixup_policy = self.params.get('fixup') | |
3157 | vid = info_dict['id'] | |
3158 | ||
3159 | if fixup_policy in ('ignore', 'never'): | |
3160 | return | |
3161 | elif fixup_policy == 'warn': | |
3162 | do_fixup = False | |
3163 | elif fixup_policy != 'force': | |
3164 | assert fixup_policy in ('detect_or_warn', None) | |
3165 | if not info_dict.get('__real_download'): | |
3166 | do_fixup = False | |
3167 | ||
3168 | def ffmpeg_fixup(cndn, msg, cls): | |
3169 | if not cndn: | |
3170 | return | |
3171 | if not do_fixup: | |
3172 | self.report_warning(f'{vid}: {msg}') | |
3173 | return | |
3174 | pp = cls(self) | |
3175 | if pp.available: | |
3176 | info_dict['__postprocessors'].append(pp) | |
3177 | else: | |
3178 | self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically') | |
3179 | ||
3180 | stretched_ratio = info_dict.get('stretched_ratio') | |
3181 | ffmpeg_fixup( | |
3182 | stretched_ratio not in (1, None), | |
3183 | f'Non-uniform pixel ratio {stretched_ratio}', | |
3184 | FFmpegFixupStretchedPP) | |
3185 | ||
3186 | ffmpeg_fixup( | |
3187 | (info_dict.get('requested_formats') is None | |
3188 | and info_dict.get('container') == 'm4a_dash' | |
3189 | and info_dict.get('ext') == 'm4a'), | |
3190 | 'writing DASH m4a. Only some players support this container', | |
3191 | FFmpegFixupM4aPP) | |
3192 | ||
3193 | downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None | |
3194 | downloader = downloader.__name__ if downloader else None | |
3195 | ||
3196 | if info_dict.get('requested_formats') is None: # Not necessary if doing merger | |
3197 | ffmpeg_fixup(downloader == 'HlsFD', | |
3198 | 'Possible MPEG-TS in MP4 container or malformed AAC timestamps', | |
3199 | FFmpegFixupM3u8PP) | |
3200 | ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD', | |
3201 | 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP) | |
3202 | ||
3203 | ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP) | |
3204 | ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP) | |
3205 | ||
3206 | fixup() | |
3207 | try: | |
3208 | replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move)) | |
3209 | except PostProcessingError as err: | |
3210 | self.report_error('Postprocessing: %s' % str(err)) | |
3211 | return | |
3212 | try: | |
3213 | for ph in self._post_hooks: | |
3214 | ph(info_dict['filepath']) | |
3215 | except Exception as err: | |
3216 | self.report_error('post hooks: %s' % str(err)) | |
3217 | return | |
3218 | info_dict['__write_download_archive'] = True | |
3219 | ||
3220 | if self.params.get('force_write_download_archive'): | |
3221 | info_dict['__write_download_archive'] = True | |
3222 | ||
3223 | # Make sure the info_dict was modified in-place | |
3224 | assert info_dict is original_infodict | |
3225 | ||
3226 | max_downloads = self.params.get('max_downloads') | |
3227 | if max_downloads is not None and self._num_downloads >= int(max_downloads): | |
3228 | raise MaxDownloadsReached() | |
3229 | ||
3230 | def __download_wrapper(self, func): | |
3231 | @functools.wraps(func) | |
3232 | def wrapper(*args, **kwargs): | |
3233 | try: | |
3234 | res = func(*args, **kwargs) | |
3235 | except UnavailableVideoError as e: | |
3236 | self.report_error(e) | |
3237 | except MaxDownloadsReached as e: | |
3238 | self.to_screen(f'[info] {e}') | |
3239 | raise | |
3240 | except DownloadCancelled as e: | |
3241 | self.to_screen(f'[info] {e}') | |
3242 | if not self.params.get('break_per_url'): | |
3243 | raise | |
3244 | else: | |
3245 | if self.params.get('dump_single_json', False): | |
3246 | self.post_extract(res) | |
3247 | self.to_stdout(json.dumps(self.sanitize_info(res))) | |
3248 | return wrapper | |
3249 | ||
3250 | def download(self, url_list): | |
3251 | """Download a given list of URLs.""" | |
3252 | url_list = variadic(url_list) # Passing a single URL is a common mistake | |
3253 | outtmpl = self.outtmpl_dict['default'] | |
3254 | if (len(url_list) > 1 | |
3255 | and outtmpl != '-' | |
3256 | and '%' not in outtmpl | |
3257 | and self.params.get('max_downloads') != 1): | |
3258 | raise SameFileError(outtmpl) | |
3259 | ||
3260 | for url in url_list: | |
