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