]>
Commit | Line | Data |
---|---|---|
1 | import collections | |
2 | import contextlib | |
3 | import datetime | |
4 | import errno | |
5 | import fileinput | |
6 | import functools | |
7 | import io | |
8 | import itertools | |
9 | import json | |
10 | import locale | |
11 | import operator | |
12 | import os | |
13 | import random | |
14 | import re | |
15 | import shutil | |
16 | import subprocess | |
17 | import sys | |
18 | import tempfile | |
19 | import time | |
20 | import tokenize | |
21 | import traceback | |
22 | import unicodedata | |
23 | import urllib.request | |
24 | from string import ascii_letters | |
25 | ||
26 | from .cache import Cache | |
27 | from .compat import compat_os_name, compat_shlex_quote | |
28 | from .cookies import load_cookies | |
29 | from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name | |
30 | from .downloader.rtmp import rtmpdump_version | |
31 | from .extractor import gen_extractor_classes, get_info_extractor | |
32 | from .extractor.common import UnsupportedURLIE | |
33 | from .extractor.openload import PhantomJSwrapper | |
34 | from .minicurses import format_text | |
35 | from .plugins import directories as plugin_directories | |
36 | from .postprocessor import _PLUGIN_CLASSES as plugin_pps | |
37 | from .postprocessor import ( | |
38 | EmbedThumbnailPP, | |
39 | FFmpegFixupDuplicateMoovPP, | |
40 | FFmpegFixupDurationPP, | |
41 | FFmpegFixupM3u8PP, | |
42 | FFmpegFixupM4aPP, | |
43 | FFmpegFixupStretchedPP, | |
44 | FFmpegFixupTimestampPP, | |
45 | FFmpegMergerPP, | |
46 | FFmpegPostProcessor, | |
47 | FFmpegVideoConvertorPP, | |
48 | MoveFilesAfterDownloadPP, | |
49 | get_postprocessor, | |
50 | ) | |
51 | from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping | |
52 | from .update import REPOSITORY, current_git_head, detect_variant | |
53 | from .utils import ( | |
54 | DEFAULT_OUTTMPL, | |
55 | IDENTITY, | |
56 | LINK_TEMPLATES, | |
57 | MEDIA_EXTENSIONS, | |
58 | NO_DEFAULT, | |
59 | NUMBER_RE, | |
60 | OUTTMPL_TYPES, | |
61 | POSTPROCESS_WHEN, | |
62 | STR_FORMAT_RE_TMPL, | |
63 | STR_FORMAT_TYPES, | |
64 | ContentTooShortError, | |
65 | DateRange, | |
66 | DownloadCancelled, | |
67 | DownloadError, | |
68 | EntryNotInPlaylist, | |
69 | ExistingVideoReached, | |
70 | ExtractorError, | |
71 | FormatSorter, | |
72 | GeoRestrictedError, | |
73 | HEADRequest, | |
74 | ISO3166Utils, | |
75 | LazyList, | |
76 | MaxDownloadsReached, | |
77 | Namespace, | |
78 | PagedList, | |
79 | PerRequestProxyHandler, | |
80 | PlaylistEntries, | |
81 | Popen, | |
82 | PostProcessingError, | |
83 | ReExtractInfo, | |
84 | RejectedVideoReached, | |
85 | SameFileError, | |
86 | UnavailableVideoError, | |
87 | UserNotLive, | |
88 | YoutubeDLCookieProcessor, | |
89 | YoutubeDLHandler, | |
90 | YoutubeDLRedirectHandler, | |
91 | age_restricted, | |
92 | args_to_str, | |
93 | bug_reports_message, | |
94 | date_from_str, | |
95 | deprecation_warning, | |
96 | determine_ext, | |
97 | determine_protocol, | |
98 | encode_compat_str, | |
99 | encodeFilename, | |
100 | error_to_compat_str, | |
101 | escapeHTML, | |
102 | expand_path, | |
103 | filter_dict, | |
104 | float_or_none, | |
105 | format_bytes, | |
106 | format_decimal_suffix, | |
107 | format_field, | |
108 | formatSeconds, | |
109 | get_compatible_ext, | |
110 | get_domain, | |
111 | int_or_none, | |
112 | iri_to_uri, | |
113 | is_path_like, | |
114 | join_nonempty, | |
115 | locked_file, | |
116 | make_archive_id, | |
117 | make_dir, | |
118 | make_HTTPS_handler, | |
119 | merge_headers, | |
120 | network_exceptions, | |
121 | number_of_digits, | |
122 | orderedSet, | |
123 | orderedSet_from_options, | |
124 | parse_filesize, | |
125 | preferredencoding, | |
126 | prepend_extension, | |
127 | register_socks_protocols, | |
128 | remove_terminal_sequences, | |
129 | render_table, | |
130 | replace_extension, | |
131 | sanitize_filename, | |
132 | sanitize_path, | |
133 | sanitize_url, | |
134 | sanitized_Request, | |
135 | std_headers, | |
136 | str_or_none, | |
137 | strftime_or_none, | |
138 | subtitles_filename, | |
139 | supports_terminal_sequences, | |
140 | system_identifier, | |
141 | timetuple_from_msec, | |
142 | to_high_limit_path, | |
143 | traverse_obj, | |
144 | try_call, | |
145 | try_get, | |
146 | url_basename, | |
147 | variadic, | |
148 | version_tuple, | |
149 | windows_enable_vt_mode, | |
150 | write_json_file, | |
151 | write_string, | |
152 | ) | |
153 | from .version import RELEASE_GIT_HEAD, VARIANT, __version__ | |
154 | ||
155 | if compat_os_name == 'nt': | |
156 | import ctypes | |
157 | ||
158 | ||
159 | class YoutubeDL: | |
160 | """YoutubeDL class. | |
161 | ||
162 | YoutubeDL objects are the ones responsible of downloading the | |
163 | actual video file and writing it to disk if the user has requested | |
164 | it, among some other tasks. In most cases there should be one per | |
165 | program. As, given a video URL, the downloader doesn't know how to | |
166 | extract all the needed information, task that InfoExtractors do, it | |
167 | has to pass the URL to one of them. | |
168 | ||
169 | For this, YoutubeDL objects have a method that allows | |
170 | InfoExtractors to be registered in a given order. When it is passed | |
171 | a URL, the YoutubeDL object handles it to the first InfoExtractor it | |
172 | finds that reports being able to handle it. The InfoExtractor extracts | |
173 | all the information about the video or videos the URL refers to, and | |
174 | YoutubeDL process the extracted information, possibly using a File | |
175 | Downloader to download the video. | |
176 | ||
177 | YoutubeDL objects accept a lot of parameters. In order not to saturate | |
178 | the object constructor with arguments, it receives a dictionary of | |
179 | options instead. These options are available through the params | |
180 | attribute for the InfoExtractors to use. The YoutubeDL also | |
181 | registers itself as the downloader in charge for the InfoExtractors | |
182 | that are added to it, so this is a "mutual registration". | |
183 | ||
184 | Available options: | |
185 | ||
186 | username: Username for authentication purposes. | |
187 | password: Password for authentication purposes. | |
188 | videopassword: Password for accessing a video. | |
189 | ap_mso: Adobe Pass multiple-system operator identifier. | |
190 | ap_username: Multiple-system operator account username. | |
191 | ap_password: Multiple-system operator account password. | |
192 | usenetrc: Use netrc for authentication instead. | |
193 | verbose: Print additional info to stdout. | |
194 | quiet: Do not print messages to stdout. | |
195 | no_warnings: Do not print out anything for warnings. | |
196 | forceprint: A dict with keys WHEN mapped to a list of templates to | |
197 | print to stdout. The allowed keys are video or any of the | |
198 | items in utils.POSTPROCESS_WHEN. | |
199 | For compatibility, a single list is also accepted | |
200 | print_to_file: A dict with keys WHEN (same as forceprint) mapped to | |
201 | a list of tuples with (template, filename) | |
202 | forcejson: Force printing info_dict as JSON. | |
203 | dump_single_json: Force printing the info_dict of the whole playlist | |
204 | (or video) as a single JSON line. | |
205 | force_write_download_archive: Force writing download archive regardless | |
206 | of 'skip_download' or 'simulate'. | |
207 | simulate: Do not download the video files. If unset (or None), | |
208 | simulate only if listsubtitles, listformats or list_thumbnails is used | |
209 | format: Video format code. see "FORMAT SELECTION" for more details. | |
210 | You can also pass a function. The function takes 'ctx' as | |
211 | argument and returns the formats to download. | |
212 | See "build_format_selector" for an implementation | |
213 | allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded. | |
214 | ignore_no_formats_error: Ignore "No video formats" error. Usefull for | |
215 | extracting metadata even if the video is not actually | |
216 | available for download (experimental) | |
217 | format_sort: A list of fields by which to sort the video formats. | |
218 | See "Sorting Formats" for more details. | |
219 | format_sort_force: Force the given format_sort. see "Sorting Formats" | |
220 | for more details. | |
221 | prefer_free_formats: Whether to prefer video formats with free containers | |
222 | over non-free ones of same quality. | |
223 | allow_multiple_video_streams: Allow multiple video streams to be merged | |
224 | into a single file | |
225 | allow_multiple_audio_streams: Allow multiple audio streams to be merged | |
226 | into a single file | |
227 | check_formats Whether to test if the formats are downloadable. | |
228 | Can be True (check all), False (check none), | |
229 | 'selected' (check selected formats), | |
230 | or None (check only if requested by extractor) | |
231 | paths: Dictionary of output paths. The allowed keys are 'home' | |
232 | 'temp' and the keys of OUTTMPL_TYPES (in utils.py) | |
233 | outtmpl: Dictionary of templates for output names. Allowed keys | |
234 | are 'default' and the keys of OUTTMPL_TYPES (in utils.py). | |
235 | For compatibility with youtube-dl, a single string can also be used | |
236 | outtmpl_na_placeholder: Placeholder for unavailable meta fields. | |
237 | restrictfilenames: Do not allow "&" and spaces in file names | |
238 | trim_file_name: Limit length of filename (extension excluded) | |
239 | windowsfilenames: Force the filenames to be windows compatible | |
240 | ignoreerrors: Do not stop on download/postprocessing errors. | |
241 | Can be 'only_download' to ignore only download errors. | |
242 | Default is 'only_download' for CLI, but False for API | |
243 | skip_playlist_after_errors: Number of allowed failures until the rest of | |
244 | the playlist is skipped | |
245 | allowed_extractors: List of regexes to match against extractor names that are allowed | |
246 | overwrites: Overwrite all video and metadata files if True, | |
247 | overwrite only non-video files if None | |
248 | and don't overwrite any file if False | |
249 | For compatibility with youtube-dl, | |
250 | "nooverwrites" may also be used instead | |
251 | playlist_items: Specific indices of playlist to download. | |
252 | playlistrandom: Download playlist items in random order. | |
253 | lazy_playlist: Process playlist entries as they are received. | |
254 | matchtitle: Download only matching titles. | |
255 | rejecttitle: Reject downloads for matching titles. | |
256 | logger: Log messages to a logging.Logger instance. | |
257 | logtostderr: Print everything to stderr instead of stdout. | |
258 | consoletitle: Display progress in console window's titlebar. | |
259 | writedescription: Write the video description to a .description file | |
260 | writeinfojson: Write the video description to a .info.json file | |
261 | clean_infojson: Remove private fields from the infojson | |
262 | getcomments: Extract video comments. This will not be written to disk | |
263 | unless writeinfojson is also given | |
264 | writeannotations: Write the video annotations to a .annotations.xml file | |
265 | writethumbnail: Write the thumbnail image to a file | |
266 | allow_playlist_files: Whether to write playlists' description, infojson etc | |
267 | also to disk when using the 'write*' options | |
268 | write_all_thumbnails: Write all thumbnail formats to files | |
269 | writelink: Write an internet shortcut file, depending on the | |
270 | current platform (.url/.webloc/.desktop) | |
271 | writeurllink: Write a Windows internet shortcut file (.url) | |
272 | writewebloclink: Write a macOS internet shortcut file (.webloc) | |
273 | writedesktoplink: Write a Linux internet shortcut file (.desktop) | |
274 | writesubtitles: Write the video subtitles to a file | |
275 | writeautomaticsub: Write the automatically generated subtitles to a file | |
276 | listsubtitles: Lists all available subtitles for the video | |
277 | subtitlesformat: The format code for subtitles | |
278 | subtitleslangs: List of languages of the subtitles to download (can be regex). | |
279 | The list may contain "all" to refer to all the available | |
280 | subtitles. The language can be prefixed with a "-" to | |
281 | exclude it from the requested languages, e.g. ['all', '-live_chat'] | |
282 | keepvideo: Keep the video file after post-processing | |
283 | daterange: A DateRange object, download only if the upload_date is in the range. | |
284 | skip_download: Skip the actual download of the video file | |
285 | cachedir: Location of the cache files in the filesystem. | |
286 | False to disable filesystem cache. | |
287 | noplaylist: Download single video instead of a playlist if in doubt. | |
288 | age_limit: An integer representing the user's age in years. | |
289 | Unsuitable videos for the given age are skipped. | |
290 | min_views: An integer representing the minimum view count the video | |
291 | must have in order to not be skipped. | |
292 | Videos without view count information are always | |
293 | downloaded. None for no limit. | |
294 | max_views: An integer representing the maximum view count. | |
295 | Videos that are more popular than that are not | |
296 | downloaded. | |
297 | Videos without view count information are always | |
298 | downloaded. None for no limit. | |
299 | download_archive: A set, or the name of a file where all downloads are recorded. | |
300 | Videos already present in the file are not downloaded again. | |
301 | break_on_existing: Stop the download process after attempting to download a | |
302 | file that is in the archive. | |
303 | break_on_reject: Stop the download process when encountering a video that | |
304 | has been filtered out. | |
305 | break_per_url: Whether break_on_reject and break_on_existing | |
306 | should act on each input URL as opposed to for the entire queue | |
307 | cookiefile: File name or text stream from where cookies should be read and dumped to | |
308 | cookiesfrombrowser: A tuple containing the name of the browser, the profile | |
309 | name/path from where cookies are loaded, the name of the keyring, | |
310 | and the container name, e.g. ('chrome', ) or | |
311 | ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta') | |
312 | legacyserverconnect: Explicitly allow HTTPS connection to servers that do not | |
313 | support RFC 5746 secure renegotiation | |
314 | nocheckcertificate: Do not verify SSL certificates | |
315 | client_certificate: Path to client certificate file in PEM format. May include the private key | |
316 | client_certificate_key: Path to private key file for client certificate | |
317 | client_certificate_password: Password for client certificate private key, if encrypted. | |
318 | If not provided and the key is encrypted, yt-dlp will ask interactively | |
319 | prefer_insecure: Use HTTP instead of HTTPS to retrieve information. | |
320 | (Only supported by some extractors) | |
321 | enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons. | |
322 | http_headers: A dictionary of custom headers to be used for all requests | |
323 | proxy: URL of the proxy server to use | |
324 | geo_verification_proxy: URL of the proxy to use for IP address verification | |
325 | on geo-restricted sites. | |
326 | socket_timeout: Time to wait for unresponsive hosts, in seconds | |
327 | bidi_workaround: Work around buggy terminals without bidirectional text | |
328 | support, using fridibi | |
329 | debug_printtraffic:Print out sent and received HTTP traffic | |
330 | default_search: Prepend this string if an input url is not valid. | |
331 | 'auto' for elaborate guessing | |
332 | encoding: Use this encoding instead of the system-specified. | |
333 | extract_flat: Whether to resolve and process url_results further | |
334 | * False: Always process (default) | |
335 | * True: Never process | |
336 | * 'in_playlist': Do not process inside playlist/multi_video | |
337 | * 'discard': Always process, but don't return the result | |
338 | from inside playlist/multi_video | |
339 | * 'discard_in_playlist': Same as "discard", but only for | |
340 | playlists (not multi_video) | |
341 | wait_for_video: If given, wait for scheduled streams to become available. | |
342 | The value should be a tuple containing the range | |
343 | (min_secs, max_secs) to wait between retries | |
344 | postprocessors: A list of dictionaries, each with an entry | |
345 | * key: The name of the postprocessor. See | |
346 | yt_dlp/postprocessor/__init__.py for a list. | |
347 | * when: When to run the postprocessor. Allowed values are | |
348 | the entries of utils.POSTPROCESS_WHEN | |
349 | Assumed to be 'post_process' if not given | |
350 | progress_hooks: A list of functions that get called on download | |
351 | progress, with a dictionary with the entries | |
352 | * status: One of "downloading", "error", or "finished". | |
353 | Check this first and ignore unknown values. | |
354 | * info_dict: The extracted info_dict | |
355 | ||
356 | If status is one of "downloading", or "finished", the | |
357 | following properties may also be present: | |
358 | * filename: The final filename (always present) | |
359 | * tmpfilename: The filename we're currently writing to | |
360 | * downloaded_bytes: Bytes on disk | |
361 | * total_bytes: Size of the whole file, None if unknown | |
362 | * total_bytes_estimate: Guess of the eventual file size, | |
363 | None if unavailable. | |
364 | * elapsed: The number of seconds since download started. | |
365 | * eta: The estimated time in seconds, None if unknown | |
366 | * speed: The download speed in bytes/second, None if | |
367 | unknown | |
368 | * fragment_index: The counter of the currently | |
369 | downloaded video fragment. | |
370 | * fragment_count: The number of fragments (= individual | |
371 | files that will be merged) | |
372 | ||
373 | Progress hooks are guaranteed to be called at least once | |
374 | (with status "finished") if the download is successful. | |
375 | postprocessor_hooks: A list of functions that get called on postprocessing | |
376 | progress, with a dictionary with the entries | |
377 | * status: One of "started", "processing", or "finished". | |
378 | Check this first and ignore unknown values. | |
379 | * postprocessor: Name of the postprocessor | |
380 | * info_dict: The extracted info_dict | |
381 | ||
382 | Progress hooks are guaranteed to be called at least twice | |
383 | (with status "started" and "finished") if the processing is successful. | |
384 | merge_output_format: "/" separated list of extensions to use when merging formats. | |
385 | final_ext: Expected final extension; used to detect when the file was | |
386 | already downloaded and converted | |
387 | fixup: Automatically correct known faults of the file. | |
388 | One of: | |
389 | - "never": do nothing | |
390 | - "warn": only emit a warning | |
391 | - "detect_or_warn": check whether we can do anything | |
392 | about it, warn otherwise (default) | |
393 | source_address: Client-side IP address to bind to. | |
394 | sleep_interval_requests: Number of seconds to sleep between requests | |
395 | during extraction | |
396 | sleep_interval: Number of seconds to sleep before each download when | |
397 | used alone or a lower bound of a range for randomized | |
398 | sleep before each download (minimum possible number | |
399 | of seconds to sleep) when used along with | |
400 | max_sleep_interval. | |
401 | max_sleep_interval:Upper bound of a range for randomized sleep before each | |
402 | download (maximum possible number of seconds to sleep). | |
403 | Must only be used along with sleep_interval. | |
404 | Actual sleep time will be a random float from range | |
405 | [sleep_interval; max_sleep_interval]. | |
406 | sleep_interval_subtitles: Number of seconds to sleep before each subtitle download | |
407 | listformats: Print an overview of available video formats and exit. | |
408 | list_thumbnails: Print a table of all thumbnails and exit. | |
409 | match_filter: A function that gets called for every video with the signature | |
410 | (info_dict, *, incomplete: bool) -> Optional[str] | |
411 | For backward compatibility with youtube-dl, the signature | |
412 | (info_dict) -> Optional[str] is also allowed. | |
413 | - If it returns a message, the video is ignored. | |
414 | - If it returns None, the video is downloaded. | |
415 | - If it returns utils.NO_DEFAULT, the user is interactively | |
416 | asked whether to download the video. | |
417 | match_filter_func in utils.py is one example for this. | |
418 | no_color: Do not emit color codes in output. | |
419 | geo_bypass: Bypass geographic restriction via faking X-Forwarded-For | |
420 | HTTP header | |
421 | geo_bypass_country: | |
422 | Two-letter ISO 3166-2 country code that will be used for | |
423 | explicit geographic restriction bypassing via faking | |
424 | X-Forwarded-For HTTP header | |
425 | geo_bypass_ip_block: | |
426 | IP range in CIDR notation that will be used similarly to | |
427 | geo_bypass_country | |
428 | external_downloader: A dictionary of protocol keys and the executable of the | |
429 | external downloader to use for it. The allowed protocols | |
430 | are default|http|ftp|m3u8|dash|rtsp|rtmp|mms. | |
431 | Set the value to 'native' to use the native downloader | |
432 | compat_opts: Compatibility options. See "Differences in default behavior". | |
433 | The following options do not work when used through the API: | |
434 | filename, abort-on-error, multistreams, no-live-chat, format-sort | |
435 | no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json. | |
436 | Refer __init__.py for their implementation | |
437 | progress_template: Dictionary of templates for progress outputs. | |
438 | Allowed keys are 'download', 'postprocess', | |
439 | 'download-title' (console title) and 'postprocess-title'. | |
440 | The template is mapped on a dictionary with keys 'progress' and 'info' | |
441 | retry_sleep_functions: Dictionary of functions that takes the number of attempts | |
442 | as argument and returns the time to sleep in seconds. | |
443 | Allowed keys are 'http', 'fragment', 'file_access' | |
444 | download_ranges: A callback function that gets called for every video with | |
445 | the signature (info_dict, ydl) -> Iterable[Section]. | |
446 | Only the returned sections will be downloaded. | |
447 | Each Section is a dict with the following keys: | |
448 | * start_time: Start time of the section in seconds | |
449 | * end_time: End time of the section in seconds | |
450 | * title: Section title (Optional) | |
451 | * index: Section number (Optional) | |