3261 | self.__download_wrapper(self.extract_info)( | |
3262 | url, force_generic_extractor=self.params.get('force_generic_extractor', False)) | |
3263 | ||
3264 | return self._download_retcode | |
3265 | ||
3266 | def download_with_info_file(self, info_filename): | |
3267 | with contextlib.closing(fileinput.FileInput( | |
3268 | [info_filename], mode='r', | |
3269 | openhook=fileinput.hook_encoded('utf-8'))) as f: | |
3270 | # FileInput doesn't have a read method, we can't call json.load | |
3271 | info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True)) | |
3272 | try: | |
3273 | self.__download_wrapper(self.process_ie_result)(info, download=True) | |
3274 | except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e: | |
3275 | if not isinstance(e, EntryNotInPlaylist): | |
3276 | self.to_stderr('\r') | |
3277 | webpage_url = info.get('webpage_url') | |
3278 | if webpage_url is not None: | |
3279 | self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}') | |
3280 | return self.download([webpage_url]) | |
3281 | else: | |
3282 | raise | |
3283 | return self._download_retcode | |
3284 | ||
3285 | @staticmethod | |
3286 | def sanitize_info(info_dict, remove_private_keys=False): | |
3287 | ''' Sanitize the infodict for converting to json ''' | |
3288 | if info_dict is None: | |
3289 | return info_dict | |
3290 | info_dict.setdefault('epoch', int(time.time())) | |
3291 | info_dict.setdefault('_type', 'video') | |
3292 | ||
3293 | if remove_private_keys: | |
3294 | reject = lambda k, v: v is None or (k.startswith('_') and k != '_type') or k in { | |
3295 | 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries', | |
3296 | 'entries', 'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber', | |
3297 | } | |
3298 | else: | |
3299 | reject = lambda k, v: False | |
3300 | ||
3301 | def filter_fn(obj): | |
3302 | if isinstance(obj, dict): | |
3303 | return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)} | |
3304 | elif isinstance(obj, (list, tuple, set, LazyList)): | |
3305 | return list(map(filter_fn, obj)) | |
3306 | elif obj is None or isinstance(obj, (str, int, float, bool)): | |
3307 | return obj | |
3308 | else: | |
3309 | return repr(obj) | |
3310 | ||
3311 | return filter_fn(info_dict) | |
3312 | ||
3313 | @staticmethod | |
3314 | def filter_requested_info(info_dict, actually_filter=True): | |
3315 | ''' Alias of sanitize_info for backward compatibility ''' | |
3316 | return YoutubeDL.sanitize_info(info_dict, actually_filter) | |
3317 | ||
3318 | @staticmethod | |
3319 | def post_extract(info_dict): | |
3320 | def actual_post_extract(info_dict): | |
3321 | if info_dict.get('_type') in ('playlist', 'multi_video'): | |
3322 | for video_dict in info_dict.get('entries', {}): | |
3323 | actual_post_extract(video_dict or {}) | |
3324 | return | |
3325 | ||
3326 | post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {}) | |
3327 | info_dict.update(post_extractor()) | |
3328 | ||
3329 | actual_post_extract(info_dict or {}) | |
3330 | ||
3331 | def run_pp(self, pp, infodict): | |
3332 | files_to_delete = [] | |
3333 | if '__files_to_move' not in infodict: | |
3334 | infodict['__files_to_move'] = {} | |
3335 | try: | |
3336 | files_to_delete, infodict = pp.run(infodict) | |
3337 | except PostProcessingError as e: | |
3338 | # Must be True and not 'only_download' | |
3339 | if self.params.get('ignoreerrors') is True: | |
3340 | self.report_error(e) | |
3341 | return infodict | |
3342 | raise | |
3343 | ||
3344 | if not files_to_delete: | |
3345 | return infodict | |
3346 | if self.params.get('keepvideo', False): | |
3347 | for f in files_to_delete: | |
3348 | infodict['__files_to_move'].setdefault(f, '') | |
3349 | else: | |
3350 | for old_filename in set(files_to_delete): | |
3351 | self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename) | |
3352 | try: | |
3353 | os.remove(encodeFilename(old_filename)) | |
3354 | except (IOError, OSError): | |
3355 | self.report_warning('Unable to remove downloaded original file') | |
3356 | if old_filename in infodict['__files_to_move']: | |
3357 | del infodict['__files_to_move'][old_filename] | |
3358 | return infodict | |
3359 | ||
3360 | def run_all_pps(self, key, info, *, additional_pps=None): | |
3361 | self._forceprint(key, info) | |
3362 | for pp in (additional_pps or []) + self._pps[key]: | |
3363 | info = self.run_pp(pp, info) | |
3364 | return info | |
3365 | ||
3366 | def pre_process(self, ie_info, key='pre_process', files_to_move=None): | |
3367 | info = dict(ie_info) | |
3368 | info['__files_to_move'] = files_to_move or {} | |
3369 | info = self.run_all_pps(key, info) | |
3370 | return info, info.pop('__files_to_move', None) | |
3371 | ||
3372 | def post_process(self, filename, info, files_to_move=None): | |
3373 | """Run all the postprocessors on the given file.""" | |
3374 | info['filepath'] = filename | |
3375 | info['__files_to_move'] = files_to_move or {} | |
3376 | info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors')) | |
3377 | info = self.run_pp(MoveFilesAfterDownloadPP(self), info) | |
3378 | del info['__files_to_move'] | |
3379 | return self.run_all_pps('after_move', info) | |
3380 | ||
3381 | def _make_archive_id(self, info_dict): | |
3382 | video_id = info_dict.get('id') | |
3383 | if not video_id: | |
3384 | return | |
3385 | # Future-proof against any change in case | |
3386 | # and backwards compatibility with prior versions | |
3387 | extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist | |
3388 | if extractor is None: | |
3389 | url = str_or_none(info_dict.get('url')) | |
3390 | if not url: | |
3391 | return | |
3392 | # Try to find matching extractor for the URL and take its ie_key | |
3393 | for ie_key, ie in self._ies.items(): | |
3394 | if ie.suitable(url): | |
3395 | extractor = ie_key | |
3396 | break | |
3397 | else: | |
3398 | return | |
3399 | return '%s %s' % (extractor.lower(), video_id) | |
3400 | ||
3401 | def in_download_archive(self, info_dict): | |
3402 | fn = self.params.get('download_archive') | |
3403 | if fn is None: | |
3404 | return False | |
3405 | ||
3406 | vid_id = self._make_archive_id(info_dict) | |
3407 | if not vid_id: | |
3408 | return False # Incomplete video information | |
3409 | ||
3410 | return vid_id in self.archive | |
3411 | ||
3412 | def record_download_archive(self, info_dict): | |
3413 | fn = self.params.get('download_archive') | |
3414 | if fn is None: | |
3415 | return | |
3416 | vid_id = self._make_archive_id(info_dict) | |
3417 | assert vid_id | |
3418 | self.write_debug(f'Adding to archive: {vid_id}') | |
3419 | with locked_file(fn, 'a', encoding='utf-8') as archive_file: | |
3420 | archive_file.write(vid_id + '\n') | |
3421 | self.archive.add(vid_id) | |
3422 | ||
3423 | @staticmethod | |
3424 | def format_resolution(format, default='unknown'): | |
3425 | if format.get('vcodec') == 'none' and format.get('acodec') != 'none': | |
3426 | return 'audio only' | |
3427 | if format.get('resolution') is not None: | |
3428 | return format['resolution'] | |
3429 | if format.get('width') and format.get('height'): | |
3430 | return '%dx%d' % (format['width'], format['height']) | |
3431 | elif format.get('height'): | |
3432 | return '%sp' % format['height'] | |
3433 | elif format.get('width'): | |
3434 | return '%dx?' % format['width'] | |
3435 | return default | |
3436 | ||
3437 | def _list_format_headers(self, *headers): | |
3438 | if self.params.get('listformats_table', True) is not False: | |
3439 | return [self._format_screen(header, self.Styles.HEADERS) for header in headers] | |
3440 | return headers | |
3441 | ||
3442 | def _format_note(self, fdict): | |
3443 | res = '' | |
3444 | if fdict.get('ext') in ['f4f', 'f4m']: | |
3445 | res += '(unsupported)' | |
3446 | if fdict.get('language'): | |
3447 | if res: | |
3448 | res += ' ' | |
3449 | res += '[%s]' % fdict['language'] | |
3450 | if fdict.get('format_note') is not None: | |
3451 | if res: | |
3452 | res += ' ' | |
3453 | res += fdict['format_note'] | |
3454 | if fdict.get('tbr') is not None: | |
3455 | if res: | |
3456 | res += ', ' | |
3457 | res += '%4dk' % fdict['tbr'] | |
3458 | if fdict.get('container') is not None: | |
3459 | if res: | |
3460 | res += ', ' | |
3461 | res += '%s container' % fdict['container'] | |
3462 | if (fdict.get('vcodec') is not None | |