452 | force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts | |
453 | noprogress: Do not print the progress bar | |
454 | live_from_start: Whether to download livestreams videos from the start | |
455 | ||
456 | The following parameters are not used by YoutubeDL itself, they are used by | |
457 | the downloader (see yt_dlp/downloader/common.py): | |
458 | nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize, | |
459 | max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries, | |
460 | continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size, | |
461 | external_downloader_args, concurrent_fragment_downloads. | |
462 | ||
463 | The following options are used by the post processors: | |
464 | ffmpeg_location: Location of the ffmpeg/avconv binary; either the path | |
465 | to the binary or its containing directory. | |
466 | postprocessor_args: A dictionary of postprocessor/executable keys (in lower case) | |
467 | and a list of additional command-line arguments for the | |
468 | postprocessor/executable. The dict can also have "PP+EXE" keys | |
469 | which are used when the given exe is used by the given PP. | |
470 | Use 'default' as the name for arguments to passed to all PP | |
471 | For compatibility with youtube-dl, a single list of args | |
472 | can also be used | |
473 | ||
474 | The following options are used by the extractors: | |
475 | extractor_retries: Number of times to retry for known errors | |
476 | dynamic_mpd: Whether to process dynamic DASH manifests (default: True) | |
477 | hls_split_discontinuity: Split HLS playlists to different formats at | |
478 | discontinuities such as ad breaks (default: False) | |
479 | extractor_args: A dictionary of arguments to be passed to the extractors. | |
480 | See "EXTRACTOR ARGUMENTS" for details. | |
481 | E.g. {'youtube': {'skip': ['dash', 'hls']}} | |
482 | mark_watched: Mark videos watched (even with --simulate). Only for YouTube | |
483 | ||
484 | The following options are deprecated and may be removed in the future: | |
485 | ||
486 | force_generic_extractor: Force downloader to use the generic extractor | |
487 | - Use allowed_extractors = ['generic', 'default'] | |
488 | playliststart: - Use playlist_items | |
489 | Playlist item to start at. | |
490 | playlistend: - Use playlist_items | |
491 | Playlist item to end at. | |
492 | playlistreverse: - Use playlist_items | |
493 | Download playlist items in reverse order. | |
494 | forceurl: - Use forceprint | |
495 | Force printing final URL. | |
496 | forcetitle: - Use forceprint | |
497 | Force printing title. | |
498 | forceid: - Use forceprint | |
499 | Force printing ID. | |
500 | forcethumbnail: - Use forceprint | |
501 | Force printing thumbnail URL. | |
502 | forcedescription: - Use forceprint | |
503 | Force printing description. | |
504 | forcefilename: - Use forceprint | |
505 | Force printing final filename. | |
506 | forceduration: - Use forceprint | |
507 | Force printing duration. | |
508 | allsubtitles: - Use subtitleslangs = ['all'] | |
509 | Downloads all the subtitles of the video | |
510 | (requires writesubtitles or writeautomaticsub) | |
511 | include_ads: - Doesn't work | |
512 | Download ads as well | |
513 | call_home: - Not implemented | |
514 | Boolean, true iff we are allowed to contact the | |
515 | yt-dlp servers for debugging. | |
516 | post_hooks: - Register a custom postprocessor | |
517 | A list of functions that get called as the final step | |
518 | for each video file, after all postprocessors have been | |
519 | called. The filename will be passed as the only argument. | |
520 | hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}. | |
521 | Use the native HLS downloader instead of ffmpeg/avconv | |
522 | if True, otherwise use ffmpeg/avconv if False, otherwise | |
523 | use downloader suggested by extractor if None. | |
524 | prefer_ffmpeg: - avconv support is deprecated | |
525 | If False, use avconv instead of ffmpeg if both are available, | |
526 | otherwise prefer ffmpeg. | |
527 | youtube_include_dash_manifest: - Use extractor_args | |
528 | If True (default), DASH manifests and related | |
529 | data will be downloaded and processed by extractor. | |
530 | You can reduce network I/O by disabling it if you don't | |
531 | care about DASH. (only for youtube) | |
532 | youtube_include_hls_manifest: - Use extractor_args | |
533 | If True (default), HLS manifests and related | |
534 | data will be downloaded and processed by extractor. | |
535 | You can reduce network I/O by disabling it if you don't | |
536 | care about HLS. (only for youtube) | |
537 | """ | |
538 | ||
539 | _NUMERIC_FIELDS = { | |
540 | 'width', 'height', 'asr', 'audio_channels', 'fps', | |
541 | 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx', | |
542 | 'timestamp', 'release_timestamp', | |
543 | 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count', | |
544 | 'average_rating', 'comment_count', 'age_limit', | |
545 | 'start_time', 'end_time', | |
546 | 'chapter_number', 'season_number', 'episode_number', | |
547 | 'track_number', 'disc_number', 'release_year', | |
548 | } | |
549 | ||
550 | _format_fields = { | |
551 | # NB: Keep in sync with the docstring of extractor/common.py | |
552 | 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note', | |
553 | 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels', | |
554 | 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns', | |
555 | 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start', | |
556 | 'preference', 'language', 'language_preference', 'quality', 'source_preference', | |
557 | 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options', | |
558 | 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time' | |
559 | } | |
560 | _format_selection_exts = { | |
561 | 'audio': set(MEDIA_EXTENSIONS.common_audio), | |
562 | 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )), | |
563 | 'storyboards': set(MEDIA_EXTENSIONS.storyboards), | |
564 | } | |
565 | ||
566 | def __init__(self, params=None, auto_init=True): | |
567 | """Create a FileDownloader object with the given options. | |
568 | @param auto_init Whether to load the default extractors and print header (if verbose). | |
569 | Set to 'no_verbose_header' to not print the header | |
570 | """ | |
571 | if params is None: | |
572 | params = {} | |
573 | self.params = params | |
574 | self._ies = {} | |
575 | self._ies_instances = {} | |
576 | self._pps = {k: [] for k in POSTPROCESS_WHEN} | |
577 | self._printed_messages = set() | |
578 | self._first_webpage_request = True | |
579 | self._post_hooks = [] | |
580 | self._progress_hooks = [] | |
581 | self._postprocessor_hooks = [] | |
582 | self._download_retcode = 0 | |
583 | self._num_downloads = 0 | |
584 | self._num_videos = 0 | |
585 | self._playlist_level = 0 | |
586 | self._playlist_urls = set() | |
587 | self.cache = Cache(self) | |
588 | ||
589 | stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout | |
590 | self._out_files = Namespace( | |
591 | out=stdout, | |
592 | error=sys.stderr, | |
593 | screen=sys.stderr if self.params.get('quiet') else stdout, | |
594 | console=None if compat_os_name == 'nt' else next( | |
595 | filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None) | |
596 | ) | |
597 | ||
598 | try: | |
599 | windows_enable_vt_mode() | |
600 | except Exception as e: | |
601 | self.write_debug(f'Failed to enable VT mode: {e}') | |
602 | ||
603 | self._allow_colors = Namespace(**{ | |
604 | type_: not self.params.get('no_color') and supports_terminal_sequences(stream) | |
605 | for type_, stream in self._out_files.items_ if type_ != 'console' | |
606 | }) | |
607 | ||
608 | # The code is left like this to be reused for future deprecations | |
609 | MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7) | |
610 | current_version = sys.version_info[:2] | |
611 | if current_version < MIN_RECOMMENDED: | |
612 | msg = ('Support for Python version %d.%d has been deprecated. ' | |
613 | 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.' | |
614 | '\n You will no longer receive updates on this version') | |
615 | if current_version < MIN_SUPPORTED: | |
616 | msg = 'Python version %d.%d is no longer supported' | |
617 | self.deprecation_warning( | |
618 | f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED)) | |
619 | ||
620 | if self.params.get('allow_unplayable_formats'): | |
621 | self.report_warning( | |
622 | f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. ' | |
623 | 'This is a developer option intended for debugging. \n' | |
624 | ' If you experience any issues while using this option, ' | |
625 | f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report') | |
626 | ||
627 | if self.params.get('bidi_workaround', False): | |
628 | try: | |
629 | import pty | |
630 | master, slave = pty.openpty() | |
631 | width = shutil.get_terminal_size().columns | |
632 | width_args = [] if width is None else ['-w', str(width)] | |
633 | sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error} | |
634 | try: | |
635 | self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs) | |
636 | except OSError: | |
637 | self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs) | |
638 | self._output_channel = os.fdopen(master, 'rb') | |
639 | except OSError as ose: | |
640 | if ose.errno == errno.ENOENT: | |
641 | self.report_warning( | |
642 | 'Could not find fribidi executable, ignoring --bidi-workaround. ' | |
643 | 'Make sure that fribidi is an executable file in one of the directories in your $PATH.') | |
644 | else: | |
645 | raise | |
646 | ||
647 | self.params['compat_opts'] = set(self.params.get('compat_opts', ())) | |
648 | if auto_init and auto_init != 'no_verbose_header': | |
649 | self.print_debug_header() | |
650 | ||
651 | def check_deprecated(param, option, suggestion): | |
652 | if self.params.get(param) is not None: | |
653 | self.report_warning(f'{option} is deprecated. Use {suggestion} instead') | |
654 | return True | |
655 | return False | |
656 | ||
657 | if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'): | |
658 | if self.params.get('geo_verification_proxy') is None: | |
659 | self.params['geo_verification_proxy'] = self.params['cn_verification_proxy'] | |
660 | ||
661 | check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"') | |
662 | check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"') | |
663 | check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"') | |
664 | ||
665 | for msg in self.params.get('_warnings', []): | |
666 | self.report_warning(msg) | |
667 | for msg in self.params.get('_deprecation_warnings', []): | |
668 | self.deprecated_feature(msg) | |
669 | ||
670 | if 'list-formats' in self.params['compat_opts']: | |
671 | self.params['listformats_table'] = False | |
672 | ||
673 | if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None: | |
674 | # nooverwrites was unnecessarily changed to overwrites | |
675 | # in 0c3d0f51778b153f65c21906031c2e091fcfb641 | |
676 | # This ensures compatibility with both keys | |
677 | self.params['overwrites'] = not self.params['nooverwrites'] | |
678 | elif self.params.get('overwrites') is None: | |
679 | self.params.pop('overwrites', None) | |
680 | else: | |
681 | self.params['nooverwrites'] = not self.params['overwrites'] | |
682 | ||
683 | if self.params.get('simulate') is None and any(( | |
684 | self.params.get('list_thumbnails'), | |
685 | self.params.get('listformats'), | |
686 | self.params.get('listsubtitles'), | |
687 | )): | |
688 | self.params['simulate'] = 'list_only' | |
689 | ||
690 | self.params.setdefault('forceprint', {}) | |
691 | self.params.setdefault('print_to_file', {}) | |
692 | ||
693 | # Compatibility with older syntax | |
694 | if not isinstance(params['forceprint'], dict): | |
695 | self.params['forceprint'] = {'video': params['forceprint']} | |
696 | ||
697 | if auto_init: | |
698 | self.add_default_info_extractors() | |
699 | ||
700 | if (sys.platform != 'win32' | |
701 | and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] | |
702 | and not self.params.get('restrictfilenames', False)): | |
703 | # Unicode filesystem API will throw errors (#1474, #13027) | |
704 | self.report_warning( | |
705 | 'Assuming --restrict-filenames since file system encoding ' | |
706 | 'cannot encode all characters. ' | |
707 | 'Set the LC_ALL environment variable to fix this.') | |
708 | self.params['restrictfilenames'] = True | |
709 | ||
710 | self._parse_outtmpl() | |
711 | ||
712 | # Creating format selector here allows us to catch syntax errors before the extraction | |
713 | self.format_selector = ( | |
714 | self.params.get('format') if self.params.get('format') in (None, '-') | |
715 | else self.params['format'] if callable(self.params['format']) | |
716 | else self.build_format_selector(self.params['format'])) | |
717 | ||
718 | # Set http_headers defaults according to std_headers | |
719 | self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {})) | |
720 | ||
721 | hooks = { | |
722 | 'post_hooks': self.add_post_hook, | |
723 | 'progress_hooks': self.add_progress_hook, | |
724 | 'postprocessor_hooks': self.add_postprocessor_hook, | |
725 | } | |
726 | for opt, fn in hooks.items(): | |
727 | for ph in self.params.get(opt, []): | |
728 | fn(ph) | |
729 | ||
730 | for pp_def_raw in self.params.get('postprocessors', []): | |
731 | pp_def = dict(pp_def_raw) | |
732 | when = pp_def.pop('when', 'post_process') | |
733 | self.add_post_processor( | |
734 | get_postprocessor(pp_def.pop('key'))(self, **pp_def), | |
735 | when=when) | |
736 | ||
737 | self._setup_opener() | |
738 | register_socks_protocols() | |
739 | ||
740 | def preload_download_archive(fn): | |
741 | """Preload the archive, if any is specified""" | |
742 | archive = set() | |
743 | if fn is None: | |
744 | return archive | |
745 | elif not is_path_like(fn): | |
746 | return fn | |
747 | ||
748 | self.write_debug(f'Loading archive file {fn!r}') | |
749 | try: | |
750 | with locked_file(fn, 'r', encoding='utf-8') as archive_file: | |
751 | for line in archive_file: | |
752 | archive.add(line.strip()) | |
753 | except OSError as ioe: | |
754 | if ioe.errno != errno.ENOENT: | |
755 | raise | |
756 | return archive | |
757 | ||
758 | self.archive = preload_download_archive(self.params.get('download_archive')) | |
759 | ||
760 | def warn_if_short_id(self, argv): | |
761 | # short YouTube ID starting with dash? | |
762 | idxs = [ | |
763 | i for i, a in enumerate(argv) | |
764 | if re.match(r'^-[0-9A-Za-z_-]{10}$', a)] | |
765 | if idxs: | |
766 | correct_argv = ( | |
767 | ['yt-dlp'] | |
768 | + [a for i, a in enumerate(argv) if i not in idxs] | |
769 | + ['--'] + [argv[i] for i in idxs] | |
770 | ) | |
771 | self.report_warning( | |
772 | 'Long argument string detected. ' | |
773 | 'Use -- to separate parameters and URLs, like this:\n%s' % | |
774 | args_to_str(correct_argv)) | |
775 | ||
776 | def add_info_extractor(self, ie): | |
777 | """Add an InfoExtractor object to the end of the list.""" | |
778 | ie_key = ie.ie_key() | |
779 | self._ies[ie_key] = ie | |
780 | if not isinstance(ie, type): | |
781 | self._ies_instances[ie_key] = ie | |
782 | ie.set_downloader(self) | |
783 | ||
784 | def get_info_extractor(self, ie_key): | |
785 | """ | |
786 | Get an instance of an IE with name ie_key, it will try to get one from | |
787 | the _ies list, if there's no instance it will create a new one and add | |
788 | it to the extractor list. | |
789 | """ | |
790 | ie = self._ies_instances.get(ie_key) | |
791 | if ie is None: | |
792 | ie = get_info_extractor(ie_key)() | |
793 | self.add_info_extractor(ie) | |
794 | return ie | |
795 | ||
796 | def add_default_info_extractors(self): | |
797 | """ | |
798 | Add the InfoExtractors returned by gen_extractors to the end of the list | |
799 | """ | |
800 | all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()} | |
801 | all_ies['end'] = UnsupportedURLIE() | |
802 | try: | |
803 | ie_names = orderedSet_from_options( | |
804 | self.params.get('allowed_extractors', ['default']), { | |
805 | 'all': list(all_ies), | |
806 | 'default': [name for name, ie in all_ies.items() if ie._ENABLED], | |
807 | }, use_regex=True) | |
808 | except re.error as e: | |
809 | raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}') | |
810 | for name in ie_names: | |
811 | self.add_info_extractor(all_ies[name]) | |
812 | self.write_debug(f'Loaded {len(ie_names)} extractors') | |
813 | ||
814 | def add_post_processor(self, pp, when='post_process'): | |
815 | """Add a PostProcessor object to the end of the chain.""" | |
816 | assert when in POSTPROCESS_WHEN, f'Invalid when={when}' | |
817 | self._pps[when].append(pp) | |
818 | pp.set_downloader(self) | |
819 | ||
820 | def add_post_hook(self, ph): | |
821 | """Add the post hook""" | |
822 | self._post_hooks.append(ph) | |
823 | ||
824 | def add_progress_hook(self, ph): | |
825 | """Add the download progress hook""" | |
826 | self._progress_hooks.append(ph) | |
827 | ||
828 | def add_postprocessor_hook(self, ph): | |
829 | """Add the postprocessing progress hook""" | |
830 | self._postprocessor_hooks.append(ph) | |
831 | for pps in self._pps.values(): | |
832 | for pp in pps: | |
833 | pp.add_progress_hook(ph) | |
834 | ||
835 | def _bidi_workaround(self, message): | |
836 | if not hasattr(self, '_output_channel'): | |
837 | return message | |
838 | ||
839 | assert hasattr(self, '_output_process') | |
840 | assert isinstance(message, str) | |
841 | line_count = message.count('\n') + 1 | |
842 | self._output_process.stdin.write((message + '\n').encode()) | |
843 | self._output_process.stdin.flush() | |
844 | res = ''.join(self._output_channel.readline().decode() | |
845 | for _ in range(line_count)) | |
846 | return res[:-len('\n')] | |
847 | ||
848 | def _write_string(self, message, out=None, only_once=False): | |
849 | if only_once: | |
850 | if message in self._printed_messages: | |
851 | return | |
852 | self._printed_messages.add(message) | |
853 | write_string(message, out=out, encoding=self.params.get('encoding')) | |
854 | ||
855 | def to_stdout(self, message, skip_eol=False, quiet=None): | |
856 | """Print message to stdout""" | |
857 | if quiet is not None: | |
858 | self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. ' | |
859 | 'Use "YoutubeDL.to_screen" instead') | |
860 | if skip_eol is not False: | |
861 | self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. ' | |
862 | 'Use "YoutubeDL.to_screen" instead') | |
863 | self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out) | |
864 | ||
865 | def to_screen(self, message, skip_eol=False, quiet=None, only_once=False): | |
866 | """Print message to screen if not in quiet mode""" | |
867 | if self.params.get('logger'): | |
868 | self.params['logger'].debug(message) | |
869 | return | |
870 | if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'): | |
871 | return | |
872 | self._write_string( | |
873 | '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')), | |
874 | self._out_files.screen, only_once=only_once) | |
875 | ||
876 | def to_stderr(self, message, only_once=False): | |
877 | """Print message to stderr""" | |
878 | assert isinstance(message, str) | |
879 | if self.params.get('logger'): | |
880 | self.params['logger'].error(message) | |
881 | else: | |
882 | self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once) | |
883 | ||
884 | def _send_console_code(self, code): | |
885 | if compat_os_name == 'nt' or not self._out_files.console: | |
886 | return | |
887 | self._write_string(code, self._out_files.console) | |
888 | ||
889 | def to_console_title(self, message): | |
890 | if not self.params.get('consoletitle', False): | |
891 | return | |
892 | message = remove_terminal_sequences(message) | |
893 | if compat_os_name == 'nt': | |
894 | if ctypes.windll.kernel32.GetConsoleWindow(): | |
895 | # c_wchar_p() might not be necessary if `message` is | |
896 | # already of type unicode() | |
897 | ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message)) | |
898 | else: | |
899 | self._send_console_code(f'\033]0;{message}\007') | |
900 | ||
901 | def save_console_title(self): | |
902 | if not self.params.get('consoletitle') or self.params.get('simulate'): | |
903 | return | |
904 | self._send_console_code('\033[22;0t') # Save the title on stack | |
905 | ||
906 | def restore_console_title(self): | |
907 | if not self.params.get('consoletitle') or self.params.get('simulate'): | |
908 | return | |
909 | self._send_console_code('\033[23;0t') # Restore the title from stack | |
910 | ||
911 | def __enter__(self): | |
912 | self.save_console_title() | |
913 | return self | |
914 | ||
915 | def __exit__(self, *args): | |
916 | self.restore_console_title() | |
917 | ||
918 | if self.params.get('cookiefile') is not None: | |
919 | self.cookiejar.save(ignore_discard=True, ignore_expires=True) | |
920 | ||
921 | def trouble(self, message=None, tb=None, is_error=True): | |
922 | """Determine action to take when a download problem appears. | |
923 | ||
924 | Depending on if the downloader has been configured to ignore | |
925 | download errors or not, this method may throw an exception or | |
926 | not when errors are found, after printing the message. | |
927 | ||
928 | @param tb If given, is additional traceback information | |
929 | @param is_error Whether to raise error according to ignorerrors | |
930 | """ | |
931 | if message is not None: | |
932 | self.to_stderr(message) | |
933 | if self.params.get('verbose'): | |
934 | if tb is None: | |
935 | if sys.exc_info()[0]: # if .trouble has been called from an except block | |
936 | tb = '' | |
937 | if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]: | |
938 | tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info)) | |
939 | tb += encode_compat_str(traceback.format_exc()) | |
940 | else: | |
941 | tb_data = traceback.format_list(traceback.extract_stack()) | |
942 | tb = ''.join(tb_data) | |
943 | if tb: | |
944 | self.to_stderr(tb) | |
945 | if not is_error: | |
946 | return | |
947 | if not self.params.get('ignoreerrors'): | |
948 | if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]: | |
949 | exc_info = sys.exc_info()[1].exc_info | |
950 | else: | |
951 | exc_info = sys.exc_info() | |
952 | raise DownloadError(message, exc_info) | |
953 | self._download_retcode = 1 | |
954 | ||
955 | Styles = Namespace( | |
956 | HEADERS='yellow', | |
957 | EMPHASIS='light blue', | |
958 | FILENAME='green', | |
959 | ID='green', | |
960 | DELIM='blue', | |
961 | ERROR='red', | |
962 | WARNING='yellow', | |
963 | SUPPRESS='light black', | |
964 | ) | |
965 | ||
966 | def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False): | |