3463 | and fdict.get('vcodec') != 'none'): | |
3464 | if res: | |
3465 | res += ', ' | |
3466 | res += fdict['vcodec'] | |
3467 | if fdict.get('vbr') is not None: | |
3468 | res += '@' | |
3469 | elif fdict.get('vbr') is not None and fdict.get('abr') is not None: | |
3470 | res += 'video@' | |
3471 | if fdict.get('vbr') is not None: | |
3472 | res += '%4dk' % fdict['vbr'] | |
3473 | if fdict.get('fps') is not None: | |
3474 | if res: | |
3475 | res += ', ' | |
3476 | res += '%sfps' % fdict['fps'] | |
3477 | if fdict.get('acodec') is not None: | |
3478 | if res: | |
3479 | res += ', ' | |
3480 | if fdict['acodec'] == 'none': | |
3481 | res += 'video only' | |
3482 | else: | |
3483 | res += '%-5s' % fdict['acodec'] | |
3484 | elif fdict.get('abr') is not None: | |
3485 | if res: | |
3486 | res += ', ' | |
3487 | res += 'audio' | |
3488 | if fdict.get('abr') is not None: | |
3489 | res += '@%3dk' % fdict['abr'] | |
3490 | if fdict.get('asr') is not None: | |
3491 | res += ' (%5dHz)' % fdict['asr'] | |
3492 | if fdict.get('filesize') is not None: | |
3493 | if res: | |
3494 | res += ', ' | |
3495 | res += format_bytes(fdict['filesize']) | |
3496 | elif fdict.get('filesize_approx') is not None: | |
3497 | if res: | |
3498 | res += ', ' | |
3499 | res += '~' + format_bytes(fdict['filesize_approx']) | |
3500 | return res | |
3501 | ||
3502 | def render_formats_table(self, info_dict): | |
3503 | if not info_dict.get('formats') and not info_dict.get('url'): | |
3504 | return None | |
3505 | ||
3506 | formats = info_dict.get('formats', [info_dict]) | |
3507 | if not self.params.get('listformats_table', True) is not False: | |
3508 | table = [ | |
3509 | [ | |
3510 | format_field(f, 'format_id'), | |
3511 | format_field(f, 'ext'), | |
3512 | self.format_resolution(f), | |
3513 | self._format_note(f) | |
3514 | ] for f in formats if f.get('preference') is None or f['preference'] >= -1000] | |
3515 | return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1) | |
3516 | ||
3517 | delim = self._format_screen('\u2502', self.Styles.DELIM, '|', test_encoding=True) | |
3518 | table = [ | |
3519 | [ | |
3520 | self._format_screen(format_field(f, 'format_id'), self.Styles.ID), | |
3521 | format_field(f, 'ext'), | |
3522 | format_field(f, func=self.format_resolution, ignore=('audio only', 'images')), | |
3523 | format_field(f, 'fps', '\t%d'), | |
3524 | format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''), | |
3525 | delim, | |
3526 | format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes), | |
3527 | format_field(f, 'tbr', '\t%dk'), | |
3528 | shorten_protocol_name(f.get('protocol', '')), | |
3529 | delim, | |
3530 | format_field(f, 'vcodec', default='unknown').replace( | |
3531 | 'none', 'images' if f.get('acodec') == 'none' | |
3532 | else self._format_screen('audio only', self.Styles.SUPPRESS)), | |
3533 | format_field(f, 'vbr', '\t%dk'), | |
3534 | format_field(f, 'acodec', default='unknown').replace( | |
3535 | 'none', '' if f.get('vcodec') == 'none' | |
3536 | else self._format_screen('video only', self.Styles.SUPPRESS)), | |
3537 | format_field(f, 'abr', '\t%dk'), | |
3538 | format_field(f, 'asr', '\t%dHz'), | |
3539 | join_nonempty( | |
3540 | self._format_screen('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None, | |
3541 | format_field(f, 'language', '[%s]'), | |
3542 | join_nonempty(format_field(f, 'format_note'), | |
3543 | format_field(f, 'container', ignore=(None, f.get('ext'))), | |
3544 | delim=', '), | |
3545 | delim=' '), | |
3546 | ] for f in formats if f.get('preference') is None or f['preference'] >= -1000] | |
3547 | header_line = self._list_format_headers( | |
3548 | 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO', | |
3549 | delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO') | |
3550 | ||
3551 | return render_table( | |
3552 | header_line, table, hide_empty=True, | |
3553 | delim=self._format_screen('\u2500', self.Styles.DELIM, '-', test_encoding=True)) | |
3554 | ||
3555 | def render_thumbnails_table(self, info_dict): | |
3556 | thumbnails = list(info_dict.get('thumbnails') or []) | |
3557 | if not thumbnails: | |