967 | text = str(text) | |
968 | if test_encoding: | |
969 | original_text = text | |
970 | # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711 | |
971 | encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii' | |
972 | text = text.encode(encoding, 'ignore').decode(encoding) | |
973 | if fallback is not None and text != original_text: | |
974 | text = fallback | |
975 | return format_text(text, f) if allow_colors else text if fallback is None else fallback | |
976 | ||
977 | def _format_out(self, *args, **kwargs): | |
978 | return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs) | |
979 | ||
980 | def _format_screen(self, *args, **kwargs): | |
981 | return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs) | |
982 | ||
983 | def _format_err(self, *args, **kwargs): | |
984 | return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs) | |
985 | ||
986 | def report_warning(self, message, only_once=False): | |
987 | ''' | |
988 | Print the message to stderr, it will be prefixed with 'WARNING:' | |
989 | If stderr is a tty file the 'WARNING:' will be colored | |
990 | ''' | |
991 | if self.params.get('logger') is not None: | |
992 | self.params['logger'].warning(message) | |
993 | else: | |
994 | if self.params.get('no_warnings'): | |
995 | return | |
996 | self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once) | |
997 | ||
998 | def deprecation_warning(self, message, *, stacklevel=0): | |
999 | deprecation_warning( | |
1000 | message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False) | |
1001 | ||
1002 | def deprecated_feature(self, message): | |
1003 | if self.params.get('logger') is not None: | |
1004 | self.params['logger'].warning(f'Deprecated Feature: {message}') | |
1005 | self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True) | |
1006 | ||
1007 | def report_error(self, message, *args, **kwargs): | |
1008 | ''' | |
1009 | Do the same as trouble, but prefixes the message with 'ERROR:', colored | |
1010 | in red if stderr is a tty file. | |
1011 | ''' | |
1012 | self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs) | |
1013 | ||
1014 | def write_debug(self, message, only_once=False): | |
1015 | '''Log debug message or Print message to stderr''' | |
1016 | if not self.params.get('verbose', False): | |
1017 | return | |
1018 | message = f'[debug] {message}' | |
1019 | if self.params.get('logger'): | |
1020 | self.params['logger'].debug(message) | |
1021 | else: | |
1022 | self.to_stderr(message, only_once) | |
1023 | ||
1024 | def report_file_already_downloaded(self, file_name): | |
1025 | """Report file has already been fully downloaded.""" | |
1026 | try: | |
1027 | self.to_screen('[download] %s has already been downloaded' % file_name) | |
1028 | except UnicodeEncodeError: | |
1029 | self.to_screen('[download] The file has already been downloaded') | |
1030 | ||
1031 | def report_file_delete(self, file_name): | |
1032 | """Report that existing file will be deleted.""" | |
1033 | try: | |
1034 | self.to_screen('Deleting existing file %s' % file_name) | |
1035 | except UnicodeEncodeError: | |
1036 | self.to_screen('Deleting existing file') | |
1037 | ||
1038 | def raise_no_formats(self, info, forced=False, *, msg=None): | |
1039 | has_drm = info.get('_has_drm') | |
1040 | ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg) | |
1041 | msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!' | |
1042 | if forced or not ignored: | |
1043 | raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'], | |
1044 | expected=has_drm or ignored or expected) | |
1045 | else: | |
1046 | self.report_warning(msg) | |
1047 | ||
1048 | def parse_outtmpl(self): | |
1049 | self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version') | |
1050 | self._parse_outtmpl() | |
1051 | return self.params['outtmpl'] | |
1052 | ||
1053 | def _parse_outtmpl(self): | |
1054 | sanitize = IDENTITY | |
1055 | if self.params.get('restrictfilenames'): # Remove spaces in the default template | |
1056 | sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-') | |
1057 | ||
1058 | outtmpl = self.params.setdefault('outtmpl', {}) | |
1059 | if not isinstance(outtmpl, dict): | |
1060 | self.params['outtmpl'] = outtmpl = {'default': outtmpl} | |
1061 | outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None}) | |
1062 | ||
1063 | def get_output_path(self, dir_type='', filename=None): | |
1064 | paths = self.params.get('paths', {}) | |
1065 | assert isinstance(paths, dict), '"paths" parameter must be a dictionary' | |
1066 | path = os.path.join( | |
1067 | expand_path(paths.get('home', '').strip()), | |
1068 | expand_path(paths.get(dir_type, '').strip()) if dir_type else '', | |
1069 | filename or '') | |
1070 | return sanitize_path(path, force=self.params.get('windowsfilenames')) | |
1071 | ||
1072 | @staticmethod | |
1073 | def _outtmpl_expandpath(outtmpl): | |
1074 | # expand_path translates '%%' into '%' and '$$' into '$' | |
1075 | # correspondingly that is not what we want since we need to keep | |
1076 | # '%%' intact for template dict substitution step. Working around | |
1077 | # with boundary-alike separator hack. | |
1078 | sep = ''.join(random.choices(ascii_letters, k=32)) | |
1079 | outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$') | |
1080 | ||
1081 | # outtmpl should be expand_path'ed before template dict substitution | |
1082 | # because meta fields may contain env variables we don't want to | |
1083 | # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and | |
1084 | # title "Hello $PATH", we don't want `$PATH` to be expanded. | |
1085 | return expand_path(outtmpl).replace(sep, '') | |
1086 | ||
1087 | @staticmethod | |
1088 | def escape_outtmpl(outtmpl): | |
1089 | ''' Escape any remaining strings like %s, %abc% etc. ''' | |
1090 | return re.sub( | |
1091 | STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'), | |
1092 | lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0), | |
1093 | outtmpl) | |
1094 | ||
1095 | @classmethod | |
1096 | def validate_outtmpl(cls, outtmpl): | |
1097 | ''' @return None or Exception object ''' | |
1098 | outtmpl = re.sub( | |
1099 | STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'), | |
1100 | lambda mobj: f'{mobj.group(0)[:-1]}s', | |
1101 | cls._outtmpl_expandpath(outtmpl)) | |
1102 | try: | |
1103 | cls.escape_outtmpl(outtmpl) % collections.defaultdict(int) | |
1104 | return None | |
1105 | except ValueError as err: | |
1106 | return err | |
1107 | ||
1108 | @staticmethod | |
1109 | def _copy_infodict(info_dict): | |
1110 | info_dict = dict(info_dict) | |
1111 | info_dict.pop('__postprocessors', None) | |
1112 | info_dict.pop('__pending_error', None) | |
1113 | return info_dict | |
1114 | ||
1115 | def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False): | |
1116 | """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict | |
1117 | @param sanitize Whether to sanitize the output as a filename. | |
1118 | For backward compatibility, a function can also be passed | |
1119 | """ | |
1120 | ||
1121 | info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set | |
1122 | ||
1123 | info_dict = self._copy_infodict(info_dict) | |
1124 | info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs | |
1125 | formatSeconds(info_dict['duration'], '-' if sanitize else ':') | |
1126 | if info_dict.get('duration', None) is not None | |
1127 | else None) | |
1128 | info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads) | |
1129 | info_dict['video_autonumber'] = self._num_videos | |
1130 | if info_dict.get('resolution') is None: | |
1131 | info_dict['resolution'] = self.format_resolution(info_dict, default=None) | |
1132 | ||
1133 | # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences | |
1134 | # of %(field)s to %(field)0Nd for backward compatibility | |
1135 | field_size_compat_map = { | |
1136 | 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0), | |
1137 | 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0), | |
1138 | 'autonumber': self.params.get('autonumber_size') or 5, | |
1139 | } | |
1140 | ||
1141 | TMPL_DICT = {} | |
1142 | EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]')) | |
1143 | MATH_FUNCTIONS = { | |
1144 | '+': float.__add__, | |
1145 | '-': float.__sub__, | |
1146 | } | |
1147 | # Field is of the form key1.key2... | |
1148 | # where keys (except first) can be string, int, slice or "{field, ...}" | |
1149 | FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'} | |
1150 | FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % { | |
1151 | 'inner': FIELD_INNER_RE, | |
1152 | 'field': rf'\w*(?:\.{FIELD_INNER_RE})*' | |
1153 | } | |
1154 | MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})' | |
1155 | MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys())) | |
1156 | INTERNAL_FORMAT_RE = re.compile(rf'''(?x) | |
1157 | (?P<negate>-)? | |
1158 | (?P<fields>{FIELD_RE}) | |
1159 | (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*) | |
1160 | (?:>(?P<strf_format>.+?))? | |
1161 | (?P<remaining> | |
1162 | (?P<alternate>(?<!\\),[^|&)]+)? | |
1163 | (?:&(?P<replacement>.*?))? | |
1164 | (?:\|(?P<default>.*?))? | |
1165 | )$''') | |
1166 | ||
1167 | def _traverse_infodict(fields): | |
1168 | fields = [f for x in re.split(r'\.({.+?})\.?', fields) | |
1169 | for f in ([x] if x.startswith('{') else x.split('.'))] | |
1170 | for i in (0, -1): | |
1171 | if fields and not fields[i]: | |
1172 | fields.pop(i) | |
1173 | ||
1174 | for i, f in enumerate(fields): | |
1175 | if not f.startswith('{'): | |
1176 | continue | |
1177 | assert f.endswith('}'), f'No closing brace for {f} in {fields}' | |
1178 | fields[i] = {k: k.split('.') for k in f[1:-1].split(',')} | |
1179 | ||
1180 | return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True) | |
1181 | ||
1182 | def get_value(mdict): | |
1183 | # Object traversal | |
1184 | value = _traverse_infodict(mdict['fields']) | |
1185 | # Negative | |
1186 | if mdict['negate']: | |
1187 | value = float_or_none(value) | |
1188 | if value is not None: | |
1189 | value *= -1 | |
1190 | # Do maths | |
1191 | offset_key = mdict['maths'] | |
1192 | if offset_key: | |
1193 | value = float_or_none(value) | |
1194 | operator = None | |
1195 | while offset_key: | |
1196 | item = re.match( | |
1197 | MATH_FIELD_RE if operator else MATH_OPERATORS_RE, | |
1198 | offset_key).group(0) | |
1199 | offset_key = offset_key[len(item):] | |
1200 | if operator is None: | |
1201 | operator = MATH_FUNCTIONS[item] | |
1202 | continue | |
1203 | item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1) | |
1204 | offset = float_or_none(item) | |
1205 | if offset is None: | |
1206 | offset = float_or_none(_traverse_infodict(item)) | |
1207 | try: | |
1208 | value = operator(value, multiplier * offset) | |
1209 | except (TypeError, ZeroDivisionError): | |
1210 | return None | |
1211 | operator = None | |
1212 | # Datetime formatting | |
1213 | if mdict['strf_format']: | |
1214 | value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ',')) | |
1215 | ||
1216 | # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485 | |
1217 | if sanitize and value == '': | |
1218 | value = None | |
1219 | return value | |
1220 | ||
1221 | na = self.params.get('outtmpl_na_placeholder', 'NA') | |
1222 | ||
1223 | def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')): | |
1224 | return sanitize_filename(str(value), restricted=restricted, is_id=( | |
1225 | bool(re.search(r'(^|[_.])id(\.|$)', key)) | |
1226 | if 'filename-sanitization' in self.params['compat_opts'] | |
1227 | else NO_DEFAULT)) | |
1228 | ||
1229 | sanitizer = sanitize if callable(sanitize) else filename_sanitizer | |
1230 | sanitize = bool(sanitize) | |
1231 | ||
1232 | def _dumpjson_default(obj): | |
1233 | if isinstance(obj, (set, LazyList)): | |
1234 | return list(obj) | |
1235 | return repr(obj) | |
1236 | ||
1237 | def create_key(outer_mobj): | |
1238 | if not outer_mobj.group('has_key'): | |
1239 | return outer_mobj.group(0) | |
1240 | key = outer_mobj.group('key') | |
1241 | mobj = re.match(INTERNAL_FORMAT_RE, key) | |
1242 | initial_field = mobj.group('fields') if mobj else '' | |
1243 | value, replacement, default = None, None, na | |
1244 | while mobj: | |
1245 | mobj = mobj.groupdict() | |
1246 | default = mobj['default'] if mobj['default'] is not None else default | |
1247 | value = get_value(mobj) | |
1248 | replacement = mobj['replacement'] | |
1249 | if value is None and mobj['alternate']: | |
1250 | mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:]) | |
1251 | else: | |
1252 | break | |
1253 | ||
1254 | fmt = outer_mobj.group('format') | |
1255 | if fmt == 's' and value is not None and key in field_size_compat_map.keys(): | |
1256 | fmt = f'0{field_size_compat_map[key]:d}d' | |
1257 | ||
1258 | value = default if value is None else value if replacement is None else replacement | |
1259 | ||
1260 | flags = outer_mobj.group('conversion') or '' | |
1261 | str_fmt = f'{fmt[:-1]}s' | |
1262 | if fmt[-1] == 'l': # list | |
1263 | delim = '\n' if '#' in flags else ', ' | |
1264 | value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt | |
1265 | elif fmt[-1] == 'j': # json | |
1266 | value, fmt = json.dumps( | |
1267 | value, default=_dumpjson_default, | |
1268 | indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt | |
1269 | elif fmt[-1] == 'h': # html | |
1270 | value, fmt = escapeHTML(str(value)), str_fmt | |
1271 | elif fmt[-1] == 'q': # quoted | |
1272 | value = map(str, variadic(value) if '#' in flags else [value]) | |
1273 | value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt | |
1274 | elif fmt[-1] == 'B': # bytes | |
1275 | value = f'%{str_fmt}'.encode() % str(value).encode() | |
1276 | value, fmt = value.decode('utf-8', 'ignore'), 's' | |
1277 | elif fmt[-1] == 'U': # unicode normalized | |
1278 | value, fmt = unicodedata.normalize( | |
1279 | # "+" = compatibility equivalence, "#" = NFD | |
1280 | 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'), | |
1281 | value), str_fmt | |
1282 | elif fmt[-1] == 'D': # decimal suffix | |
1283 | num_fmt, fmt = fmt[:-1].replace('#', ''), 's' | |
1284 | value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s', | |
1285 | factor=1024 if '#' in flags else 1000) | |
1286 | elif fmt[-1] == 'S': # filename sanitization | |
1287 | value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt | |
1288 | elif fmt[-1] == 'c': | |
1289 | if value: | |
1290 | value = str(value)[0] | |
1291 | else: | |
1292 | fmt = str_fmt | |
1293 | elif fmt[-1] not in 'rs': # numeric | |
1294 | value = float_or_none(value) | |
1295 | if value is None: | |
1296 | value, fmt = default, 's' | |
1297 | ||
1298 | if sanitize: | |
1299 | if fmt[-1] == 'r': | |
1300 | # If value is an object, sanitize might convert it to a string | |
1301 | # So we convert it to repr first | |
1302 | value, fmt = repr(value), str_fmt | |
1303 | if fmt[-1] in 'csr': | |
1304 | value = sanitizer(initial_field, value) | |
1305 | ||
1306 | key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format')) | |
1307 | TMPL_DICT[key] = value | |
1308 | return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix')) | |
1309 | ||
1310 | return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT | |
1311 | ||
1312 | def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs): | |
1313 | outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs) | |
1314 | return self.escape_outtmpl(outtmpl) % info_dict | |
1315 | ||
1316 | def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None): | |
1317 | assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive' | |
1318 | if outtmpl is None: | |
1319 | outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default']) | |
1320 | try: | |
1321 | outtmpl = self._outtmpl_expandpath(outtmpl) | |
1322 | filename = self.evaluate_outtmpl(outtmpl, info_dict, True) | |
1323 | if not filename: | |
1324 | return None | |
1325 | ||
1326 | if tmpl_type in ('', 'temp'): | |
1327 | final_ext, ext = self.params.get('final_ext'), info_dict.get('ext') | |
1328 | if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'): | |
1329 | filename = replace_extension(filename, ext, final_ext) | |
1330 | elif tmpl_type: | |
1331 | force_ext = OUTTMPL_TYPES[tmpl_type] | |
1332 | if force_ext: | |
1333 | filename = replace_extension(filename, force_ext, info_dict.get('ext')) | |
1334 | ||
1335 | # https://github.com/blackjack4494/youtube-dlc/issues/85 | |
1336 | trim_file_name = self.params.get('trim_file_name', False) | |
1337 | if trim_file_name: | |
1338 | no_ext, *ext = filename.rsplit('.', 2) | |
1339 | filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.') | |
1340 | ||
1341 | return filename | |
1342 | except ValueError as err: | |
1343 | self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')') | |
1344 | return None | |
1345 | ||
1346 | def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False): | |
1347 | """Generate the output filename""" | |
1348 | if outtmpl: | |
1349 | assert not dir_type, 'outtmpl and dir_type are mutually exclusive' | |
1350 | dir_type = None | |
1351 | filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl) | |
1352 | if not filename and dir_type not in ('', 'temp'): | |
1353 | return '' | |
1354 | ||
1355 | if warn: | |
1356 | if not self.params.get('paths'): | |
1357 | pass | |
1358 | elif filename == '-': | |
1359 | self.report_warning('--paths is ignored when an outputting to stdout', only_once=True) | |
1360 | elif os.path.isabs(filename): | |
1361 | self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True) | |
1362 | if filename == '-' or not filename: | |
1363 | return filename | |
1364 | ||
1365 | return self.get_output_path(dir_type, filename) | |
1366 | ||
1367 | def _match_entry(self, info_dict, incomplete=False, silent=False): | |
1368 | """Returns None if the file should be downloaded""" | |
1369 | _type = info_dict.get('_type', 'video') | |
1370 | assert incomplete or _type == 'video', 'Only video result can be considered complete' | |
1371 | ||
1372 | video_title = info_dict.get('title', info_dict.get('id', 'entry')) | |
1373 | ||
1374 | def check_filter(): | |
1375 | if _type in ('playlist', 'multi_video'): | |
1376 | return | |
1377 | elif _type in ('url', 'url_transparent') and not try_call( | |
1378 | lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])): | |
1379 | return | |
1380 | ||
1381 | if 'title' in info_dict: | |
1382 | # This can happen when we're just evaluating the playlist | |
1383 | title = info_dict['title'] | |
1384 | matchtitle = self.params.get('matchtitle', False) | |
1385 | if matchtitle: | |
1386 | if not re.search(matchtitle, title, re.IGNORECASE): | |
1387 | return '"' + title + '" title did not match pattern "' + matchtitle + '"' | |
1388 | rejecttitle = self.params.get('rejecttitle', False) | |
1389 | if rejecttitle: | |
1390 | if re.search(rejecttitle, title, re.IGNORECASE): | |
1391 | return '"' + title + '" title matched reject pattern "' + rejecttitle + '"' | |
1392 | ||
1393 | date = info_dict.get('upload_date') | |
1394 | if date is not None: | |
1395 | dateRange = self.params.get('daterange', DateRange()) | |
1396 | if date not in dateRange: | |
1397 | return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}' | |
1398 | view_count = info_dict.get('view_count') | |
1399 | if view_count is not None: | |
1400 | min_views = self.params.get('min_views') | |
1401 | if min_views is not None and view_count < min_views: | |
1402 | return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views) | |
1403 | max_views = self.params.get('max_views') | |
1404 | if max_views is not None and view_count > max_views: | |
1405 | return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views) | |
1406 | if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')): | |
1407 | return 'Skipping "%s" because it is age restricted' % video_title | |
1408 | ||
1409 | match_filter = self.params.get('match_filter') | |
1410 | if match_filter is not None: | |
1411 | try: | |
1412 | ret = match_filter(info_dict, incomplete=incomplete) | |
1413 | except TypeError: | |
1414 | # For backward compatibility | |
1415 | ret = None if incomplete else match_filter(info_dict) | |
1416 | if ret is NO_DEFAULT: | |
1417 | while True: | |
1418 | filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME) | |
1419 | reply = input(self._format_screen( | |
1420 | f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip() | |
1421 | if reply in {'y', ''}: | |
1422 | return None | |
1423 | elif reply == 'n': | |
1424 | return f'Skipping {video_title}' | |
1425 | elif ret is not None: | |
1426 | return ret | |
1427 | return None | |
1428 | ||
1429 | if self.in_download_archive(info_dict): | |
1430 | reason = '%s has already been recorded in the archive' % video_title | |
1431 | break_opt, break_err = 'break_on_existing', ExistingVideoReached | |
1432 | else: | |
1433 | reason = check_filter() | |
1434 | break_opt, break_err = 'break_on_reject', RejectedVideoReached | |
1435 | if reason is not None: | |
1436 | if not silent: | |
1437 | self.to_screen('[download] ' + reason) | |
1438 | if self.params.get(break_opt, False): | |
1439 | raise break_err() | |
1440 | return reason | |
1441 | ||
1442 | @staticmethod | |
1443 | def add_extra_info(info_dict, extra_info): | |
1444 | '''Set the keys from extra_info in info dict if they are missing''' | |
1445 | for key, value in extra_info.items(): | |
1446 | info_dict.setdefault(key, value) | |
1447 | ||
1448 | def extract_info(self, url, download=True, ie_key=None, extra_info=None, | |
1449 | process=True, force_generic_extractor=False): | |
1450 | """ | |
1451 | Extract and return the information dictionary of the URL | |
1452 | ||
1453 | Arguments: | |
1454 | @param url URL to extract | |
1455 | ||
1456 | Keyword arguments: | |
1457 | @param download Whether to download videos | |
1458 | @param process Whether to resolve all unresolved references (URLs, playlist items). | |
1459 | Must be True for download to work | |
1460 | @param ie_key Use only the extractor with this key | |
1461 | ||
1462 | @param extra_info Dictionary containing the extra values to add to the info (For internal use only) | |
1463 | @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic') | |
1464 | """ | |
1465 | ||
1466 | if extra_info is None: | |
1467 | extra_info = {} | |
1468 | ||
1469 | if not ie_key and force_generic_extractor: | |
1470 | ie_key = 'Generic' | |
1471 | ||
1472 | if ie_key: | |
1473 | ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {} | |
1474 | else: | |
1475 | ies = self._ies | |
1476 | ||
1477 | for key, ie in ies.items(): | |
1478 | if not ie.suitable(url): | |
1479 | continue | |
1480 | ||
1481 | if not ie.working(): | |
1482 | self.report_warning('The program functionality for this site has been marked as broken, ' | |
1483 | 'and will probably not work.') | |
1484 | ||
1485 | temp_id = ie.get_temp_id(url) | |
1486 | if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}): | |
1487 | self.to_screen(f'[{key}] {temp_id}: has already been recorded in the archive') | |
1488 | if self.params.get('break_on_existing', False): | |
1489 | raise ExistingVideoReached() | |
1490 | break | |
1491 | return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process) | |
1492 | else: | |
1493 | extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default']) | |
1494 | self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}', | |
1495 | tb=False if extractors_restricted else None) | |
1496 | ||
1497 | def _handle_extraction_exceptions(func): | |
1498 | @functools.wraps(func) | |
1499 | def wrapper(self, *args, **kwargs): | |
1500 | while True: | |
1501 | try: | |
1502 | return func(self, *args, **kwargs) | |
1503 | except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError): | |
1504 | raise | |
1505 | except ReExtractInfo as e: | |
1506 | if e.expected: | |
1507 | self.to_screen(f'{e}; Re-extracting data') | |
1508 | else: | |
1509 | self.to_stderr('\r') | |
1510 | self.report_warning(f'{e}; Re-extracting data') | |
1511 | continue | |
1512 | except GeoRestrictedError as e: | |
1513 | msg = e.msg | |
1514 | if e.countries: | |
1515 | msg += '\nThis video is available in %s.' % ', '.join( | |
1516 | map(ISO3166Utils.short2full, e.countries)) | |
1517 | msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.' | |
1518 | self.report_error(msg) | |
1519 | except ExtractorError as e: # An error we somewhat expected | |
1520 | self.report_error(str(e), e.format_traceback()) | |
1521 | except Exception as e: | |
1522 | if self.params.get('ignoreerrors'): | |
1523 | self.report_error(str(e), tb=encode_compat_str(traceback.format_exc())) | |
1524 | else: | |
1525 | raise | |
1526 | break | |
1527 | return wrapper | |
1528 | ||
1529 | def _wait_for_video(self, ie_result={}): | |
1530 | if (not self.params.get('wait_for_video') | |
1531 | or ie_result.get('_type', 'video') != 'video' | |
1532 | or ie_result.get('formats') or ie_result.get('url')): | |
1533 | return | |
1534 | ||
1535 | format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1] | |
1536 | last_msg = '' | |
1537 | ||
1538 | def progress(msg): | |
1539 | nonlocal last_msg | |
1540 | full_msg = f'{msg}\n' | |
1541 | if not self.params.get('noprogress'): | |
1542 | full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r' | |
1543 | elif last_msg: | |
1544 | return | |
1545 | self.to_screen(full_msg, skip_eol=True) | |
1546 | last_msg = msg | |
1547 | ||
1548 | min_wait, max_wait = self.params.get('wait_for_video') | |
1549 | diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time()) | |
1550 | if diff is None and ie_result.get('live_status') == 'is_upcoming': | |
1551 | diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0) | |
1552 | self.report_warning('Release time of video is not known') | |
1553 | elif ie_result and (diff or 0) <= 0: | |
1554 | self.report_warning('Video should already be available according to extracted info') | |
1555 | diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf')) | |
1556 | self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now') | |
1557 | ||
1558 | wait_till = time.time() + diff | |
1559 | try: | |
1560 | while True: | |
1561 | diff = wait_till - time.time() | |
1562 | if diff <= 0: | |
1563 | progress('') | |
1564 | raise ReExtractInfo('[wait] Wait period ended', expected=True) | |
1565 | progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}') | |
1566 | time.sleep(1) | |
1567 | except KeyboardInterrupt: | |
1568 | progress('') | |
1569 | raise ReExtractInfo('[wait] Interrupted by user', expected=True) | |
1570 | except BaseException as e: | |
1571 | if not isinstance(e, ReExtractInfo): | |
1572 | self.to_screen('') | |
1573 | raise | |
1574 | ||
1575 | @_handle_extraction_exceptions | |
1576 | def __extract_info(self, url, ie, download, extra_info, process): | |
1577 | try: | |
1578 | ie_result = ie.extract(url) | |
1579 | except UserNotLive as e: | |
1580 | if process: | |
1581 | if self.params.get('wait_for_video'): | |
1582 | self.report_warning(e) | |
1583 | self._wait_for_video() | |
1584 | raise | |
1585 | if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here) | |
1586 | self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}') | |
1587 | return | |
1588 | if isinstance(ie_result, list): | |
1589 | # Backwards compatibility: old IE result format | |
1590 | ie_result = { | |
1591 | '_type': 'compat_list', | |
1592 | 'entries': ie_result, | |
1593 | } | |
1594 | if extra_info.get('original_url'): | |
1595 | ie_result.setdefault('original_url', extra_info['original_url']) | |
1596 | self.add_default_extra_info(ie_result, ie, url) | |
1597 | if process: | |
1598 | self._wait_for_video(ie_result) | |
1599 | return self.process_ie_result(ie_result, download, extra_info) | |
1600 | else: | |
1601 | return ie_result | |
1602 | ||
1603 | def add_default_extra_info(self, ie_result, ie, url): | |
1604 | if url is not None: | |
1605 | self.add_extra_info(ie_result, { | |
1606 | 'webpage_url': url, | |
1607 | 'original_url': url, | |
1608 | }) | |
1609 | webpage_url = ie_result.get('webpage_url') | |
1610 | if webpage_url: | |
1611 | self.add_extra_info(ie_result, { | |
1612 | 'webpage_url_basename': url_basename(webpage_url), | |
1613 | 'webpage_url_domain': get_domain(webpage_url), | |
1614 | }) | |
1615 | if ie is not None: | |
1616 | self.add_extra_info(ie_result, { | |
1617 | 'extractor': ie.IE_NAME, | |
1618 | 'extractor_key': ie.ie_key(), | |
1619 | }) | |
1620 | ||
1621 | def process_ie_result(self, ie_result, download=True, extra_info=None): | |
1622 | """ | |
1623 | Take the result of the ie(may be modified) and resolve all unresolved | |
1624 | references (URLs, playlist items). | |
1625 | ||
1626 | It will also download the videos if 'download'. | |
1627 | Returns the resolved ie_result. | |
1628 | """ | |
1629 | if extra_info is None: | |
1630 | extra_info = {} | |
1631 | result_type = ie_result.get('_type', 'video') | |
1632 | ||
1633 | if result_type in ('url', 'url_transparent'): | |
1634 | ie_result['url'] = sanitize_url( | |
1635 | ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https') | |
1636 | if ie_result.get('original_url') and not extra_info.get('original_url'): | |
1637 | extra_info = {'original_url': ie_result['original_url'], **extra_info} | |
1638 | ||
1639 | extract_flat = self.params.get('extract_flat', False) | |
1640 | if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) | |
1641 | or extract_flat is True): | |
1642 | info_copy = ie_result.copy() | |
1643 | ie = try_get(ie_result.get('ie_key'), self.get_info_extractor) | |
1644 | if ie and not ie_result.get('id'): | |
1645 | info_copy['id'] = ie.get_temp_id(ie_result['url']) | |
1646 | self.add_default_extra_info(info_copy, ie, ie_result['url']) | |
1647 | self.add_extra_info(info_copy, extra_info) | |
1648 | info_copy, _ = self.pre_process(info_copy) | |
1649 | self._fill_common_fields(info_copy, False) | |
1650 | self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True) | |
1651 | self._raise_pending_errors(info_copy) | |
1652 | if self.params.get('force_write_download_archive', False): | |
1653 | self.record_download_archive(info_copy) | |
1654 | return ie_result | |
1655 | ||
1656 | if result_type == 'video': | |
1657 | self.add_extra_info(ie_result, extra_info) | |
1658 | ie_result = self.process_video_result(ie_result, download=download) | |
1659 | self._raise_pending_errors(ie_result) | |
1660 | additional_urls = (ie_result or {}).get('additional_urls') | |
1661 | if additional_urls: | |
1662 | # TODO: Improve MetadataParserPP to allow setting a list | |
1663 | if isinstance(additional_urls, str): | |
1664 | additional_urls = [additional_urls] | |
1665 | self.to_screen( | |
1666 | '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls))) | |
1667 | self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls)) | |
1668 | ie_result['additional_entries'] = [ | |
1669 | self.extract_info( | |
1670 | url, download, extra_info=extra_info, | |
1671 | force_generic_extractor=self.params.get('force_generic_extractor')) | |
1672 | for url in additional_urls | |
1673 | ] | |
1674 | return ie_result | |
1675 | elif result_type == 'url': | |
1676 | # We have to add extra_info to the results because it may be | |
1677 | # contained in a playlist | |
1678 | return self.extract_info( | |
1679 | ie_result['url'], download, | |
1680 | ie_key=ie_result.get('ie_key'), | |
1681 | extra_info=extra_info) | |
1682 | elif result_type == 'url_transparent': | |
1683 | # Use the information from the embedding page | |
1684 | info = self.extract_info( | |
1685 | ie_result['url'], ie_key=ie_result.get('ie_key'), | |
1686 | extra_info=extra_info, download=False, process=False) | |
1687 | ||
1688 | # extract_info may return None when ignoreerrors is enabled and | |
1689 | # extraction failed with an error, don't crash and return early | |
1690 | # in this case | |
1691 | if not info: | |
1692 | return info | |
1693 | ||
1694 | exempted_fields = {'_type', 'url', 'ie_key'} | |
1695 | if not ie_result.get('section_end') and ie_result.get('section_start') is None: | |
1696 | # For video clips, the id etc of the clip extractor should be used | |
1697 | exempted_fields |= {'id', 'extractor', 'extractor_key'} | |
1698 | ||
1699 | new_result = info.copy() | |
1700 | new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields)) | |
1701 | ||
1702 | # Extracted info may not be a video result (i.e. | |
1703 | # info.get('_type', 'video') != video) but rather an url or | |
1704 | # url_transparent. In such cases outer metadata (from ie_result) | |
1705 | # should be propagated to inner one (info). For this to happen | |
1706 | # _type of info should be overridden with url_transparent. This | |
1707 | # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163. | |
1708 | if new_result.get('_type') == 'url': | |
1709 | new_result['_type'] = 'url_transparent' | |
1710 | ||
1711 | return self.process_ie_result( | |
1712 | new_result, download=download, extra_info=extra_info) | |
1713 | elif result_type in ('playlist', 'multi_video'): | |
1714 | # Protect from infinite recursion due to recursively nested playlists | |
1715 | # (see https://github.com/ytdl-org/youtube-dl/issues/27833) | |
1716 | webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url | |
1717 | if webpage_url and webpage_url in self._playlist_urls: | |
1718 | self.to_screen( | |
1719 | '[download] Skipping already downloaded playlist: %s' | |
1720 | % ie_result.get('title') or ie_result.get('id')) | |
1721 | return | |
1722 | ||
1723 | self._playlist_level += 1 | |
1724 | self._playlist_urls.add(webpage_url) | |
1725 | self._fill_common_fields(ie_result, False) | |
1726 | self._sanitize_thumbnails(ie_result) | |
1727 | try: | |
1728 | return self.__process_playlist(ie_result, download) | |
1729 | finally: | |
1730 | self._playlist_level -= 1 | |
1731 | if not self._playlist_level: | |
1732 | self._playlist_urls.clear() | |
1733 | elif result_type == 'compat_list': | |
1734 | self.report_warning( | |
1735 | 'Extractor %s returned a compat_list result. ' | |
1736 | 'It needs to be updated.' % ie_result.get('extractor')) | |
1737 | ||
1738 | def _fixup(r): | |
1739 | self.add_extra_info(r, { | |
1740 | 'extractor': ie_result['extractor'], | |
1741 | 'webpage_url': ie_result['webpage_url'], | |
1742 | 'webpage_url_basename': url_basename(ie_result['webpage_url']), | |
1743 | 'webpage_url_domain': get_domain(ie_result['webpage_url']), | |
1744 | 'extractor_key': ie_result['extractor_key'], | |
1745 | }) | |
1746 | return r | |
1747 | ie_result['entries'] = [ | |
1748 | self.process_ie_result(_fixup(r), download, extra_info) | |
1749 | for r in ie_result['entries'] | |
1750 | ] | |
1751 | return ie_result | |
1752 | else: | |
1753 | raise Exception('Invalid result type: %s' % result_type) | |
1754 | ||
1755 | def _ensure_dir_exists(self, path): | |
1756 | return make_dir(path, self.report_error) | |
1757 | ||
1758 | @staticmethod | |
1759 | def _playlist_infodict(ie_result, strict=False, **kwargs): | |
1760 | info = { | |
1761 | 'playlist_count': ie_result.get('playlist_count'), | |
1762 | 'playlist': ie_result.get('title') or ie_result.get('id'), | |
1763 | 'playlist_id': ie_result.get('id'), | |
1764 | 'playlist_title': ie_result.get('title'), | |
1765 | 'playlist_uploader': ie_result.get('uploader'), | |
1766 | 'playlist_uploader_id': ie_result.get('uploader_id'), | |
1767 | **kwargs, | |
1768 | } | |
1769 | if strict: | |
1770 | return info | |
1771 | if ie_result.get('webpage_url'): | |
1772 | info.update({ | |
1773 | 'webpage_url': ie_result['webpage_url'], | |
1774 | 'webpage_url_basename': url_basename(ie_result['webpage_url']), | |
1775 | 'webpage_url_domain': get_domain(ie_result['webpage_url']), | |
1776 | }) | |
1777 | return { | |
1778 | **info, | |
1779 | 'playlist_index': 0, | |
1780 | '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)), | |
1781 | 'extractor': ie_result['extractor'], | |
1782 | 'extractor_key': ie_result['extractor_key'], | |
1783 | } | |
1784 | ||
1785 | def __process_playlist(self, ie_result, download): | |
1786 | """Process each entry in the playlist""" | |
1787 | assert ie_result['_type'] in ('playlist', 'multi_video') | |
1788 | ||
1789 | common_info = self._playlist_infodict(ie_result, strict=True) | |
1790 | title = common_info.get('playlist') or '<Untitled>' | |
1791 | if self._match_entry(common_info, incomplete=True) is not None: | |
1792 | return | |
1793 | self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}') | |
1794 | ||
1795 | all_entries = PlaylistEntries(self, ie_result) | |
1796 | entries = orderedSet(all_entries.get_requested_items(), lazy=True) | |
1797 | ||
1798 | lazy = self.params.get('lazy_playlist') | |
1799 | if lazy: | |
1800 | resolved_entries, n_entries = [], 'N/A' | |
1801 | ie_result['requested_entries'], ie_result['entries'] = None, None | |
1802 | else: | |
1803 | entries = resolved_entries = list(entries) | |
1804 | n_entries = len(resolved_entries) | |
1805 | ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], []) | |
1806 | if not ie_result.get('playlist_count'): | |
1807 | # Better to do this after potentially exhausting entries | |
1808 | ie_result['playlist_count'] = all_entries.get_full_count() | |
1809 | ||
1810 | extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries)) | |
1811 | ie_copy = collections.ChainMap(ie_result, extra) | |
1812 | ||
1813 | _infojson_written = False | |
1814 | write_playlist_files = self.params.get('allow_playlist_files', True) | |
1815 | if write_playlist_files and self.params.get('list_thumbnails'): | |
1816 | self.list_thumbnails(ie_result) | |
1817 | if write_playlist_files and not self.params.get('simulate'): | |
1818 | _infojson_written = self._write_info_json( | |
1819 | 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson')) | |
1820 | if _infojson_written is None: | |
1821 | return | |
1822 | if self._write_description('playlist', ie_result, | |
1823 | self.prepare_filename(ie_copy, 'pl_description')) is None: | |
1824 | return | |
1825 | # TODO: This should be passed to ThumbnailsConvertor if necessary | |
1826 | self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail')) | |
1827 | ||
1828 | if lazy: | |
1829 | if self.params.get('playlistreverse') or self.params.get('playlistrandom'): | |
1830 | self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True) | |
1831 | elif self.params.get('playlistreverse'): | |
1832 | entries.reverse() | |
1833 | elif self.params.get('playlistrandom'): | |
1834 | random.shuffle(entries) | |
1835 | ||
1836 | self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items' | |
1837 | f'{format_field(ie_result, "playlist_count", " of %s")}') | |
1838 | ||
1839 | keep_resolved_entries = self.params.get('extract_flat') != 'discard' | |
1840 | if self.params.get('extract_flat') == 'discard_in_playlist': | |
1841 | keep_resolved_entries = ie_result['_type'] != 'playlist' | |
1842 | if keep_resolved_entries: | |
1843 | self.write_debug('The information of all playlist entries will be held in memory') | |
1844 | ||
1845 | failures = 0 | |
1846 | max_failures = self.params.get('skip_playlist_after_errors') or float('inf') | |
1847 | for i, (playlist_index, entry) in enumerate(entries): | |
1848 | if lazy: | |
1849 | resolved_entries.append((playlist_index, entry)) | |
1850 | if not entry: | |
1851 | continue | |
1852 | ||
1853 | entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip') | |
1854 | if not lazy and 'playlist-index' in self.params.get('compat_opts', []): | |
1855 | playlist_index = ie_result['requested_entries'][i] | |
1856 | ||
1857 | entry_copy = collections.ChainMap(entry, { | |
1858 | **common_info, | |
1859 | 'n_entries': int_or_none(n_entries), | |
1860 | 'playlist_index': playlist_index, | |
1861 | 'playlist_autonumber': i + 1, | |
1862 | }) | |
1863 | ||
1864 | if self._match_entry(entry_copy, incomplete=True) is not None: | |
1865 | # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369 | |
1866 | resolved_entries[i] = (playlist_index, NO_DEFAULT) | |
1867 | continue | |
1868 | ||
1869 | self.to_screen('[download] Downloading item %s of %s' % ( | |
1870 | self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS))) | |
1871 | ||
1872 | entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({ | |
1873 | 'playlist_index': playlist_index, | |
1874 | 'playlist_autonumber': i + 1, | |
1875 | }, extra)) | |
1876 | if not entry_result: | |
1877 | failures += 1 | |
1878 | if failures >= max_failures: | |
1879 | self.report_error( | |
1880 | f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction') | |
1881 | break | |
1882 | if keep_resolved_entries: | |
1883 | resolved_entries[i] = (playlist_index, entry_result) | |
1884 | ||
1885 | # Update with processed data | |
1886 | ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT] | |
1887 | ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT] | |
1888 | if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))): | |
1889 | # Do not set for full playlist | |
1890 | ie_result.pop('requested_entries') | |
1891 | ||
1892 | # Write the updated info to json | |
1893 | if _infojson_written is True and self._write_info_json( | |
1894 | 'updated playlist', ie_result, | |
1895 | self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None: | |
1896 | return | |
1897 | ||
1898 | ie_result = self.run_all_pps('playlist', ie_result) | |
1899 | self.to_screen(f'[download] Finished downloading playlist: {title}') | |
1900 | return ie_result | |
1901 | ||
1902 | @_handle_extraction_exceptions | |
1903 | def __process_iterable_entry(self, entry, download, extra_info): | |
1904 | return self.process_ie_result( | |
1905 | entry, download=download, extra_info=extra_info) | |
1906 | ||
1907 | def _build_format_filter(self, filter_spec): | |
1908 | " Returns a function to filter the formats according to the filter_spec " | |
1909 | ||
1910 | OPERATORS = { | |
1911 | '<': operator.lt, | |
1912 | '<=': operator.le, | |
1913 | '>': operator.gt, | |
1914 | '>=': operator.ge, | |
1915 | '=': operator.eq, | |
1916 | '!=': operator.ne, | |
1917 | } | |
1918 | operator_rex = re.compile(r'''(?x)\s* | |
1919 | (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s* | |
1920 | (?P<op>%s)(?P<none_inclusive>\s*\?)?\s* | |
1921 | (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s* | |
1922 | ''' % '|'.join(map(re.escape, OPERATORS.keys()))) | |
1923 | m = operator_rex.fullmatch(filter_spec) | |
1924 | if m: | |
1925 | try: | |
1926 | comparison_value = int(m.group('value')) | |
1927 | except ValueError: | |
1928 | comparison_value = parse_filesize(m.group('value')) | |
1929 | if comparison_value is None: | |
1930 | comparison_value = parse_filesize(m.group('value') + 'B') | |
1931 | if comparison_value is None: | |
1932 | raise ValueError( | |
1933 | 'Invalid value %r in format specification %r' % ( | |
1934 | m.group('value'), filter_spec)) | |
1935 | op = OPERATORS[m.group('op')] | |
1936 | ||
1937 | if not m: | |
1938 | STR_OPERATORS = { | |
1939 | '=': operator.eq, | |
1940 | '^=': lambda attr, value: attr.startswith(value), | |
1941 | '$=': lambda attr, value: attr.endswith(value), | |
1942 | '*=': lambda attr, value: value in attr, | |
1943 | '~=': lambda attr, value: value.search(attr) is not None | |
1944 | } | |
1945 | str_operator_rex = re.compile(r'''(?x)\s* | |
1946 | (?P<key>[a-zA-Z0-9._-]+)\s* | |
1947 | (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)? | |
1948 | (?P<quote>["'])? | |
1949 | (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+)) | |
1950 | (?(quote)(?P=quote))\s* | |
1951 | ''' % '|'.join(map(re.escape, STR_OPERATORS.keys()))) | |
1952 | m = str_operator_rex.fullmatch(filter_spec) | |
1953 | if m: | |
1954 | if m.group('op') == '~=': | |
1955 | comparison_value = re.compile(m.group('value')) | |
1956 | else: | |
1957 | comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value')) | |
1958 | str_op = STR_OPERATORS[m.group('op')] | |
1959 | if m.group('negation'): | |
1960 | op = lambda attr, value: not str_op(attr, value) | |
1961 | else: | |
1962 | op = str_op | |
1963 | ||
1964 | if not m: | |
1965 | raise SyntaxError('Invalid filter specification %r' % filter_spec) | |
1966 | ||
1967 | def _filter(f): | |
1968 | actual_value = f.get(m.group('key')) | |
1969 | if actual_value is None: | |
1970 | return m.group('none_inclusive') | |
1971 | return op(actual_value, comparison_value) | |
1972 | return _filter | |
1973 | ||
1974 | def _check_formats(self, formats): | |
1975 | for f in formats: | |
1976 | self.to_screen('[info] Testing format %s' % f['format_id']) | |
1977 | path = self.get_output_path('temp') | |
1978 | if not self._ensure_dir_exists(f'{path}/'): | |
1979 | continue | |
1980 | temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None) | |
1981 | temp_file.close() | |
1982 | try: | |
1983 | success, _ = self.dl(temp_file.name, f, test=True) | |
1984 | except (DownloadError, OSError, ValueError) + network_exceptions: | |
1985 | success = False | |
1986 | finally: | |
1987 | if os.path.exists(temp_file.name): | |
1988 | try: | |
1989 | os.remove(temp_file.name) | |
1990 | except OSError: | |
1991 | self.report_warning('Unable to delete temporary file "%s"' % temp_file.name) | |
1992 | if success: | |
1993 | yield f | |
1994 | else: | |