3558 | return None | |
3559 | return render_table( | |
3560 | self._list_format_headers('ID', 'Width', 'Height', 'URL'), | |
3561 | [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]) | |
3562 | ||
3563 | def render_subtitles_table(self, video_id, subtitles): | |
3564 | def _row(lang, formats): | |
3565 | exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats))) | |
3566 | if len(set(names)) == 1: | |
3567 | names = [] if names[0] == 'unknown' else names[:1] | |
3568 | return [lang, ', '.join(names), ', '.join(exts)] | |
3569 | ||
3570 | if not subtitles: | |
3571 | return None | |
3572 | return render_table( | |
3573 | self._list_format_headers('Language', 'Name', 'Formats'), | |
3574 | [_row(lang, formats) for lang, formats in subtitles.items()], | |
3575 | hide_empty=True) | |
3576 | ||
3577 | def __list_table(self, video_id, name, func, *args): | |
3578 | table = func(*args) | |
3579 | if not table: | |
3580 | self.to_screen(f'{video_id} has no {name}') | |
3581 | return | |
3582 | self.to_screen(f'[info] Available {name} for {video_id}:') | |
3583 | self.to_stdout(table) | |
3584 | ||
3585 | def list_formats(self, info_dict): | |
3586 | self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict) | |
3587 | ||
3588 | def list_thumbnails(self, info_dict): | |
3589 | self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict) | |
3590 | ||
3591 | def list_subtitles(self, video_id, subtitles, name='subtitles'): | |
3592 | self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles) | |
3593 | ||
3594 | def urlopen(self, req): | |
3595 | """ Start an HTTP download """ | |
3596 | if isinstance(req, compat_basestring): | |
3597 | req = sanitized_Request(req) | |
3598 | return self._opener.open(req, timeout=self._socket_timeout) | |
3599 | ||
3600 | def print_debug_header(self): | |
3601 | if not self.params.get('verbose'): | |
3602 | return | |
3603 | ||
3604 | def get_encoding(stream): | |
3605 | ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__)) | |
3606 | if not supports_terminal_sequences(stream): | |
3607 | from .compat import WINDOWS_VT_MODE | |
3608 | ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)' | |
3609 | return ret | |
3610 | ||
3611 | encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % ( | |
3612 | locale.getpreferredencoding(), | |
3613 | sys.getfilesystemencoding(), | |
3614 | get_encoding(self._out_files['screen']), get_encoding(self._out_files['error']), | |
3615 | self.get_encoding()) | |
3616 | ||
3617 | logger = self.params.get('logger') | |
3618 | if logger: | |
3619 | write_debug = lambda msg: logger.debug(f'[debug] {msg}') | |
3620 | write_debug(encoding_str) | |
3621 | else: | |
3622 | write_string(f'[debug] {encoding_str}\n', encoding=None) | |
3623 | write_debug = lambda msg: self._write_string(f'[debug] {msg}\n') | |
3624 | ||
3625 | source = detect_variant() | |
3626 | write_debug(join_nonempty( | |
3627 | 'yt-dlp version', __version__, | |
3628 | f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '', | |
3629 | '' if source == 'unknown' else f'({source})', | |
3630 | delim=' ')) | |
3631 | if not _LAZY_LOADER: | |
3632 | if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'): | |
3633 | write_debug('Lazy loading extractors is forcibly disabled') | |
3634 | else: | |
3635 | write_debug('Lazy loading extractors is disabled') | |
3636 | if plugin_extractors or plugin_postprocessors: | |
3637 | write_debug('Plugins: %s' % [ | |
3638 | '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}') | |
3639 | for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())]) | |
3640 | if self.params.get('compat_opts'): | |
3641 | write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts'))) | |
3642 | ||
3643 | if source == 'source': | |
3644 | try: | |
3645 | sp = Popen( | |
3646 | ['git', 'rev-parse', '--short', 'HEAD'], | |
3647 | stdout=subprocess.PIPE, stderr=subprocess.PIPE, | |
3648 | cwd=os.path.dirname(os.path.abspath(__file__))) | |
3649 | out, err = sp.communicate_or_kill() | |
3650 | out = out.decode().strip() | |
3651 | if re.match('[0-9a-f]+', out): | |
3652 | write_debug('Git HEAD: %s' % out) | |