1995 | self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id']) | |
1996 | ||
1997 | def _default_format_spec(self, info_dict, download=True): | |
1998 | ||
1999 | def can_merge(): | |
2000 | merger = FFmpegMergerPP(self) | |
2001 | return merger.available and merger.can_merge() | |
2002 | ||
2003 | prefer_best = ( | |
2004 | not self.params.get('simulate') | |
2005 | and download | |
2006 | and ( | |
2007 | not can_merge() | |
2008 | or info_dict.get('is_live') and not self.params.get('live_from_start') | |
2009 | or self.params['outtmpl']['default'] == '-')) | |
2010 | compat = ( | |
2011 | prefer_best | |
2012 | or self.params.get('allow_multiple_audio_streams', False) | |
2013 | or 'format-spec' in self.params['compat_opts']) | |
2014 | ||
2015 | return ( | |
2016 | 'best/bestvideo+bestaudio' if prefer_best | |
2017 | else 'bestvideo*+bestaudio/best' if not compat | |
2018 | else 'bestvideo+bestaudio/best') | |
2019 | ||
2020 | def build_format_selector(self, format_spec): | |
2021 | def syntax_error(note, start): | |
2022 | message = ( | |
2023 | 'Invalid format specification: ' | |
2024 | '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1])) | |
2025 | return SyntaxError(message) | |
2026 | ||
2027 | PICKFIRST = 'PICKFIRST' | |
2028 | MERGE = 'MERGE' | |
2029 | SINGLE = 'SINGLE' | |
2030 | GROUP = 'GROUP' | |
2031 | FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters']) | |
2032 | ||
2033 | allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False), | |
2034 | 'video': self.params.get('allow_multiple_video_streams', False)} | |
2035 | ||
2036 | check_formats = self.params.get('check_formats') == 'selected' | |
2037 | ||
2038 | def _parse_filter(tokens): | |
2039 | filter_parts = [] | |
2040 | for type, string, start, _, _ in tokens: | |
2041 | if type == tokenize.OP and string == ']': | |
2042 | return ''.join(filter_parts) | |
2043 | else: | |
2044 | filter_parts.append(string) | |
2045 | ||
2046 | def _remove_unused_ops(tokens): | |
2047 | # Remove operators that we don't use and join them with the surrounding strings. | |
2048 | # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9' | |
2049 | ALLOWED_OPS = ('/', '+', ',', '(', ')') | |
2050 | last_string, last_start, last_end, last_line = None, None, None, None | |
2051 | for type, string, start, end, line in tokens: | |
2052 | if type == tokenize.OP and string == '[': | |
2053 | if last_string: | |
2054 | yield tokenize.NAME, last_string, last_start, last_end, last_line | |
2055 | last_string = None | |
2056 | yield type, string, start, end, line | |
2057 | # everything inside brackets will be handled by _parse_filter | |
2058 | for type, string, start, end, line in tokens: | |
2059 | yield type, string, start, end, line | |
2060 | if type == tokenize.OP and string == ']': | |
2061 | break | |
2062 | elif type == tokenize.OP and string in ALLOWED_OPS: | |
2063 | if last_string: | |
2064 | yield tokenize.NAME, last_string, last_start, last_end, last_line | |
2065 | last_string = None | |
2066 | yield type, string, start, end, line | |
2067 | elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]: | |
2068 | if not last_string: | |
2069 | last_string = string | |
2070 | last_start = start | |
2071 | last_end = end | |
2072 | else: | |
2073 | last_string += string | |
2074 | if last_string: | |
2075 | yield tokenize.NAME, last_string, last_start, last_end, last_line | |
2076 | ||
2077 | def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False): | |
2078 | selectors = [] | |
2079 | current_selector = None | |
2080 | for type, string, start, _, _ in tokens: | |
2081 | # ENCODING is only defined in python 3.x | |
2082 | if type == getattr(tokenize, 'ENCODING', None): | |
2083 | continue | |
2084 | elif type in [tokenize.NAME, tokenize.NUMBER]: | |
2085 | current_selector = FormatSelector(SINGLE, string, []) | |
2086 | elif type == tokenize.OP: | |
2087 | if string == ')': | |
2088 | if not inside_group: | |
2089 | # ')' will be handled by the parentheses group | |
2090 | tokens.restore_last_token() | |
2091 | break | |
2092 | elif inside_merge and string in ['/', ',']: | |
2093 | tokens.restore_last_token() | |
2094 | break | |
2095 | elif inside_choice and string == ',': | |
2096 | tokens.restore_last_token() | |
2097 | break | |
2098 | elif string == ',': | |
2099 | if not current_selector: | |
2100 | raise syntax_error('"," must follow a format selector', start) | |
2101 | selectors.append(current_selector) | |
2102 | current_selector = None | |
2103 | elif string == '/': | |
2104 | if not current_selector: | |
2105 | raise syntax_error('"/" must follow a format selector', start) | |
2106 | first_choice = current_selector | |
2107 | second_choice = _parse_format_selection(tokens, inside_choice=True) | |
2108 | current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), []) | |
2109 | elif string == '[': | |
2110 | if not current_selector: | |
2111 | current_selector = FormatSelector(SINGLE, 'best', []) | |
2112 | format_filter = _parse_filter(tokens) | |
2113 | current_selector.filters.append(format_filter) | |
2114 | elif string == '(': | |
2115 | if current_selector: | |
2116 | raise syntax_error('Unexpected "("', start) | |
2117 | group = _parse_format_selection(tokens, inside_group=True) | |
2118 | current_selector = FormatSelector(GROUP, group, []) | |
2119 | elif string == '+': | |
2120 | if not current_selector: | |
2121 | raise syntax_error('Unexpected "+"', start) | |
2122 | selector_1 = current_selector | |
2123 | selector_2 = _parse_format_selection(tokens, inside_merge=True) | |
2124 | if not selector_2: | |
2125 | raise syntax_error('Expected a selector', start) | |
2126 | current_selector = FormatSelector(MERGE, (selector_1, selector_2), []) | |
2127 | else: | |
2128 | raise syntax_error(f'Operator not recognized: "{string}"', start) | |
2129 | elif type == tokenize.ENDMARKER: | |
2130 | break | |
2131 | if current_selector: | |
2132 | selectors.append(current_selector) | |
2133 | return selectors | |
2134 | ||
2135 | def _merge(formats_pair): | |
2136 | format_1, format_2 = formats_pair | |
2137 | ||
2138 | formats_info = [] | |
2139 | formats_info.extend(format_1.get('requested_formats', (format_1,))) | |
2140 | formats_info.extend(format_2.get('requested_formats', (format_2,))) | |
2141 | ||
2142 | if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']: | |
2143 | get_no_more = {'video': False, 'audio': False} | |
2144 | for (i, fmt_info) in enumerate(formats_info): | |
2145 | if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none': | |
2146 | formats_info.pop(i) | |
2147 | continue | |
2148 | for aud_vid in ['audio', 'video']: | |
2149 | if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none': | |
2150 | if get_no_more[aud_vid]: | |
2151 | formats_info.pop(i) | |
2152 | break | |
2153 | get_no_more[aud_vid] = True | |
2154 | ||
2155 | if len(formats_info) == 1: | |
2156 | return formats_info[0] | |
2157 | ||
2158 | video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none'] | |
2159 | audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none'] | |
2160 | ||
2161 | the_only_video = video_fmts[0] if len(video_fmts) == 1 else None | |
2162 | the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None | |
2163 | ||
2164 | output_ext = get_compatible_ext( | |
2165 | vcodecs=[f.get('vcodec') for f in video_fmts], | |
2166 | acodecs=[f.get('acodec') for f in audio_fmts], | |
2167 | vexts=[f['ext'] for f in video_fmts], | |
2168 | aexts=[f['ext'] for f in audio_fmts], | |
2169 | preferences=(try_call(lambda: self.params['merge_output_format'].split('/')) | |
2170 | or self.params.get('prefer_free_formats') and ('webm', 'mkv'))) | |
2171 | ||
2172 | filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info)) | |
2173 | ||
2174 | new_dict = { | |
2175 | 'requested_formats': formats_info, | |
2176 | 'format': '+'.join(filtered('format')), | |
2177 | 'format_id': '+'.join(filtered('format_id')), | |
2178 | 'ext': output_ext, | |
2179 | 'protocol': '+'.join(map(determine_protocol, formats_info)), | |
2180 | 'language': '+'.join(orderedSet(filtered('language'))) or None, | |
2181 | 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None, | |
2182 | 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None, | |
2183 | 'tbr': sum(filtered('tbr', 'vbr', 'abr')), | |
2184 | } | |
2185 | ||
2186 | if the_only_video: | |
2187 | new_dict.update({ | |
2188 | 'width': the_only_video.get('width'), | |
2189 | 'height': the_only_video.get('height'), | |
2190 | 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video), | |
2191 | 'fps': the_only_video.get('fps'), | |
2192 | 'dynamic_range': the_only_video.get('dynamic_range'), | |
2193 | 'vcodec': the_only_video.get('vcodec'), | |
2194 | 'vbr': the_only_video.get('vbr'), | |
2195 | 'stretched_ratio': the_only_video.get('stretched_ratio'), | |
2196 | 'aspect_ratio': the_only_video.get('aspect_ratio'), | |
2197 | }) | |
2198 | ||
2199 | if the_only_audio: | |
2200 | new_dict.update({ | |
2201 | 'acodec': the_only_audio.get('acodec'), | |
2202 | 'abr': the_only_audio.get('abr'), | |
2203 | 'asr': the_only_audio.get('asr'), | |
2204 | 'audio_channels': the_only_audio.get('audio_channels') | |
2205 | }) | |
2206 | ||
2207 | return new_dict | |
2208 | ||
2209 | def _check_formats(formats): | |
2210 | if not check_formats: | |
2211 | yield from formats | |
2212 | return | |
2213 | yield from self._check_formats(formats) | |
2214 | ||
2215 | def _build_selector_function(selector): | |
2216 | if isinstance(selector, list): # , | |
2217 | fs = [_build_selector_function(s) for s in selector] | |
2218 | ||
2219 | def selector_function(ctx): | |
2220 | for f in fs: | |
2221 | yield from f(ctx) | |
2222 | return selector_function | |
2223 | ||
2224 | elif selector.type == GROUP: # () | |
2225 | selector_function = _build_selector_function(selector.selector) | |
2226 | ||
2227 | elif selector.type == PICKFIRST: # / | |
2228 | fs = [_build_selector_function(s) for s in selector.selector] | |
2229 | ||
2230 | def selector_function(ctx): | |
2231 | for f in fs: | |
2232 | picked_formats = list(f(ctx)) | |
2233 | if picked_formats: | |
2234 | return picked_formats | |
2235 | return [] | |
2236 | ||
2237 | elif selector.type == MERGE: # + | |
2238 | selector_1, selector_2 = map(_build_selector_function, selector.selector) | |
2239 | ||
2240 | def selector_function(ctx): | |
2241 | for pair in itertools.product(selector_1(ctx), selector_2(ctx)): | |
2242 | yield _merge(pair) | |
2243 | ||
2244 | elif selector.type == SINGLE: # atom | |
2245 | format_spec = selector.selector or 'best' | |
2246 | ||
2247 | # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector | |
2248 | if format_spec == 'all': | |
2249 | def selector_function(ctx): | |
2250 | yield from _check_formats(ctx['formats'][::-1]) | |
2251 | elif format_spec == 'mergeall': | |
2252 | def selector_function(ctx): | |
2253 | formats = list(_check_formats( | |
2254 | f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none')) | |
2255 | if not formats: | |
2256 | return | |
2257 | merged_format = formats[-1] | |
2258 | for f in formats[-2::-1]: | |
2259 | merged_format = _merge((merged_format, f)) | |
2260 | yield merged_format | |
2261 | ||
2262 | else: | |
2263 | format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1 | |
2264 | mobj = re.match( | |
2265 | r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$', | |
2266 | format_spec) | |
2267 | if mobj is not None: | |
2268 | format_idx = int_or_none(mobj.group('n'), default=1) | |
2269 | format_reverse = mobj.group('bw')[0] == 'b' | |
2270 | format_type = (mobj.group('type') or [None])[0] | |
2271 | not_format_type = {'v': 'a', 'a': 'v'}.get(format_type) | |
2272 | format_modified = mobj.group('mod') is not None | |
2273 | ||
2274 | format_fallback = not format_type and not format_modified # for b, w | |
2275 | _filter_f = ( | |
2276 | (lambda f: f.get('%scodec' % format_type) != 'none') | |
2277 | if format_type and format_modified # bv*, ba*, wv*, wa* | |
2278 | else (lambda f: f.get('%scodec' % not_format_type) == 'none') | |
2279 | if format_type # bv, ba, wv, wa | |
2280 | else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none') | |
2281 | if not format_modified # b, w | |
2282 | else lambda f: True) # b*, w* | |
2283 | filter_f = lambda f: _filter_f(f) and ( | |
2284 | f.get('vcodec') != 'none' or f.get('acodec') != 'none') | |
2285 | else: | |
2286 | if format_spec in self._format_selection_exts['audio']: | |
2287 | filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' | |
2288 | elif format_spec in self._format_selection_exts['video']: | |
2289 | filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none' | |
2290 | seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none' | |
2291 | elif format_spec in self._format_selection_exts['storyboards']: | |
2292 | filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none' | |
2293 | else: | |
2294 | filter_f = lambda f: f.get('format_id') == format_spec # id | |
2295 | ||
2296 | def selector_function(ctx): | |
2297 | formats = list(ctx['formats']) | |
2298 | matches = list(filter(filter_f, formats)) if filter_f is not None else formats | |
2299 | if not matches: | |
2300 | if format_fallback and ctx['incomplete_formats']: | |
2301 | # for extractors with incomplete formats (audio only (soundcloud) | |
2302 | # or video only (imgur)) best/worst will fallback to | |
2303 | # best/worst {video,audio}-only format | |
2304 | matches = formats | |
2305 | elif seperate_fallback and not ctx['has_merged_format']: | |
2306 | # for compatibility with youtube-dl when there is no pre-merged format | |
2307 | matches = list(filter(seperate_fallback, formats)) | |
2308 | matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1])) | |
2309 | try: | |
2310 | yield matches[format_idx - 1] | |
2311 | except LazyList.IndexError: | |
2312 | return | |
2313 | ||
2314 | filters = [self._build_format_filter(f) for f in selector.filters] | |
2315 | ||
2316 | def final_selector(ctx): | |
2317 | ctx_copy = dict(ctx) | |
2318 | for _filter in filters: | |
2319 | ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats'])) | |
2320 | return selector_function(ctx_copy) | |
2321 | return final_selector | |
2322 | ||
2323 | stream = io.BytesIO(format_spec.encode()) | |
2324 | try: | |
2325 | tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline))) | |
2326 | except tokenize.TokenError: | |
2327 | raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec))) | |
2328 | ||
2329 | class TokenIterator: | |
2330 | def __init__(self, tokens): | |
2331 | self.tokens = tokens | |
2332 | self.counter = 0 | |
2333 | ||
2334 | def __iter__(self): | |
2335 | return self | |
2336 | ||
2337 | def __next__(self): | |
2338 | if self.counter >= len(self.tokens): | |
2339 | raise StopIteration() | |
2340 | value = self.tokens[self.counter] | |
2341 | self.counter += 1 | |
2342 | return value | |
2343 | ||
2344 | next = __next__ | |
2345 | ||
2346 | def restore_last_token(self): | |
2347 | self.counter -= 1 | |
2348 | ||
2349 | parsed_selector = _parse_format_selection(iter(TokenIterator(tokens))) | |
2350 | return _build_selector_function(parsed_selector) | |
2351 | ||
2352 | def _calc_headers(self, info_dict): | |
2353 | res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {}) | |
2354 | ||
2355 | cookies = self._calc_cookies(info_dict['url']) | |
2356 | if cookies: | |
2357 | res['Cookie'] = cookies | |
2358 | ||
2359 | if 'X-Forwarded-For' not in res: | |
2360 | x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip') | |
2361 | if x_forwarded_for_ip: | |
2362 | res['X-Forwarded-For'] = x_forwarded_for_ip | |
2363 | ||
2364 | return res | |
2365 | ||
2366 | def _calc_cookies(self, url): | |
2367 | pr = sanitized_Request(url) | |
2368 | self.cookiejar.add_cookie_header(pr) | |
2369 | return pr.get_header('Cookie') | |
2370 | ||
2371 | def _sort_thumbnails(self, thumbnails): | |
2372 | thumbnails.sort(key=lambda t: ( | |
2373 | t.get('preference') if t.get('preference') is not None else -1, | |
2374 | t.get('width') if t.get('width') is not None else -1, | |
2375 | t.get('height') if t.get('height') is not None else -1, | |
2376 | t.get('id') if t.get('id') is not None else '', | |
2377 | t.get('url'))) | |
2378 | ||
2379 | def _sanitize_thumbnails(self, info_dict): | |
2380 | thumbnails = info_dict.get('thumbnails') | |
2381 | if thumbnails is None: | |
2382 | thumbnail = info_dict.get('thumbnail') | |
2383 | if thumbnail: | |
2384 | info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}] | |
2385 | if not thumbnails: | |
2386 | return | |
2387 | ||
2388 | def check_thumbnails(thumbnails): | |
2389 | for t in thumbnails: | |
2390 | self.to_screen(f'[info] Testing thumbnail {t["id"]}') | |
2391 | try: | |
2392 | self.urlopen(HEADRequest(t['url'])) | |
2393 | except network_exceptions as err: | |
2394 | self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...') | |
2395 | continue | |
2396 | yield t | |
2397 | ||
2398 | self._sort_thumbnails(thumbnails) | |
2399 | for i, t in enumerate(thumbnails): | |
2400 | if t.get('id') is None: | |
2401 | t['id'] = '%d' % i | |
2402 | if t.get('width') and t.get('height'): | |
2403 | t['resolution'] = '%dx%d' % (t['width'], t['height']) | |
2404 | t['url'] = sanitize_url(t['url']) | |
2405 | ||
2406 | if self.params.get('check_formats') is True: | |
2407 | info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True) | |
2408 | else: | |
2409 | info_dict['thumbnails'] = thumbnails | |
2410 | ||
2411 | def _fill_common_fields(self, info_dict, final=True): | |
2412 | # TODO: move sanitization here | |
2413 | if final: | |
2414 | title = info_dict['fulltitle'] = info_dict.get('title') | |
2415 | if not title: | |
2416 | if title == '': | |
2417 | self.write_debug('Extractor gave empty title. Creating a generic title') | |
2418 | else: | |
2419 | self.report_warning('Extractor failed to obtain "title". Creating a generic title instead') | |
2420 | info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}' | |
2421 | ||
2422 | if info_dict.get('duration') is not None: | |
2423 | info_dict['duration_string'] = formatSeconds(info_dict['duration']) | |
2424 | ||
2425 | for ts_key, date_key in ( | |
2426 | ('timestamp', 'upload_date'), | |
2427 | ('release_timestamp', 'release_date'), | |
2428 | ('modified_timestamp', 'modified_date'), | |
2429 | ): | |
2430 | if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None: | |
2431 | # Working around out-of-range timestamp values (e.g. negative ones on Windows, | |
2432 | # see http://bugs.python.org/issue1646728) | |
2433 | with contextlib.suppress(ValueError, OverflowError, OSError): | |
2434 | upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key]) | |
2435 | info_dict[date_key] = upload_date.strftime('%Y%m%d') | |
2436 | ||
2437 | live_keys = ('is_live', 'was_live') | |
2438 | live_status = info_dict.get('live_status') | |
2439 | if live_status is None: | |
2440 | for key in live_keys: | |
2441 | if info_dict.get(key) is False: | |
2442 | continue | |
2443 | if info_dict.get(key): | |
2444 | live_status = key | |
2445 | break | |
2446 | if all(info_dict.get(key) is False for key in live_keys): | |
2447 | live_status = 'not_live' | |
2448 | if live_status: | |
2449 | info_dict['live_status'] = live_status | |
2450 | for key in live_keys: | |
2451 | if info_dict.get(key) is None: | |
2452 | info_dict[key] = (live_status == key) | |
2453 | if live_status == 'post_live': | |
2454 | info_dict['was_live'] = True | |
2455 | ||
2456 | # Auto generate title fields corresponding to the *_number fields when missing | |
2457 | # in order to always have clean titles. This is very common for TV series. | |
2458 | for field in ('chapter', 'season', 'episode'): | |
2459 | if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field): | |
2460 | info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field]) | |
2461 | ||
2462 | def _raise_pending_errors(self, info): | |
2463 | err = info.pop('__pending_error', None) | |
2464 | if err: | |
2465 | self.report_error(err, tb=False) | |
2466 | ||
2467 | def sort_formats(self, info_dict): | |
2468 | formats = self._get_formats(info_dict) | |
2469 | formats.sort(key=FormatSorter( | |
2470 | self, info_dict.get('_format_sort_fields') or []).calculate_preference) | |
2471 | ||
2472 | def process_video_result(self, info_dict, download=True): | |
2473 | assert info_dict.get('_type', 'video') == 'video' | |
2474 | self._num_videos += 1 | |
2475 | ||
2476 | if 'id' not in info_dict: | |
2477 | raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor']) | |
2478 | elif not info_dict.get('id'): | |
2479 | raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor']) | |
2480 | ||
2481 | def report_force_conversion(field, field_not, conversion): | |
2482 | self.report_warning( | |
2483 | '"%s" field is not %s - forcing %s conversion, there is an error in extractor' | |
2484 | % (field, field_not, conversion)) | |
2485 | ||
2486 | def sanitize_string_field(info, string_field): | |
2487 | field = info.get(string_field) | |
2488 | if field is None or isinstance(field, str): | |
2489 | return | |
2490 | report_force_conversion(string_field, 'a string', 'string') | |
2491 | info[string_field] = str(field) | |
2492 | ||
2493 | def sanitize_numeric_fields(info): | |
2494 | for numeric_field in self._NUMERIC_FIELDS: | |
2495 | field = info.get(numeric_field) | |
2496 | if field is None or isinstance(field, (int, float)): | |
2497 | continue | |
2498 | report_force_conversion(numeric_field, 'numeric', 'int') | |
2499 | info[numeric_field] = int_or_none(field) | |
2500 | ||
2501 | sanitize_string_field(info_dict, 'id') | |
2502 | sanitize_numeric_fields(info_dict) | |
2503 | if info_dict.get('section_end') and info_dict.get('section_start') is not None: | |
2504 | info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3) | |
2505 | if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None): | |
2506 | self.report_warning('"duration" field is negative, there is an error in extractor') | |
2507 | ||
2508 | chapters = info_dict.get('chapters') or [] | |
2509 | if chapters and chapters[0].get('start_time'): | |
2510 | chapters.insert(0, {'start_time': 0}) | |
2511 | ||
2512 | dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')} | |
2513 | for idx, (prev, current, next_) in enumerate(zip( | |
2514 | (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1): | |
2515 | if current.get('start_time') is None: | |
2516 | current['start_time'] = prev.get('end_time') | |
2517 | if not current.get('end_time'): | |
2518 | current['end_time'] = next_.get('start_time') | |
2519 | if not current.get('title'): | |
2520 | current['title'] = f'<Untitled Chapter {idx}>' | |
2521 | ||
2522 | if 'playlist' not in info_dict: | |
2523 | # It isn't part of a playlist | |
2524 | info_dict['playlist'] = None | |
2525 | info_dict['playlist_index'] = None | |
2526 | ||
2527 | self._sanitize_thumbnails(info_dict) | |
2528 | ||
2529 | thumbnail = info_dict.get('thumbnail') | |
2530 | thumbnails = info_dict.get('thumbnails') | |
2531 | if thumbnail: | |
2532 | info_dict['thumbnail'] = sanitize_url(thumbnail) | |
2533 | elif thumbnails: | |
2534 | info_dict['thumbnail'] = thumbnails[-1]['url'] | |
2535 | ||
2536 | if info_dict.get('display_id') is None and 'id' in info_dict: | |
2537 | info_dict['display_id'] = info_dict['id'] | |
2538 | ||
2539 | self._fill_common_fields(info_dict) | |
2540 | ||
2541 | for cc_kind in ('subtitles', 'automatic_captions'): | |
2542 | cc = info_dict.get(cc_kind) | |
2543 | if cc: | |
2544 | for _, subtitle in cc.items(): | |
2545 | for subtitle_format in subtitle: | |
2546 | if subtitle_format.get('url'): | |
2547 | subtitle_format['url'] = sanitize_url(subtitle_format['url']) | |
2548 | if subtitle_format.get('ext') is None: | |
2549 | subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower() | |
2550 | ||
2551 | automatic_captions = info_dict.get('automatic_captions') | |
2552 | subtitles = info_dict.get('subtitles') | |
2553 | ||
2554 | info_dict['requested_subtitles'] = self.process_subtitles( | |
2555 | info_dict['id'], subtitles, automatic_captions) | |
2556 | ||
2557 | formats = self._get_formats(info_dict) | |
2558 | ||
2559 | # Backward compatibility with InfoExtractor._sort_formats | |
2560 | field_preference = (formats or [{}])[0].pop('__sort_fields', None) | |
2561 | if field_preference: | |
2562 | info_dict['_format_sort_fields'] = field_preference | |
2563 | ||
2564 | # or None ensures --clean-infojson removes it | |
2565 | info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None | |
2566 | if not self.params.get('allow_unplayable_formats'): | |
2567 | formats = [f for f in formats if not f.get('has_drm')] | |
2568 | ||
2569 | if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats): | |
2570 | self.report_warning( | |
2571 | f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}' | |
2572 | 'only images are available for download. Use --list-formats to see them'.capitalize()) | |
2573 | ||
2574 | get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start')) | |
2575 | if not get_from_start: | |
2576 | info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M') | |
2577 | if info_dict.get('is_live') and formats: | |
2578 | formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start] | |
2579 | if get_from_start and not formats: | |
2580 | self.raise_no_formats(info_dict, msg=( | |
2581 | '--live-from-start is passed, but there are no formats that can be downloaded from the start. ' | |
2582 | 'If you want to download from the current time, use --no-live-from-start')) | |
2583 | ||
2584 | def is_wellformed(f): | |
2585 | url = f.get('url') | |
2586 | if not url: | |
2587 | self.report_warning( | |
2588 | '"url" field is missing or empty - skipping format, ' | |
2589 | 'there is an error in extractor') | |
2590 | return False | |
2591 | if isinstance(url, bytes): | |
2592 | sanitize_string_field(f, 'url') | |
2593 | return True | |
2594 | ||
2595 | # Filter out malformed formats for better extraction robustness | |
2596 | formats = list(filter(is_wellformed, formats or [])) | |
2597 | ||
2598 | if not formats: | |
2599 | self.raise_no_formats(info_dict) | |
2600 | ||
2601 | for format in formats: | |
2602 | sanitize_string_field(format, 'format_id') | |
2603 | sanitize_numeric_fields(format) | |
2604 | format['url'] = sanitize_url(format['url']) | |
2605 | if format.get('ext') is None: | |
2606 | format['ext'] = determine_ext(format['url']).lower() | |
2607 | if format.get('protocol') is None: | |
2608 | format['protocol'] = determine_protocol(format) | |
2609 | if format.get('resolution') is None: | |
2610 | format['resolution'] = self.format_resolution(format, default=None) | |
2611 | if format.get('dynamic_range') is None and format.get('vcodec') != 'none': | |
2612 | format['dynamic_range'] = 'SDR' | |
2613 | if format.get('aspect_ratio') is None: | |
2614 | format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2)) | |
2615 | if (info_dict.get('duration') and format.get('tbr') | |
2616 | and not format.get('filesize') and not format.get('filesize_approx')): | |
2617 | format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8)) | |
2618 | format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict)) | |
2619 | ||
2620 | # This is copied to http_headers by the above _calc_headers and can now be removed | |
2621 | if '__x_forwarded_for_ip' in info_dict: | |
2622 | del info_dict['__x_forwarded_for_ip'] | |
2623 | ||
2624 | self.sort_formats({ | |
2625 | 'formats': formats, | |
2626 | '_format_sort_fields': info_dict.get('_format_sort_fields') | |
2627 | }) | |
2628 | ||
2629 | # Sanitize and group by format_id | |
2630 | formats_dict = {} | |
2631 | for i, format in enumerate(formats): | |
2632 | if not format.get('format_id'): | |
2633 | format['format_id'] = str(i) | |
2634 | else: | |
2635 | # Sanitize format_id from characters used in format selector expression | |
2636 | format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id']) | |
2637 | formats_dict.setdefault(format['format_id'], []).append(format) | |
2638 | ||
2639 | # Make sure all formats have unique format_id | |
2640 | common_exts = set(itertools.chain(*self._format_selection_exts.values())) | |
2641 | for format_id, ambiguous_formats in formats_dict.items(): | |
2642 | ambigious_id = len(ambiguous_formats) > 1 | |
2643 | for i, format in enumerate(ambiguous_formats): | |
2644 | if ambigious_id: | |
2645 | format['format_id'] = '%s-%d' % (format_id, i) | |
2646 | # Ensure there is no conflict between id and ext in format selection | |
2647 | # See https://github.com/yt-dlp/yt-dlp/issues/1282 | |
2648 | if format['format_id'] != format['ext'] and format['format_id'] in common_exts: | |
2649 | format['format_id'] = 'f%s' % format['format_id'] | |
2650 | ||
2651 | if format.get('format') is None: | |
2652 | format['format'] = '{id} - {res}{note}'.format( | |
2653 | id=format['format_id'], | |
2654 | res=self.format_resolution(format), | |
2655 | note=format_field(format, 'format_note', ' (%s)'), | |
2656 | ) | |
2657 | ||
2658 | if self.params.get('check_formats') is True: | |
2659 | formats = LazyList(self._check_formats(formats[::-1]), reverse=True) | |
2660 | ||
2661 | if not formats or formats[0] is not info_dict: | |
2662 | # only set the 'formats' fields if the original info_dict list them | |
2663 | # otherwise we end up with a circular reference, the first (and unique) | |
2664 | # element in the 'formats' field in info_dict is info_dict itself, | |
2665 | # which can't be exported to json | |
2666 | info_dict['formats'] = formats | |
2667 | ||
2668 | info_dict, _ = self.pre_process(info_dict) | |
2669 | ||
2670 | if self._match_entry(info_dict, incomplete=self._format_fields) is not None: | |
2671 | return info_dict | |
2672 | ||
2673 | self.post_extract(info_dict) | |
2674 | info_dict, _ = self.pre_process(info_dict, 'after_filter') | |
2675 | ||
2676 | # The pre-processors may have modified the formats | |
2677 | formats = self._get_formats(info_dict) | |
2678 | ||
2679 | list_only = self.params.get('simulate') == 'list_only' | |
2680 | interactive_format_selection = not list_only and self.format_selector == '-' | |
2681 | if self.params.get('list_thumbnails'): | |
2682 | self.list_thumbnails(info_dict) | |
2683 | if self.params.get('listsubtitles'): | |
2684 | if 'automatic_captions' in info_dict: | |
2685 | self.list_subtitles( | |
2686 | info_dict['id'], automatic_captions, 'automatic captions') | |
2687 | self.list_subtitles(info_dict['id'], subtitles, 'subtitles') | |
2688 | if self.params.get('listformats') or interactive_format_selection: | |
2689 | self.list_formats(info_dict) | |
2690 | if list_only: | |
2691 | # Without this printing, -F --print-json will not work | |
2692 | self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True) | |
2693 | return info_dict | |
2694 | ||
2695 | format_selector = self.format_selector | |
2696 | if format_selector is None: | |
2697 | req_format = self._default_format_spec(info_dict, download=download) | |
2698 | self.write_debug('Default format spec: %s' % req_format) | |
2699 | format_selector = self.build_format_selector(req_format) | |
2700 | ||
2701 | while True: | |
2702 | if interactive_format_selection: | |
2703 | req_format = input( | |
2704 | self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS)) | |
2705 | try: | |
2706 | format_selector = self.build_format_selector(req_format) | |
2707 | except SyntaxError as err: | |
2708 | self.report_error(err, tb=False, is_error=False) | |
2709 | continue | |
2710 | ||
2711 | formats_to_download = list(format_selector({ | |
2712 | 'formats': formats, | |
2713 | 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats), | |
2714 | 'incomplete_formats': ( | |
2715 | # All formats are video-only or | |
2716 | all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) | |
2717 | # all formats are audio-only | |
2718 | or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)), | |
2719 | })) | |
2720 | if interactive_format_selection and not formats_to_download: | |
2721 | self.report_error('Requested format is not available', tb=False, is_error=False) | |
2722 | continue | |
2723 | break | |
2724 | ||
2725 | if not formats_to_download: | |
2726 | if not self.params.get('ignore_no_formats_error'): | |
2727 | raise ExtractorError( | |
2728 | 'Requested format is not available. Use --list-formats for a list of available formats', | |
2729 | expected=True, video_id=info_dict['id'], ie=info_dict['extractor']) | |
2730 | self.report_warning('Requested format is not available') | |
2731 | # Process what we can, even without any available formats. | |
2732 | formats_to_download = [{}] | |
2733 | ||
2734 | requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self)) | |
2735 | best_format, downloaded_formats = formats_to_download[-1], [] | |
2736 | if download: | |
2737 | if best_format and requested_ranges: | |
2738 | def to_screen(*msg): | |
2739 | self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}') | |
2740 | ||
2741 | to_screen(f'Downloading {len(formats_to_download)} format(s):', | |
2742 | (f['format_id'] for f in formats_to_download)) | |
2743 | if requested_ranges != ({}, ): | |
2744 | to_screen(f'Downloading {len(requested_ranges)} time ranges:', | |
2745 | (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges)) | |
2746 | max_downloads_reached = False | |
2747 | ||
2748 | for fmt, chapter in itertools.product(formats_to_download, requested_ranges): | |
2749 | new_info = self._copy_infodict(info_dict) | |
2750 | new_info.update(fmt) | |
2751 | offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf') | |
2752 | end_time = offset + min(chapter.get('end_time', duration), duration) | |
2753 | if chapter or offset: | |
2754 | new_info.update({ | |
2755 | 'section_start': offset + chapter.get('start_time', 0), | |
2756 | # duration may not be accurate. So allow deviations <1sec | |
2757 | 'section_end': end_time if end_time <= offset + duration + 1 else None, | |
2758 | 'section_title': chapter.get('title'), | |
2759 | 'section_number': chapter.get('index'), | |
2760 | }) | |
2761 | downloaded_formats.append(new_info) | |
2762 | try: | |
2763 | self.process_info(new_info) | |
2764 | except MaxDownloadsReached: | |
2765 | max_downloads_reached = True | |
2766 | self._raise_pending_errors(new_info) | |
2767 | # Remove copied info | |
2768 | for key, val in tuple(new_info.items()): | |
2769 | if info_dict.get(key) == val: | |
2770 | new_info.pop(key) | |
2771 | if max_downloads_reached: | |
2772 | break | |
2773 | ||
2774 | write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats} | |
2775 | assert write_archive.issubset({True, False, 'ignore'}) | |
2776 | if True in write_archive and False not in write_archive: | |
2777 | self.record_download_archive(info_dict) | |
2778 | ||
2779 | info_dict['requested_downloads'] = downloaded_formats | |
2780 | info_dict = self.run_all_pps('after_video', info_dict) | |
2781 | if max_downloads_reached: | |
2782 | raise MaxDownloadsReached() | |
2783 | ||
2784 | # We update the info dict with the selected best quality format (backwards compatibility) | |
2785 | info_dict.update(best_format) | |
2786 | return info_dict | |
2787 | ||
2788 | def process_subtitles(self, video_id, normal_subtitles, automatic_captions): | |
2789 | """Select the requested subtitles and their format""" | |
2790 | available_subs, normal_sub_langs = {}, [] | |
2791 | if normal_subtitles and self.params.get('writesubtitles'): | |
2792 | available_subs.update(normal_subtitles) | |
2793 | normal_sub_langs = tuple(normal_subtitles.keys()) | |
2794 | if automatic_captions and self.params.get('writeautomaticsub'): | |
2795 | for lang, cap_info in automatic_captions.items(): | |
2796 | if lang not in available_subs: | |
2797 | available_subs[lang] = cap_info | |
2798 | ||
2799 | if not available_subs or ( | |
2800 | not self.params.get('writesubtitles') | |
2801 | and not self.params.get('writeautomaticsub')): | |
2802 | return None | |
2803 | ||
2804 | all_sub_langs = tuple(available_subs.keys()) | |
2805 | if self.params.get('allsubtitles', False): | |
2806 | requested_langs = all_sub_langs | |
2807 | elif self.params.get('subtitleslangs', False): | |
2808 | try: | |
2809 | requested_langs = orderedSet_from_options( | |
2810 | self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True) | |
2811 | except re.error as e: | |
2812 | raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}') | |
2813 | elif normal_sub_langs: | |
2814 | requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1] | |
2815 | else: | |
2816 | requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1] | |
2817 | if requested_langs: | |
2818 | self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}') | |
2819 | ||
2820 | formats_query = self.params.get('subtitlesformat', 'best') | |
2821 | formats_preference = formats_query.split('/') if formats_query else [] | |
2822 | subs = {} | |
2823 | for lang in requested_langs: | |
2824 | formats = available_subs.get(lang) | |
2825 | if formats is None: | |
2826 | self.report_warning(f'{lang} subtitles not available for {video_id}') | |
2827 | continue | |
2828 | for ext in formats_preference: | |
2829 | if ext == 'best': | |
2830 | f = formats[-1] | |
2831 | break | |
2832 | matches = list(filter(lambda f: f['ext'] == ext, formats)) | |
2833 | if matches: | |
2834 | f = matches[-1] | |
2835 | break | |
2836 | else: | |
2837 | f = formats[-1] | |
2838 | self.report_warning( | |
2839 | 'No subtitle format found matching "%s" for language %s, ' | |
2840 | 'using %s' % (formats_query, lang, f['ext'])) | |
2841 | subs[lang] = f | |
2842 | return subs | |
2843 | ||
2844 | def _forceprint(self, key, info_dict): | |
2845 | if info_dict is None: | |
2846 | return | |
2847 | info_copy = info_dict.copy() | |
2848 | info_copy['formats_table'] = self.render_formats_table(info_dict) | |
2849 | info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict) | |
2850 | info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles')) | |
2851 | info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions')) | |
2852 | ||
2853 | def format_tmpl(tmpl): | |
2854 | mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl) | |
2855 | if not mobj: | |
2856 | return tmpl | |
2857 | ||
2858 | fmt = '%({})s' | |
2859 | if tmpl.startswith('{'): | |
2860 | tmpl = f'.{tmpl}' | |
2861 | if tmpl.endswith('='): | |
2862 | tmpl, fmt = tmpl[:-1], '{0} = %({0})#j' | |
2863 | return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(','))) | |
2864 | ||
2865 | for tmpl in self.params['forceprint'].get(key, []): | |
2866 | self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy)) | |
2867 | ||
2868 | for tmpl, file_tmpl in self.params['print_to_file'].get(key, []): | |
2869 | filename = self.prepare_filename(info_dict, outtmpl=file_tmpl) | |
2870 | tmpl = format_tmpl(tmpl) | |
2871 | self.to_screen(f'[info] Writing {tmpl!r} to: {filename}') | |
2872 | if self._ensure_dir_exists(filename): | |
2873 | with open(filename, 'a', encoding='utf-8') as f: | |
2874 | f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n') | |
2875 | ||
2876 | def __forced_printings(self, info_dict, filename, incomplete): | |
2877 | def print_mandatory(field, actual_field=None): | |
2878 | if actual_field is None: | |
2879 | actual_field = field | |
2880 | if (self.params.get('force%s' % field, False) | |
2881 | and (not incomplete or info_dict.get(actual_field) is not None)): | |
2882 | self.to_stdout(info_dict[actual_field]) | |
2883 | ||
2884 | def print_optional(field): | |
2885 | if (self.params.get('force%s' % field, False) | |
2886 | and info_dict.get(field) is not None): | |
2887 | self.to_stdout(info_dict[field]) | |
2888 | ||
2889 | info_dict = info_dict.copy() | |
2890 | if filename is not None: | |
2891 | info_dict['filename'] = filename | |
2892 | if info_dict.get('requested_formats') is not None: | |
2893 | # For RTMP URLs, also include the playpath | |
2894 | info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats']) | |
2895 | elif info_dict.get('url'): | |
2896 | info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '') | |
2897 | ||
2898 | if (self.params.get('forcejson') | |
2899 | or self.params['forceprint'].get('video') | |
2900 | or self.params['print_to_file'].get('video')): | |
2901 | self.post_extract(info_dict) | |
2902 | self._forceprint('video', info_dict) | |
2903 | ||
2904 | print_mandatory('title') | |
2905 | print_mandatory('id') | |
2906 | print_mandatory('url', 'urls') | |
2907 | print_optional('thumbnail') | |
2908 | print_optional('description') | |
2909 | print_optional('filename') | |
2910 | if self.params.get('forceduration') and info_dict.get('duration') is not None: | |
2911 | self.to_stdout(formatSeconds(info_dict['duration'])) | |
2912 | print_mandatory('format') | |
2913 | ||
2914 | if self.params.get('forcejson'): | |
2915 | self.to_stdout(json.dumps(self.sanitize_info(info_dict))) | |
2916 | ||
2917 | def dl(self, name, info, subtitle=False, test=False): | |
2918 | if not info.get('url'): | |
2919 | self.raise_no_formats(info, True) | |
2920 | ||
2921 | if test: | |
2922 | verbose = self.params.get('verbose') | |
2923 | params = { | |
2924 | 'test': True, | |
2925 | 'quiet': self.params.get('quiet') or not verbose, | |
2926 | 'verbose': verbose, | |
2927 | 'noprogress': not verbose, | |
2928 | 'nopart': True, | |
2929 | 'skip_unavailable_fragments': False, | |
2930 | 'keep_fragments': False, | |
2931 | 'overwrites': True, | |
2932 | '_no_ytdl_file': True, | |
2933 | } | |
2934 | else: | |
2935 | params = self.params | |
2936 | fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params) | |
2937 | if not test: | |
2938 | for ph in self._progress_hooks: | |
2939 | fd.add_progress_hook(ph) | |
2940 | urls = '", "'.join( | |
2941 | (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url']) | |
2942 | for f in info.get('requested_formats', []) or [info]) | |
2943 | self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"') | |
2944 | ||
2945 | # Note: Ideally info should be a deep-copied so that hooks cannot modify it. | |
2946 | # But it may contain objects that are not deep-copyable | |
2947 | new_info = self._copy_infodict(info) | |
2948 | if new_info.get('http_headers') is None: | |
2949 | new_info['http_headers'] = self._calc_headers(new_info) | |
2950 | return fd.download(name, new_info, subtitle) | |
2951 | ||
2952 | def existing_file(self, filepaths, *, default_overwrite=True): | |
2953 | existing_files = list(filter(os.path.exists, orderedSet(filepaths))) | |
2954 | if existing_files and not self.params.get('overwrites', default_overwrite): | |
2955 | return existing_files[0] | |
2956 | ||
2957 | for file in existing_files: | |
2958 | self.report_file_delete(file) | |
2959 | os.remove(file) | |
2960 | return None | |
2961 | ||
2962 | def process_info(self, info_dict): | |
2963 | """Process a single resolved IE result. (Modifies it in-place)""" | |
2964 | ||
2965 | assert info_dict.get('_type', 'video') == 'video' | |
2966 | original_infodict = info_dict | |
2967 | ||
2968 | if 'format' not in info_dict and 'ext' in info_dict: | |
2969 | info_dict['format'] = info_dict['ext'] | |
2970 | ||
2971 | if self._match_entry(info_dict) is not None: | |
2972 | info_dict['__write_download_archive'] = 'ignore' | |
2973 | return | |
2974 | ||
2975 | # Does nothing under normal operation - for backward compatibility of process_info | |
2976 | self.post_extract(info_dict) | |
2977 | ||
2978 | def replace_info_dict(new_info): | |
2979 | nonlocal info_dict | |
2980 | if new_info == info_dict: | |
2981 | return | |
2982 | info_dict.clear() | |
2983 | info_dict.update(new_info) | |
2984 | ||
2985 | new_info, _ = self.pre_process(info_dict, 'video') | |
2986 | replace_info_dict(new_info) | |
2987 | self._num_downloads += 1 | |
2988 | ||
2989 | # info_dict['_filename'] needs to be set for backward compatibility | |
2990 | info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True) | |
2991 | temp_filename = self.prepare_filename(info_dict, 'temp') | |
2992 | files_to_move = {} | |
2993 | ||
2994 | # Forced printings | |
2995 | self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict)) | |
2996 | ||
2997 | def check_max_downloads(): | |
2998 | if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'): | |
2999 | raise MaxDownloadsReached() | |
3000 | ||
3001 | if self.params.get('simulate'): | |
3002 | info_dict['__write_download_archive'] = self.params.get('force_write_download_archive') | |
3003 | check_max_downloads() | |
3004 | return | |
3005 | ||
3006 | if full_filename is None: | |
3007 | return | |
3008 | if not self._ensure_dir_exists(encodeFilename(full_filename)): | |
3009 | return | |
3010 | if not self._ensure_dir_exists(encodeFilename(temp_filename)): | |
3011 | return | |
3012 | ||
3013 | if self._write_description('video', info_dict, | |
3014 | self.prepare_filename(info_dict, 'description')) is None: | |
3015 | return | |
3016 | ||
3017 | sub_files = self._write_subtitles(info_dict, temp_filename) | |
3018 | if sub_files is None: | |
3019 | return | |
3020 | files_to_move.update(dict(sub_files)) | |
3021 | ||
3022 | thumb_files = self._write_thumbnails( | |