3653 | except Exception: | |
3654 | try: | |
3655 | sys.exc_clear() | |
3656 | except Exception: | |
3657 | pass | |
3658 | ||
3659 | def python_implementation(): | |
3660 | impl_name = platform.python_implementation() | |
3661 | if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'): | |
3662 | return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3] | |
3663 | return impl_name | |
3664 | ||
3665 | write_debug('Python version %s (%s %s) - %s' % ( | |
3666 | platform.python_version(), | |
3667 | python_implementation(), | |
3668 | platform.architecture()[0], | |
3669 | platform_name())) | |
3670 | ||
3671 | exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self) | |
3672 | ffmpeg_features = {key for key, val in ffmpeg_features.items() if val} | |
3673 | if ffmpeg_features: | |
3674 | exe_versions['ffmpeg'] += ' (%s)' % ','.join(ffmpeg_features) | |
3675 | ||
3676 | exe_versions['rtmpdump'] = rtmpdump_version() | |
3677 | exe_versions['phantomjs'] = PhantomJSwrapper._version() | |
3678 | exe_str = ', '.join( | |
3679 | f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v | |
3680 | ) or 'none' | |
3681 | write_debug('exe versions: %s' % exe_str) | |
3682 | ||
3683 | from .downloader.websocket import has_websockets | |
3684 | from .postprocessor.embedthumbnail import has_mutagen | |
3685 | from .cookies import SQLITE_AVAILABLE, SECRETSTORAGE_AVAILABLE | |
3686 | ||
3687 | lib_str = join_nonempty( | |
3688 | compat_brotli and compat_brotli.__name__, | |
3689 | compat_pycrypto_AES and compat_pycrypto_AES.__name__.split('.')[0], | |
3690 | SECRETSTORAGE_AVAILABLE and 'secretstorage', | |
3691 | has_mutagen and 'mutagen', | |
3692 | SQLITE_AVAILABLE and 'sqlite', | |
3693 | has_websockets and 'websockets', | |
3694 | delim=', ') or 'none' | |
3695 | write_debug('Optional libraries: %s' % lib_str) | |
3696 | ||
3697 | proxy_map = {} | |
3698 | for handler in self._opener.handlers: | |
3699 | if hasattr(handler, 'proxies'): | |
3700 | proxy_map.update(handler.proxies) | |
3701 | write_debug(f'Proxy map: {proxy_map}') | |
3702 | ||
3703 | # Not implemented | |
3704 | if False and self.params.get('call_home'): | |
3705 | ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8') | |
3706 | write_debug('Public IP address: %s' % ipaddr) | |
3707 | latest_version = self.urlopen( | |
3708 | 'https://yt-dl.org/latest/version').read().decode('utf-8') | |
3709 | if version_tuple(latest_version) > version_tuple(__version__): | |
3710 | self.report_warning( | |
3711 | 'You are using an outdated version (newest version: %s)! ' | |
3712 | 'See https://yt-dl.org/update if you need help updating.' % | |
3713 | latest_version) | |
3714 | ||
3715 | def _setup_opener(self): | |
3716 | timeout_val = self.params.get('socket_timeout') | |
3717 | self._socket_timeout = 20 if timeout_val is None else float(timeout_val) | |
3718 | ||
3719 | opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser') | |
3720 | opts_cookiefile = self.params.get('cookiefile') | |
3721 | opts_proxy = self.params.get('proxy') | |
3722 | ||
3723 | self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self) | |
3724 | ||
3725 | cookie_processor = YoutubeDLCookieProcessor(self.cookiejar) | |
3726 | if opts_proxy is not None: | |
3727 | if opts_proxy == '': | |
3728 | proxies = {} | |
3729 | else: | |
3730 | proxies = {'http': opts_proxy, 'https': opts_proxy} | |
3731 | else: | |
3732 | proxies = compat_urllib_request.getproxies() | |
3733 | # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805) | |
3734 | if 'http' in proxies and 'https' not in proxies: | |
3735 | proxies['https'] = proxies['http'] | |
3736 | proxy_handler = PerRequestProxyHandler(proxies) | |
3737 | ||
3738 | debuglevel = 1 if self.params.get('debug_printtraffic') else 0 | |
3739 | https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel) | |
3740 | ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel) | |
3741 | redirect_handler = YoutubeDLRedirectHandler() | |
3742 | data_handler = compat_urllib_request_DataHandler() | |
3743 | ||
3744 | # When passing our own FileHandler instance, build_opener won't add the | |