3023 | 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail')) | |
3024 | if thumb_files is None: | |
3025 | return | |
3026 | files_to_move.update(dict(thumb_files)) | |
3027 | ||
3028 | infofn = self.prepare_filename(info_dict, 'infojson') | |
3029 | _infojson_written = self._write_info_json('video', info_dict, infofn) | |
3030 | if _infojson_written: | |
3031 | info_dict['infojson_filename'] = infofn | |
3032 | # For backward compatibility, even though it was a private field | |
3033 | info_dict['__infojson_filename'] = infofn | |
3034 | elif _infojson_written is None: | |
3035 | return | |
3036 | ||
3037 | # Note: Annotations are deprecated | |
3038 | annofn = None | |
3039 | if self.params.get('writeannotations', False): | |
3040 | annofn = self.prepare_filename(info_dict, 'annotation') | |
3041 | if annofn: | |
3042 | if not self._ensure_dir_exists(encodeFilename(annofn)): | |
3043 | return | |
3044 | if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)): | |
3045 | self.to_screen('[info] Video annotations are already present') | |
3046 | elif not info_dict.get('annotations'): | |
3047 | self.report_warning('There are no annotations to write.') | |
3048 | else: | |
3049 | try: | |
3050 | self.to_screen('[info] Writing video annotations to: ' + annofn) | |
3051 | with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile: | |
3052 | annofile.write(info_dict['annotations']) | |
3053 | except (KeyError, TypeError): | |
3054 | self.report_warning('There are no annotations to write.') | |
3055 | except OSError: | |
3056 | self.report_error('Cannot write annotations file: ' + annofn) | |
3057 | return | |
3058 | ||
3059 | # Write internet shortcut files | |
3060 | def _write_link_file(link_type): | |
3061 | url = try_get(info_dict['webpage_url'], iri_to_uri) | |
3062 | if not url: | |
3063 | self.report_warning( | |
3064 | f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown') | |
3065 | return True | |
3066 | linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext')) | |
3067 | if not self._ensure_dir_exists(encodeFilename(linkfn)): | |
3068 | return False | |
3069 | if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)): | |
3070 | self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present') | |
3071 | return True | |
3072 | try: | |
3073 | self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}') | |
3074 | with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', | |
3075 | newline='\r\n' if link_type == 'url' else '\n') as linkfile: | |
3076 | template_vars = {'url': url} | |
3077 | if link_type == 'desktop': | |
3078 | template_vars['filename'] = linkfn[:-(len(link_type) + 1)] | |
3079 | linkfile.write(LINK_TEMPLATES[link_type] % template_vars) | |
3080 | except OSError: | |
3081 | self.report_error(f'Cannot write internet shortcut {linkfn}') | |
3082 | return False | |
3083 | return True | |
3084 | ||
3085 | write_links = { | |
3086 | 'url': self.params.get('writeurllink'), | |
3087 | 'webloc': self.params.get('writewebloclink'), | |
3088 | 'desktop': self.params.get('writedesktoplink'), | |
3089 | } | |
3090 | if self.params.get('writelink'): | |
3091 | link_type = ('webloc' if sys.platform == 'darwin' | |
3092 | else 'desktop' if sys.platform.startswith('linux') | |
3093 | else 'url') | |
3094 | write_links[link_type] = True | |
3095 | ||
3096 | if any(should_write and not _write_link_file(link_type) | |
3097 | for link_type, should_write in write_links.items()): | |
3098 | return | |
3099 | ||
3100 | new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move) | |
3101 | replace_info_dict(new_info) | |
3102 | ||
3103 | if self.params.get('skip_download'): | |
3104 | info_dict['filepath'] = temp_filename | |
3105 | info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename))) | |
3106 | info_dict['__files_to_move'] = files_to_move | |
3107 | replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)) | |
3108 | info_dict['__write_download_archive'] = self.params.get('force_write_download_archive') | |
3109 | else: | |
3110 | # Download | |
3111 | info_dict.setdefault('__postprocessors', []) | |
3112 | try: | |
3113 | ||
3114 | def existing_video_file(*filepaths): | |
3115 | ext = info_dict.get('ext') | |
3116 | converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext) | |
3117 | file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)), | |
3118 | default_overwrite=False) | |
3119 | if file: | |
3120 | info_dict['ext'] = os.path.splitext(file)[1][1:] | |
3121 | return file | |
3122 | ||
3123 | fd, success = None, True | |
3124 | if info_dict.get('protocol') or info_dict.get('url'): | |
3125 | fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-') | |
3126 | if fd is not FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and ( | |
3127 | info_dict.get('section_start') or info_dict.get('section_end')): | |
3128 | msg = ('This format cannot be partially downloaded' if FFmpegFD.available() | |
3129 | else 'You have requested downloading the video partially, but ffmpeg is not installed') | |
3130 | self.report_error(f'{msg}. Aborting') | |
3131 | return | |
3132 | ||
3133 | if info_dict.get('requested_formats') is not None: | |
3134 | requested_formats = info_dict['requested_formats'] | |
3135 | old_ext = info_dict['ext'] | |
3136 | if self.params.get('merge_output_format') is None: | |
3137 | if (info_dict['ext'] == 'webm' | |
3138 | and info_dict.get('thumbnails') | |
3139 | # check with type instead of pp_key, __name__, or isinstance | |
3140 | # since we dont want any custom PPs to trigger this | |
3141 | and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721 | |
3142 | info_dict['ext'] = 'mkv' | |
3143 | self.report_warning( | |
3144 | 'webm doesn\'t support embedding a thumbnail, mkv will be used') | |
3145 | new_ext = info_dict['ext'] | |
3146 | ||
3147 | def correct_ext(filename, ext=new_ext): | |
3148 | if filename == '-': | |
3149 | return filename | |
3150 | filename_real_ext = os.path.splitext(filename)[1][1:] | |
3151 | filename_wo_ext = ( | |
3152 | os.path.splitext(filename)[0] | |
3153 | if filename_real_ext in (old_ext, new_ext) | |
3154 | else filename) | |
3155 | return f'{filename_wo_ext}.{ext}' | |
3156 | ||
3157 | # Ensure filename always has a correct extension for successful merge | |
3158 | full_filename = correct_ext(full_filename) | |
3159 | temp_filename = correct_ext(temp_filename) | |
3160 | dl_filename = existing_video_file(full_filename, temp_filename) | |
3161 | info_dict['__real_download'] = False | |
3162 | ||
3163 | merger = FFmpegMergerPP(self) | |
3164 | downloaded = [] | |
3165 | if dl_filename is not None: | |
3166 | self.report_file_already_downloaded(dl_filename) | |
3167 | elif fd: | |
3168 | for f in requested_formats if fd != FFmpegFD else []: | |
3169 | f['filepath'] = fname = prepend_extension( | |
3170 | correct_ext(temp_filename, info_dict['ext']), | |
3171 | 'f%s' % f['format_id'], info_dict['ext']) | |
3172 | downloaded.append(fname) | |
3173 | info_dict['url'] = '\n'.join(f['url'] for f in requested_formats) | |
3174 | success, real_download = self.dl(temp_filename, info_dict) | |
3175 | info_dict['__real_download'] = real_download | |
3176 | else: | |
3177 | if self.params.get('allow_unplayable_formats'): | |
3178 | self.report_warning( | |
3179 | 'You have requested merging of multiple formats ' | |
3180 | 'while also allowing unplayable formats to be downloaded. ' | |
3181 | 'The formats won\'t be merged to prevent data corruption.') | |
3182 | elif not merger.available: | |
3183 | msg = 'You have requested merging of multiple formats but ffmpeg is not installed' | |
3184 | if not self.params.get('ignoreerrors'): | |
3185 | self.report_error(f'{msg}. Aborting due to --abort-on-error') | |
3186 | return | |
3187 | self.report_warning(f'{msg}. The formats won\'t be merged') | |
3188 | ||
3189 | if temp_filename == '-': | |
3190 | reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params) | |
3191 | else 'but the formats are incompatible for simultaneous download' if merger.available | |
3192 | else 'but ffmpeg is not installed') | |
3193 | self.report_warning( | |
3194 | f'You have requested downloading multiple formats to stdout {reason}. ' | |
3195 | 'The formats will be streamed one after the other') | |
3196 | fname = temp_filename | |
3197 | for f in requested_formats: | |
3198 | new_info = dict(info_dict) | |
3199 | del new_info['requested_formats'] | |
3200 | new_info.update(f) | |
3201 | if temp_filename != '-': | |
3202 | fname = prepend_extension( | |
3203 | correct_ext(temp_filename, new_info['ext']), | |
3204 | 'f%s' % f['format_id'], new_info['ext']) | |
3205 | if not self._ensure_dir_exists(fname): | |
3206 | return | |
3207 | f['filepath'] = fname | |
3208 | downloaded.append(fname) | |
3209 | partial_success, real_download = self.dl(fname, new_info) | |
3210 | info_dict['__real_download'] = info_dict['__real_download'] or real_download | |
3211 | success = success and partial_success | |
3212 | ||
3213 | if downloaded and merger.available and not self.params.get('allow_unplayable_formats'): | |
3214 | info_dict['__postprocessors'].append(merger) | |
3215 | info_dict['__files_to_merge'] = downloaded | |
3216 | # Even if there were no downloads, it is being merged only now | |
3217 | info_dict['__real_download'] = True | |
3218 | else: | |
3219 | for file in downloaded: | |
3220 | files_to_move[file] = None | |
3221 | else: | |
3222 | # Just a single file | |
3223 | dl_filename = existing_video_file(full_filename, temp_filename) | |
3224 | if dl_filename is None or dl_filename == temp_filename: | |
3225 | # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part. | |
3226 | # So we should try to resume the download | |
3227 | success, real_download = self.dl(temp_filename, info_dict) | |
3228 | info_dict['__real_download'] = real_download | |
3229 | else: | |
3230 | self.report_file_already_downloaded(dl_filename) | |
3231 | ||
3232 | dl_filename = dl_filename or temp_filename | |
3233 | info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename))) | |
3234 | ||
3235 | except network_exceptions as err: | |
3236 | self.report_error('unable to download video data: %s' % error_to_compat_str(err)) | |
3237 | return | |
3238 | except OSError as err: | |
3239 | raise UnavailableVideoError(err) | |
3240 | except (ContentTooShortError, ) as err: | |
3241 | self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})') | |
3242 | return | |
3243 | ||
3244 | self._raise_pending_errors(info_dict) | |
3245 | if success and full_filename != '-': | |
3246 | ||
3247 | def fixup(): | |
3248 | do_fixup = True | |
3249 | fixup_policy = self.params.get('fixup') | |
3250 | vid = info_dict['id'] | |
3251 | ||
3252 | if fixup_policy in ('ignore', 'never'): | |
3253 | return | |
3254 | elif fixup_policy == 'warn': | |
3255 | do_fixup = 'warn' | |
3256 | elif fixup_policy != 'force': | |
3257 | assert fixup_policy in ('detect_or_warn', None) | |
3258 | if not info_dict.get('__real_download'): | |
3259 | do_fixup = False | |
3260 | ||
3261 | def ffmpeg_fixup(cndn, msg, cls): | |
3262 | if not (do_fixup and cndn): | |
3263 | return | |
3264 | elif do_fixup == 'warn': | |
3265 | self.report_warning(f'{vid}: {msg}') | |
3266 | return | |
3267 | pp = cls(self) | |
3268 | if pp.available: | |
3269 | info_dict['__postprocessors'].append(pp) | |
3270 | else: | |
3271 | self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically') | |
3272 | ||
3273 | stretched_ratio = info_dict.get('stretched_ratio') | |
3274 | ffmpeg_fixup(stretched_ratio not in (1, None), | |
3275 | f'Non-uniform pixel ratio {stretched_ratio}', | |
3276 | FFmpegFixupStretchedPP) | |
3277 | ||
3278 | downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None | |
3279 | downloader = downloader.FD_NAME if downloader else None | |
3280 | ||
3281 | ext = info_dict.get('ext') | |
3282 | postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any(( | |
3283 | isinstance(pp, FFmpegVideoConvertorPP) | |
3284 | and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None) | |
3285 | ) for pp in self._pps['post_process']) | |
3286 | ||
3287 | if not postprocessed_by_ffmpeg: | |
3288 | ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash', | |
3289 | 'writing DASH m4a. Only some players support this container', | |
3290 | FFmpegFixupM4aPP) | |
3291 | ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts') | |
3292 | or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None, | |
3293 | 'Possible MPEG-TS in MP4 container or malformed AAC timestamps', | |
3294 | FFmpegFixupM3u8PP) | |
3295 | ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD', | |
3296 | 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP) | |
3297 | ||
3298 | ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP) | |
3299 | ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP) | |
3300 | ||
3301 | fixup() | |
3302 | try: | |
3303 | replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move)) | |
3304 | except PostProcessingError as err: | |
3305 | self.report_error('Postprocessing: %s' % str(err)) | |
3306 | return | |
3307 | try: | |
3308 | for ph in self._post_hooks: | |
3309 | ph(info_dict['filepath']) | |
3310 | except Exception as err: | |
3311 | self.report_error('post hooks: %s' % str(err)) | |
3312 | return | |
3313 | info_dict['__write_download_archive'] = True | |
3314 | ||
3315 | assert info_dict is original_infodict # Make sure the info_dict was modified in-place | |
3316 | if self.params.get('force_write_download_archive'): | |
3317 | info_dict['__write_download_archive'] = True | |
3318 | check_max_downloads() | |
3319 | ||
3320 | def __download_wrapper(self, func): | |
3321 | @functools.wraps(func) | |
3322 | def wrapper(*args, **kwargs): | |
3323 | try: | |
3324 | res = func(*args, **kwargs) | |
3325 | except UnavailableVideoError as e: | |
3326 | self.report_error(e) | |
3327 | except DownloadCancelled as e: | |
3328 | self.to_screen(f'[info] {e}') | |
3329 | if not self.params.get('break_per_url'): | |
3330 | raise | |
3331 | self._num_downloads = 0 | |
3332 | else: | |
3333 | if self.params.get('dump_single_json', False): | |
3334 | self.post_extract(res) | |
3335 | self.to_stdout(json.dumps(self.sanitize_info(res))) | |
3336 | return wrapper | |
3337 | ||
3338 | def download(self, url_list): | |
3339 | """Download a given list of URLs.""" | |
3340 | url_list = variadic(url_list) # Passing a single URL is a common mistake | |
3341 | outtmpl = self.params['outtmpl']['default'] | |
3342 | if (len(url_list) > 1 | |
3343 | and outtmpl != '-' | |
3344 | and '%' not in outtmpl | |
3345 | and self.params.get('max_downloads') != 1): | |
3346 | raise SameFileError(outtmpl) | |
3347 | ||
3348 | for url in url_list: | |
3349 | self.__download_wrapper(self.extract_info)( | |
3350 | url, force_generic_extractor=self.params.get('force_generic_extractor', False)) | |
3351 | ||
3352 | return self._download_retcode | |
3353 | ||
3354 | def download_with_info_file(self, info_filename): | |
3355 | with contextlib.closing(fileinput.FileInput( | |
3356 | [info_filename], mode='r', | |
3357 | openhook=fileinput.hook_encoded('utf-8'))) as f: | |
3358 | # FileInput doesn't have a read method, we can't call json.load | |
3359 | info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True)) | |
3360 | try: | |
3361 | self.__download_wrapper(self.process_ie_result)(info, download=True) | |
3362 | except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e: | |
3363 | if not isinstance(e, EntryNotInPlaylist): | |
3364 | self.to_stderr('\r') | |
3365 | webpage_url = info.get('webpage_url') | |
3366 | if webpage_url is not None: | |
3367 | self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}') | |
3368 | return self.download([webpage_url]) | |
3369 | else: | |
3370 | raise | |
3371 | return self._download_retcode | |
3372 | ||
3373 | @staticmethod | |
3374 | def sanitize_info(info_dict, remove_private_keys=False): | |
3375 | ''' Sanitize the infodict for converting to json ''' | |
3376 | if info_dict is None: | |
3377 | return info_dict | |
3378 | info_dict.setdefault('epoch', int(time.time())) | |
3379 | info_dict.setdefault('_type', 'video') | |
3380 | info_dict.setdefault('_version', { | |
3381 | 'version': __version__, | |
3382 | 'current_git_head': current_git_head(), | |
3383 | 'release_git_head': RELEASE_GIT_HEAD, | |
3384 | 'repository': REPOSITORY, | |
3385 | }) | |
3386 | ||
3387 | if remove_private_keys: | |
3388 | reject = lambda k, v: v is None or k.startswith('__') or k in { | |
3389 | 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries', | |
3390 | 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber', | |
3391 | '_format_sort_fields', | |
3392 | } | |
3393 | else: | |
3394 | reject = lambda k, v: False | |
3395 | ||
3396 | def filter_fn(obj): | |
3397 | if isinstance(obj, dict): | |
3398 | return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)} | |
3399 | elif isinstance(obj, (list, tuple, set, LazyList)): | |
3400 | return list(map(filter_fn, obj)) | |
3401 | elif obj is None or isinstance(obj, (str, int, float, bool)): | |
3402 | return obj | |
3403 | else: | |
3404 | return repr(obj) | |
3405 | ||
3406 | return filter_fn(info_dict) | |
3407 | ||
3408 | @staticmethod | |
3409 | def filter_requested_info(info_dict, actually_filter=True): | |
3410 | ''' Alias of sanitize_info for backward compatibility ''' | |
3411 | return YoutubeDL.sanitize_info(info_dict, actually_filter) | |
3412 | ||
3413 | def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None): | |
3414 | for filename in set(filter(None, files_to_delete)): | |
3415 | if msg: | |
3416 | self.to_screen(msg % filename) | |
3417 | try: | |
3418 | os.remove(filename) | |
3419 | except OSError: | |
3420 | self.report_warning(f'Unable to delete file {filename}') | |
3421 | if filename in info.get('__files_to_move', []): # NB: Delete even if None | |
3422 | del info['__files_to_move'][filename] | |
3423 | ||
3424 | @staticmethod | |
3425 | def post_extract(info_dict): | |
3426 | def actual_post_extract(info_dict): | |
3427 | if info_dict.get('_type') in ('playlist', 'multi_video'): | |
3428 | for video_dict in info_dict.get('entries', {}): | |
3429 | actual_post_extract(video_dict or {}) | |
3430 | return | |
3431 | ||
3432 | post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {}) | |
3433 | info_dict.update(post_extractor()) | |
3434 | ||
3435 | actual_post_extract(info_dict or {}) | |
3436 | ||
3437 | def run_pp(self, pp, infodict): | |
3438 | files_to_delete = [] | |
3439 | if '__files_to_move' not in infodict: | |
3440 | infodict['__files_to_move'] = {} | |
3441 | try: | |
3442 | files_to_delete, infodict = pp.run(infodict) | |
3443 | except PostProcessingError as e: | |
3444 | # Must be True and not 'only_download' | |
3445 | if self.params.get('ignoreerrors') is True: | |
3446 | self.report_error(e) | |
3447 | return infodict | |
3448 | raise | |
3449 | ||
3450 | if not files_to_delete: | |
3451 | return infodict | |
3452 | if self.params.get('keepvideo', False): | |
3453 | for f in files_to_delete: | |
3454 | infodict['__files_to_move'].setdefault(f, '') | |
3455 | else: | |
3456 | self._delete_downloaded_files( | |
3457 | *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)') | |
3458 | return infodict | |
3459 | ||
3460 | def run_all_pps(self, key, info, *, additional_pps=None): | |
3461 | if key != 'video': | |
3462 | self._forceprint(key, info) | |
3463 | for pp in (additional_pps or []) + self._pps[key]: | |
3464 | info = self.run_pp(pp, info) | |
3465 | return info | |
3466 | ||
3467 | def pre_process(self, ie_info, key='pre_process', files_to_move=None): | |
3468 | info = dict(ie_info) | |
3469 | info['__files_to_move'] = files_to_move or {} | |
3470 | try: | |
3471 | info = self.run_all_pps(key, info) | |
3472 | except PostProcessingError as err: | |
3473 | msg = f'Preprocessing: {err}' | |
3474 | info.setdefault('__pending_error', msg) | |
3475 | self.report_error(msg, is_error=False) | |
3476 | return info, info.pop('__files_to_move', None) | |
3477 | ||
3478 | def post_process(self, filename, info, files_to_move=None): | |
3479 | """Run all the postprocessors on the given file.""" | |
3480 | info['filepath'] = filename | |
3481 | info['__files_to_move'] = files_to_move or {} | |
3482 | info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors')) | |
3483 | info = self.run_pp(MoveFilesAfterDownloadPP(self), info) | |
3484 | del info['__files_to_move'] | |
3485 | return self.run_all_pps('after_move', info) | |
3486 | ||
3487 | def _make_archive_id(self, info_dict): | |
3488 | video_id = info_dict.get('id') | |
3489 | if not video_id: | |
3490 | return | |
3491 | # Future-proof against any change in case | |
3492 | # and backwards compatibility with prior versions | |
3493 | extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist | |
3494 | if extractor is None: | |
3495 | url = str_or_none(info_dict.get('url')) | |
3496 | if not url: | |
3497 | return | |
3498 | # Try to find matching extractor for the URL and take its ie_key | |
3499 | for ie_key, ie in self._ies.items(): | |
3500 | if ie.suitable(url): | |
3501 | extractor = ie_key | |
3502 | break | |
3503 | else: | |
3504 | return | |
3505 | return make_archive_id(extractor, video_id) | |
3506 | ||
3507 | def in_download_archive(self, info_dict): | |
3508 | if not self.archive: | |
3509 | return False | |
3510 | ||
3511 | vid_ids = [self._make_archive_id(info_dict)] | |
3512 | vid_ids.extend(info_dict.get('_old_archive_ids') or []) | |
3513 | return any(id_ in self.archive for id_ in vid_ids) | |
3514 | ||
3515 | def record_download_archive(self, info_dict): | |
3516 | fn = self.params.get('download_archive') | |
3517 | if fn is None: | |
3518 | return | |
3519 | vid_id = self._make_archive_id(info_dict) | |
3520 | assert vid_id | |
3521 | ||
3522 | self.write_debug(f'Adding to archive: {vid_id}') | |
3523 | if is_path_like(fn): | |
3524 | with locked_file(fn, 'a', encoding='utf-8') as archive_file: | |
3525 | archive_file.write(vid_id + '\n') | |
3526 | self.archive.add(vid_id) | |
3527 | ||
3528 | @staticmethod | |
3529 | def format_resolution(format, default='unknown'): | |
3530 | if format.get('vcodec') == 'none' and format.get('acodec') != 'none': | |
3531 | return 'audio only' | |
3532 | if format.get('resolution') is not None: | |
3533 | return format['resolution'] | |
3534 | if format.get('width') and format.get('height'): | |
3535 | return '%dx%d' % (format['width'], format['height']) | |
3536 | elif format.get('height'): | |
3537 | return '%sp' % format['height'] | |
3538 | elif format.get('width'): | |
3539 | return '%dx?' % format['width'] | |