3745 | # default FileHandler and allows us to disable the file protocol, which | |
3746 | # can be used for malicious purposes (see | |
3747 | # https://github.com/ytdl-org/youtube-dl/issues/8227) | |
3748 | file_handler = compat_urllib_request.FileHandler() | |
3749 | ||
3750 | def file_open(*args, **kwargs): | |
3751 | raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons') | |
3752 | file_handler.file_open = file_open | |
3753 | ||
3754 | opener = compat_urllib_request.build_opener( | |
3755 | proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler) | |
3756 | ||
3757 | # Delete the default user-agent header, which would otherwise apply in | |
3758 | # cases where our custom HTTP handler doesn't come into play | |
3759 | # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details) | |
3760 | opener.addheaders = [] | |
3761 | self._opener = opener | |
3762 | ||
3763 | def encode(self, s): | |
3764 | if isinstance(s, bytes): | |
3765 | return s # Already encoded | |
3766 | ||
3767 | try: | |
3768 | return s.encode(self.get_encoding()) | |
3769 | except UnicodeEncodeError as err: | |
3770 | err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.' | |
3771 | raise | |
3772 | ||
3773 | def get_encoding(self): | |
3774 | encoding = self.params.get('encoding') | |
3775 | if encoding is None: | |
3776 | encoding = preferredencoding() | |
3777 | return encoding | |
3778 | ||
3779 | def _write_info_json(self, label, ie_result, infofn, overwrite=None): | |
3780 | ''' Write infojson and returns True = written, False = skip, None = error ''' | |
3781 | if overwrite is None: | |
3782 | overwrite = self.params.get('overwrites', True) | |
3783 | if not self.params.get('writeinfojson'): | |
3784 | return False | |
3785 | elif not infofn: | |
3786 | self.write_debug(f'Skipping writing {label} infojson') | |
3787 | return False | |
3788 | elif not self._ensure_dir_exists(infofn): | |
3789 | return None | |
3790 | elif not overwrite and os.path.exists(infofn): | |
3791 | self.to_screen(f'[info] {label.title()} metadata is already present') | |
3792 | else: | |
3793 | self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}') | |
3794 | try: | |
3795 | write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn) | |
3796 | except (OSError, IOError): | |
3797 | self.report_error(f'Cannot write {label} metadata to JSON file {infofn}') | |
3798 | return None | |
3799 | return True | |
3800 | ||
3801 | def _write_description(self, label, ie_result, descfn): | |
3802 | ''' Write description and returns True = written, False = skip, None = error ''' | |
3803 | if not self.params.get('writedescription'): | |
3804 | return False | |
3805 | elif not descfn: | |
3806 | self.write_debug(f'Skipping writing {label} description') | |
3807 | return False | |
3808 | elif not self._ensure_dir_exists(descfn): | |
3809 | return None | |
3810 | elif not self.params.get('overwrites', True) and os.path.exists(descfn): | |
3811 | self.to_screen(f'[info] {label.title()} description is already present') | |
3812 | elif ie_result.get('description') is None: | |
3813 | self.report_warning(f'There\'s no {label} description to write') | |
3814 | return False | |
3815 | else: | |
3816 | try: | |
3817 | self.to_screen(f'[info] Writing {label} description to: {descfn}') | |
3818 | with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile: | |
3819 | descfile.write(ie_result['description']) | |
3820 | except (OSError, IOError): | |
3821 | self.report_error(f'Cannot write {label} description file {descfn}') | |
3822 | return None | |
3823 | return True | |
3824 | ||
3825 | def _write_subtitles(self, info_dict, filename): | |
3826 | ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error''' | |
3827 | ret = [] | |
3828 | subtitles = info_dict.get('requested_subtitles') | |
3829 | if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')): | |
3830 | # subtitles download errors are already managed as troubles in relevant IE | |
3831 | # that way it will silently go on when used with unsupporting IE | |
3832 | return ret | |
3833 | ||
3834 | sub_filename_base = self.prepare_filename(info_dict, 'subtitle') | |