3540 | return default | |
3541 | ||
3542 | def _list_format_headers(self, *headers): | |
3543 | if self.params.get('listformats_table', True) is not False: | |
3544 | return [self._format_out(header, self.Styles.HEADERS) for header in headers] | |
3545 | return headers | |
3546 | ||
3547 | def _format_note(self, fdict): | |
3548 | res = '' | |
3549 | if fdict.get('ext') in ['f4f', 'f4m']: | |
3550 | res += '(unsupported)' | |
3551 | if fdict.get('language'): | |
3552 | if res: | |
3553 | res += ' ' | |
3554 | res += '[%s]' % fdict['language'] | |
3555 | if fdict.get('format_note') is not None: | |
3556 | if res: | |
3557 | res += ' ' | |
3558 | res += fdict['format_note'] | |
3559 | if fdict.get('tbr') is not None: | |
3560 | if res: | |
3561 | res += ', ' | |
3562 | res += '%4dk' % fdict['tbr'] | |
3563 | if fdict.get('container') is not None: | |
3564 | if res: | |
3565 | res += ', ' | |
3566 | res += '%s container' % fdict['container'] | |
3567 | if (fdict.get('vcodec') is not None | |
3568 | and fdict.get('vcodec') != 'none'): | |
3569 | if res: | |
3570 | res += ', ' | |
3571 | res += fdict['vcodec'] | |
3572 | if fdict.get('vbr') is not None: | |
3573 | res += '@' | |
3574 | elif fdict.get('vbr') is not None and fdict.get('abr') is not None: | |
3575 | res += 'video@' | |
3576 | if fdict.get('vbr') is not None: | |
3577 | res += '%4dk' % fdict['vbr'] | |
3578 | if fdict.get('fps') is not None: | |
3579 | if res: | |
3580 | res += ', ' | |
3581 | res += '%sfps' % fdict['fps'] | |
3582 | if fdict.get('acodec') is not None: | |
3583 | if res: | |
3584 | res += ', ' | |
3585 | if fdict['acodec'] == 'none': | |
3586 | res += 'video only' | |
3587 | else: | |
3588 | res += '%-5s' % fdict['acodec'] | |
3589 | elif fdict.get('abr') is not None: | |
3590 | if res: | |
3591 | res += ', ' | |
3592 | res += 'audio' | |
3593 | if fdict.get('abr') is not None: | |
3594 | res += '@%3dk' % fdict['abr'] | |
3595 | if fdict.get('asr') is not None: | |
3596 | res += ' (%5dHz)' % fdict['asr'] | |
3597 | if fdict.get('filesize') is not None: | |
3598 | if res: | |
3599 | res += ', ' | |
3600 | res += format_bytes(fdict['filesize']) | |
3601 | elif fdict.get('filesize_approx') is not None: | |
3602 | if res: | |
3603 | res += ', ' | |
3604 | res += '~' + format_bytes(fdict['filesize_approx']) | |
3605 | return res | |
3606 | ||
3607 | def _get_formats(self, info_dict): | |
3608 | if info_dict.get('formats') is None: | |
3609 | if info_dict.get('url') and info_dict.get('_type', 'video') == 'video': | |
3610 | return [info_dict] | |
3611 | return [] | |
3612 | return info_dict['formats'] | |
3613 | ||
3614 | def render_formats_table(self, info_dict): | |
3615 | formats = self._get_formats(info_dict) | |
3616 | if not formats: | |
3617 | return | |
3618 | if not self.params.get('listformats_table', True) is not False: | |
3619 | table = [ | |
3620 | [ | |
3621 | format_field(f, 'format_id'), | |
3622 | format_field(f, 'ext'), | |
3623 | self.format_resolution(f), | |
3624 | self._format_note(f) | |
3625 | ] for f in formats if (f.get('preference') or 0) >= -1000] | |
3626 | return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1) | |
3627 | ||
3628 | def simplified_codec(f, field): | |
3629 | assert field in ('acodec', 'vcodec') | |
3630 | codec = f.get(field, 'unknown') | |
3631 | if not codec: | |
3632 | return 'unknown' | |
3633 | elif codec != 'none': | |
3634 | return '.'.join(codec.split('.')[:4]) | |
3635 | ||
3636 | if field == 'vcodec' and f.get('acodec') == 'none': | |
3637 | return 'images' | |
3638 | elif field == 'acodec' and f.get('vcodec') == 'none': | |
3639 | return '' | |
3640 | return self._format_out('audio only' if field == 'vcodec' else 'video only', | |
3641 | self.Styles.SUPPRESS) | |
3642 | ||
3643 | delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True) | |
3644 | table = [ | |
3645 | [ | |
3646 | self._format_out(format_field(f, 'format_id'), self.Styles.ID), | |
3647 | format_field(f, 'ext'), | |
3648 | format_field(f, func=self.format_resolution, ignore=('audio only', 'images')), | |
3649 | format_field(f, 'fps', '\t%d', func=round), | |
3650 | format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''), | |
3651 | format_field(f, 'audio_channels', '\t%s'), | |
3652 | delim, | |
3653 | format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes), | |
3654 | format_field(f, 'tbr', '\t%dk', func=round), | |
3655 | shorten_protocol_name(f.get('protocol', '')), | |
3656 | delim, | |
3657 | simplified_codec(f, 'vcodec'), | |
3658 | format_field(f, 'vbr', '\t%dk', func=round), | |
3659 | simplified_codec(f, 'acodec'), | |
3660 | format_field(f, 'abr', '\t%dk', func=round), | |
3661 | format_field(f, 'asr', '\t%s', func=format_decimal_suffix), | |
3662 | join_nonempty( | |
3663 | self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None, | |
3664 | format_field(f, 'language', '[%s]'), | |
3665 | join_nonempty(format_field(f, 'format_note'), | |
3666 | format_field(f, 'container', ignore=(None, f.get('ext'))), | |
3667 | delim=', '), | |
3668 | delim=' '), | |
3669 | ] for f in formats if f.get('preference') is None or f['preference'] >= -1000] | |
3670 | header_line = self._list_format_headers( | |
3671 | 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO', | |
3672 | delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO') | |
3673 | ||
3674 | return render_table( | |
3675 | header_line, table, hide_empty=True, | |
3676 | delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True)) | |
3677 | ||
3678 | def render_thumbnails_table(self, info_dict): | |
3679 | thumbnails = list(info_dict.get('thumbnails') or []) | |
3680 | if not thumbnails: | |
3681 | return None | |
3682 | return render_table( | |
3683 | self._list_format_headers('ID', 'Width', 'Height', 'URL'), | |
3684 | [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails]) | |
3685 | ||
3686 | def render_subtitles_table(self, video_id, subtitles): | |
3687 | def _row(lang, formats): | |
3688 | exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats))) | |
3689 | if len(set(names)) == 1: | |
3690 | names = [] if names[0] == 'unknown' else names[:1] | |
3691 | return [lang, ', '.join(names), ', '.join(exts)] | |
3692 | ||
3693 | if not subtitles: | |
3694 | return None | |
3695 | return render_table( | |
3696 | self._list_format_headers('Language', 'Name', 'Formats'), | |
3697 | [_row(lang, formats) for lang, formats in subtitles.items()], | |
3698 | hide_empty=True) | |
3699 | ||
3700 | def __list_table(self, video_id, name, func, *args): | |
3701 | table = func(*args) | |
3702 | if not table: | |
3703 | self.to_screen(f'{video_id} has no {name}') | |
3704 | return | |
3705 | self.to_screen(f'[info] Available {name} for {video_id}:') | |
3706 | self.to_stdout(table) | |
3707 | ||
3708 | def list_formats(self, info_dict): | |
3709 | self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict) | |
3710 | ||
3711 | def list_thumbnails(self, info_dict): | |
3712 | self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict) | |
3713 | ||
3714 | def list_subtitles(self, video_id, subtitles, name='subtitles'): | |
3715 | self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles) | |
3716 | ||
3717 | def urlopen(self, req): | |
3718 | """ Start an HTTP download """ | |
3719 | if isinstance(req, str): | |
3720 | req = sanitized_Request(req) | |
3721 | return self._opener.open(req, timeout=self._socket_timeout) | |
3722 | ||
3723 | def print_debug_header(self): | |
3724 | if not self.params.get('verbose'): | |
3725 | return | |
3726 | ||
3727 | from . import _IN_CLI # Must be delayed import | |
3728 | ||
3729 | # These imports can be slow. So import them only as needed | |
3730 | from .extractor.extractors import _LAZY_LOADER | |
3731 | from .extractor.extractors import ( | |
3732 | _PLUGIN_CLASSES as plugin_ies, | |
3733 | _PLUGIN_OVERRIDES as plugin_ie_overrides | |
3734 | ) | |
3735 | ||
3736 | def get_encoding(stream): | |
3737 | ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__)) | |
3738 | if not supports_terminal_sequences(stream): | |
3739 | from .utils import WINDOWS_VT_MODE # Must be imported locally | |
3740 | ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)' | |
3741 | return ret | |
3742 | ||
3743 | encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % ( | |
3744 | locale.getpreferredencoding(), | |
3745 | sys.getfilesystemencoding(), | |
3746 | self.get_encoding(), | |
3747 | ', '.join( | |
3748 | f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_ | |
3749 | if stream is not None and key != 'console') | |
3750 | ) | |
3751 | ||
3752 | logger = self.params.get('logger') | |
3753 | if logger: | |
3754 | write_debug = lambda msg: logger.debug(f'[debug] {msg}') | |
3755 | write_debug(encoding_str) | |
3756 | else: | |
3757 | write_string(f'[debug] {encoding_str}\n', encoding=None) | |
3758 | write_debug = lambda msg: self._write_string(f'[debug] {msg}\n') | |
3759 | ||
3760 | source = detect_variant() | |
3761 | if VARIANT not in (None, 'pip'): | |
3762 | source += '*' | |
3763 | write_debug(join_nonempty( | |
3764 | f'{"yt-dlp" if REPOSITORY == "yt-dlp/yt-dlp" else REPOSITORY} version', | |
3765 | __version__, | |
3766 | f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '', | |
3767 | '' if source == 'unknown' else f'({source})', | |
3768 | '' if _IN_CLI else 'API', | |
3769 | delim=' ')) | |
3770 | ||
3771 | if not _IN_CLI: | |
3772 | write_debug(f'params: {self.params}') | |
3773 | ||
3774 | if not _LAZY_LOADER: | |
3775 | if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'): | |
3776 | write_debug('Lazy loading extractors is forcibly disabled') | |
3777 | else: | |
3778 | write_debug('Lazy loading extractors is disabled') | |
3779 | if self.params['compat_opts']: | |
3780 | write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts'])) | |
3781 | ||
3782 | if current_git_head(): | |
3783 | write_debug(f'Git HEAD: {current_git_head()}') | |
3784 | write_debug(system_identifier()) | |
3785 | ||
3786 | exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self) | |
3787 | ffmpeg_features = {key for key, val in ffmpeg_features.items() if val} | |
3788 | if ffmpeg_features: | |
3789 | exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features)) | |
3790 | ||
3791 | exe_versions['rtmpdump'] = rtmpdump_version() | |
3792 | exe_versions['phantomjs'] = PhantomJSwrapper._version() | |
3793 | exe_str = ', '.join( | |
3794 | f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v | |
3795 | ) or 'none' | |
3796 | write_debug('exe versions: %s' % exe_str) | |
3797 | ||
3798 | from .compat.compat_utils import get_package_info | |
3799 | from .dependencies import available_dependencies | |
3800 | ||
3801 | write_debug('Optional libraries: %s' % (', '.join(sorted({ | |
3802 | join_nonempty(*get_package_info(m)) for m in available_dependencies.values() | |
3803 | })) or 'none')) | |
3804 | ||
3805 | self._setup_opener() | |
3806 | proxy_map = {} | |
3807 | for handler in self._opener.handlers: | |
3808 | if hasattr(handler, 'proxies'): | |
3809 | proxy_map.update(handler.proxies) | |
3810 | write_debug(f'Proxy map: {proxy_map}') | |
3811 | ||
3812 | for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items(): | |
3813 | display_list = ['%s%s' % ( | |
3814 | klass.__name__, '' if klass.__name__ == name else f' as {name}') | |
3815 | for name, klass in plugins.items()] | |
3816 | if plugin_type == 'Extractor': | |
3817 | display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})' | |
3818 | for parent, plugins in plugin_ie_overrides.items()) | |
3819 | if not display_list: | |
3820 | continue | |
3821 | write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}') | |
3822 | ||
3823 | plugin_dirs = plugin_directories() | |
3824 | if plugin_dirs: | |
3825 | write_debug(f'Plugin directories: {plugin_dirs}') | |
3826 | ||
3827 | # Not implemented | |
3828 | if False and self.params.get('call_home'): | |
3829 | ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode() | |
3830 | write_debug('Public IP address: %s' % ipaddr) | |
3831 | latest_version = self.urlopen( | |
3832 | 'https://yt-dl.org/latest/version').read().decode() | |
3833 | if version_tuple(latest_version) > version_tuple(__version__): | |
3834 | self.report_warning( | |
3835 | 'You are using an outdated version (newest version: %s)! ' | |
3836 | 'See https://yt-dl.org/update if you need help updating.' % | |
3837 | latest_version) | |
3838 | ||
3839 | def _setup_opener(self): | |
3840 | if hasattr(self, '_opener'): | |
3841 | return | |
3842 | timeout_val = self.params.get('socket_timeout') | |
3843 | self._socket_timeout = 20 if timeout_val is None else float(timeout_val) | |
3844 | ||
3845 | opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser') | |
3846 | opts_cookiefile = self.params.get('cookiefile') | |
3847 | opts_proxy = self.params.get('proxy') | |
3848 | ||
3849 | self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self) | |
3850 | ||
3851 | cookie_processor = YoutubeDLCookieProcessor(self.cookiejar) | |
3852 | if opts_proxy is not None: | |
3853 | if opts_proxy == '': | |
3854 | proxies = {} | |
3855 | else: | |
3856 | proxies = {'http': opts_proxy, 'https': opts_proxy} | |
3857 | else: | |
3858 | proxies = urllib.request.getproxies() | |
3859 | # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805) | |
3860 | if 'http' in proxies and 'https' not in proxies: | |
3861 | proxies['https'] = proxies['http'] | |
3862 | proxy_handler = PerRequestProxyHandler(proxies) | |
3863 | ||
3864 | debuglevel = 1 if self.params.get('debug_printtraffic') else 0 | |
3865 | https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel) | |
3866 | ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel) | |
3867 | redirect_handler = YoutubeDLRedirectHandler() | |
3868 | data_handler = urllib.request.DataHandler() | |
3869 | ||
3870 | # When passing our own FileHandler instance, build_opener won't add the | |
3871 | # default FileHandler and allows us to disable the file protocol, which | |
3872 | # can be used for malicious purposes (see | |
3873 | # https://github.com/ytdl-org/youtube-dl/issues/8227) | |
3874 | file_handler = urllib.request.FileHandler() | |
3875 | ||
3876 | if not self.params.get('enable_file_urls'): | |
3877 | def file_open(*args, **kwargs): | |
3878 | raise urllib.error.URLError( | |
3879 | 'file:// URLs are explicitly disabled in yt-dlp for security reasons. ' | |
3880 | 'Use --enable-file-urls to enable at your own risk.') | |
3881 | file_handler.file_open = file_open | |
3882 | ||
3883 | opener = urllib.request.build_opener( | |
3884 | proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler) | |
3885 | ||
3886 | # Delete the default user-agent header, which would otherwise apply in | |
3887 | # cases where our custom HTTP handler doesn't come into play | |
3888 | # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details) | |
3889 | opener.addheaders = [] | |
3890 | self._opener = opener | |
3891 | ||
3892 | def encode(self, s): | |
3893 | if isinstance(s, bytes): | |
3894 | return s # Already encoded | |
3895 | ||
3896 | try: | |
3897 | return s.encode(self.get_encoding()) | |
3898 | except UnicodeEncodeError as err: | |
3899 | err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.' | |
3900 | raise | |
3901 | ||
3902 | def get_encoding(self): | |
3903 | encoding = self.params.get('encoding') | |
3904 | if encoding is None: | |
3905 | encoding = preferredencoding() | |
3906 | return encoding | |
3907 | ||
3908 | def _write_info_json(self, label, ie_result, infofn, overwrite=None): | |
3909 | ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error ''' | |
3910 | if overwrite is None: | |
3911 | overwrite = self.params.get('overwrites', True) | |
3912 | if not self.params.get('writeinfojson'): | |
3913 | return False | |
3914 | elif not infofn: | |
3915 | self.write_debug(f'Skipping writing {label} infojson') | |
3916 | return False | |
3917 | elif not self._ensure_dir_exists(infofn): | |
3918 | return None | |
3919 | elif not overwrite and os.path.exists(infofn): | |
3920 | self.to_screen(f'[info] {label.title()} metadata is already present') | |
3921 | return 'exists' | |
3922 | ||
3923 | self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}') | |
3924 | try: | |
3925 | write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn) | |
3926 | return True | |
3927 | except OSError: | |
3928 | self.report_error(f'Cannot write {label} metadata to JSON file {infofn}') | |
3929 | return None | |
3930 | ||
3931 | def _write_description(self, label, ie_result, descfn): | |
3932 | ''' Write description and returns True = written, False = skip, None = error ''' | |
3933 | if not self.params.get('writedescription'): | |
3934 | return False | |
3935 | elif not descfn: | |
3936 | self.write_debug(f'Skipping writing {label} description') | |
3937 | return False | |
3938 | elif not self._ensure_dir_exists(descfn): | |
3939 | return None | |
3940 | elif not self.params.get('overwrites', True) and os.path.exists(descfn): | |
3941 | self.to_screen(f'[info] {label.title()} description is already present') | |
3942 | elif ie_result.get('description') is None: | |
3943 | self.to_screen(f'[info] There\'s no {label} description to write') | |
3944 | return False | |
3945 | else: | |
3946 | try: | |
3947 | self.to_screen(f'[info] Writing {label} description to: {descfn}') | |
3948 | with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile: | |
3949 | descfile.write(ie_result['description']) | |
3950 | except OSError: | |
3951 | self.report_error(f'Cannot write {label} description file {descfn}') | |
3952 | return None | |
3953 | return True | |
3954 | ||
3955 | def _write_subtitles(self, info_dict, filename): | |
3956 | ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error''' | |
3957 | ret = [] | |
3958 | subtitles = info_dict.get('requested_subtitles') | |
3959 | if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')): | |
3960 | # subtitles download errors are already managed as troubles in relevant IE | |
3961 | # that way it will silently go on when used with unsupporting IE | |
3962 | return ret | |
3963 | elif not subtitles: | |
3964 | self.to_screen('[info] There\'s no subtitles for the requested languages') | |
3965 | return ret | |
3966 | sub_filename_base = self.prepare_filename(info_dict, 'subtitle') | |
3967 | if not sub_filename_base: | |
3968 | self.to_screen('[info] Skipping writing video subtitles') | |
3969 | return ret | |
3970 | ||
3971 | for sub_lang, sub_info in subtitles.items(): | |
3972 | sub_format = sub_info['ext'] | |
3973 | sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext')) | |
3974 | sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext')) | |
3975 | existing_sub = self.existing_file((sub_filename_final, sub_filename)) | |
3976 | if existing_sub: | |
3977 | self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present') | |
3978 | sub_info['filepath'] = existing_sub | |
3979 | ret.append((existing_sub, sub_filename_final)) | |
3980 | continue | |
3981 | ||
3982 | self.to_screen(f'[info] Writing video subtitles to: {sub_filename}') | |
3983 | if sub_info.get('data') is not None: | |
3984 | try: | |
3985 | # Use newline='' to prevent conversion of newline characters | |
3986 | # See https://github.com/ytdl-org/youtube-dl/issues/10268 | |
3987 | with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile: | |
3988 | subfile.write(sub_info['data']) | |
3989 | sub_info['filepath'] = sub_filename | |
3990 | ret.append((sub_filename, sub_filename_final)) | |
3991 | continue | |
3992 | except OSError: | |
3993 | self.report_error(f'Cannot write video subtitles file {sub_filename}') | |
3994 | return None | |
3995 | ||
3996 | try: | |
3997 | sub_copy = sub_info.copy() | |
3998 | sub_copy.setdefault('http_headers', info_dict.get('http_headers')) | |
3999 | self.dl(sub_filename, sub_copy, subtitle=True) | |
4000 | sub_info['filepath'] = sub_filename | |
4001 | ret.append((sub_filename, sub_filename_final)) | |
4002 | except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err: | |
4003 | msg = f'Unable to download video subtitles for {sub_lang!r}: {err}' | |
4004 | if self.params.get('ignoreerrors') is not True: # False or 'only_download' | |
4005 | if not self.params.get('ignoreerrors'): | |
4006 | self.report_error(msg) | |
4007 | raise DownloadError(msg) | |
4008 | self.report_warning(msg) | |
4009 | return ret | |
4010 | ||
4011 | def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None): | |
4012 | ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) ''' | |
4013 | write_all = self.params.get('write_all_thumbnails', False) | |
4014 | thumbnails, ret = [], [] | |
4015 | if write_all or self.params.get('writethumbnail', False): | |
4016 | thumbnails = info_dict.get('thumbnails') or [] | |
4017 | if not thumbnails: | |
4018 | self.to_screen(f'[info] There\'s no {label} thumbnails to download') | |
4019 | return ret | |
4020 | multiple = write_all and len(thumbnails) > 1 | |
4021 | ||
4022 | if thumb_filename_base is None: | |
4023 | thumb_filename_base = filename | |
4024 | if thumbnails and not thumb_filename_base: | |
4025 | self.write_debug(f'Skipping writing {label} thumbnail') | |
4026 | return ret | |
4027 | ||
4028 | for idx, t in list(enumerate(thumbnails))[::-1]: | |
4029 | thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg') | |
4030 | thumb_display_id = f'{label} thumbnail {t["id"]}' | |
4031 | thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext')) | |
4032 | thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext')) | |
4033 | ||
4034 | existing_thumb = self.existing_file((thumb_filename_final, thumb_filename)) | |
4035 | if existing_thumb: | |
4036 | self.to_screen('[info] %s is already present' % ( | |
4037 | thumb_display_id if multiple else f'{label} thumbnail').capitalize()) | |
4038 | t['filepath'] = existing_thumb | |
4039 | ret.append((existing_thumb, thumb_filename_final)) | |
4040 | else: | |
4041 | self.to_screen(f'[info] Downloading {thumb_display_id} ...') | |
4042 | try: | |
4043 | uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {}))) | |
4044 | self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}') | |
4045 | with open(encodeFilename(thumb_filename), 'wb') as thumbf: | |
4046 | shutil.copyfileobj(uf, thumbf) | |
4047 | ret.append((thumb_filename, thumb_filename_final)) | |
4048 | t['filepath'] = thumb_filename | |
4049 | except network_exceptions as err: | |
4050 | thumbnails.pop(idx) | |
4051 | self.report_warning(f'Unable to download {thumb_display_id}: {err}') | |
4052 | if ret and not write_all: | |
4053 | break | |
4054 | return ret |