3835 | if not sub_filename_base: | |
3836 | self.to_screen('[info] Skipping writing video subtitles') | |
3837 | return ret | |
3838 | for sub_lang, sub_info in subtitles.items(): | |
3839 | sub_format = sub_info['ext'] | |
3840 | sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext')) | |
3841 | sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext')) | |
3842 | existing_sub = self.existing_file((sub_filename_final, sub_filename)) | |
3843 | if existing_sub: | |
3844 | self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present') | |
3845 | sub_info['filepath'] = existing_sub | |
3846 | ret.append((existing_sub, sub_filename_final)) | |
3847 | continue | |
3848 | ||
3849 | self.to_screen(f'[info] Writing video subtitles to: {sub_filename}') | |
3850 | if sub_info.get('data') is not None: | |
3851 | try: | |
3852 | # Use newline='' to prevent conversion of newline characters | |
3853 | # See https://github.com/ytdl-org/youtube-dl/issues/10268 | |
3854 | with io.open(sub_filename, 'w', encoding='utf-8', newline='') as subfile: | |
3855 | subfile.write(sub_info['data']) | |
3856 | sub_info['filepath'] = sub_filename | |
3857 | ret.append((sub_filename, sub_filename_final)) | |
3858 | continue | |
3859 | except (OSError, IOError): | |
3860 | self.report_error(f'Cannot write video subtitles file {sub_filename}') | |
3861 | return None | |
3862 | ||
3863 | try: | |
3864 | sub_copy = sub_info.copy() | |
3865 | sub_copy.setdefault('http_headers', info_dict.get('http_headers')) | |
3866 | self.dl(sub_filename, sub_copy, subtitle=True) | |
3867 | sub_info['filepath'] = sub_filename | |
3868 | ret.append((sub_filename, sub_filename_final)) | |
3869 | except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err: | |
3870 | if self.params.get('ignoreerrors') is not True: # False or 'only_download' | |
3871 | raise DownloadError(f'Unable to download video subtitles for {sub_lang!r}: {err}', err) | |
3872 | self.report_warning(f'Unable to download video subtitles for {sub_lang!r}: {err}') | |
3873 | return ret | |
3874 | ||
3875 | def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None): | |
3876 | ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) ''' | |
3877 | write_all = self.params.get('write_all_thumbnails', False) | |
3878 | thumbnails, ret = [], [] | |
3879 | if write_all or self.params.get('writethumbnail', False): | |
3880 | thumbnails = info_dict.get('thumbnails') or [] | |
3881 | multiple = write_all and len(thumbnails) > 1 | |
3882 | ||
3883 | if thumb_filename_base is None: | |
3884 | thumb_filename_base = filename | |
3885 | if thumbnails and not thumb_filename_base: | |
3886 | self.write_debug(f'Skipping writing {label} thumbnail') | |
3887 | return ret | |
3888 | ||
3889 | for idx, t in list(enumerate(thumbnails))[::-1]: | |
3890 | thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg') | |
3891 | thumb_display_id = f'{label} thumbnail {t["id"]}' | |
3892 | thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext')) | |
3893 | thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext')) | |
3894 | ||
3895 | existing_thumb = self.existing_file((thumb_filename_final, thumb_filename)) | |
3896 | if existing_thumb: | |
3897 | self.to_screen('[info] %s is already present' % ( | |
3898 | thumb_display_id if multiple else f'{label} thumbnail').capitalize()) | |
3899 | t['filepath'] = existing_thumb | |
3900 | ret.append((existing_thumb, thumb_filename_final)) | |
3901 | else: | |
3902 | self.to_screen(f'[info] Downloading {thumb_display_id} ...') | |
3903 | try: | |
3904 | uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {}))) | |
3905 | self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}') | |
3906 | with open(encodeFilename(thumb_filename), 'wb') as thumbf: | |
3907 | shutil.copyfileobj(uf, thumbf) | |
3908 | ret.append((thumb_filename, thumb_filename_final)) | |
3909 | t['filepath'] = thumb_filename | |
3910 | except network_exceptions as err: | |
3911 | thumbnails.pop(idx) | |
3912 | self.report_warning(f'Unable to download {thumb_display_id}: {err}') | |
3913 | if ret and not write_all: | |
3914 | break | |
3915 | return ret |