]>
Commit | Line | Data |
---|---|---|
d6983cb4 | 1 | import base64 |
234416e4 | 2 | import collections |
ac668111 | 3 | import getpass |
3ec05685 | 4 | import hashlib |
54007a45 | 5 | import http.client |
6 | import http.cookiejar | |
7 | import http.cookies | |
2314b4d8 | 8 | import inspect |
cc16383f | 9 | import itertools |
3d3538e4 | 10 | import json |
f8271158 | 11 | import math |
4094b6e3 | 12 | import netrc |
d6983cb4 | 13 | import os |
773f291d | 14 | import random |
6929b41a | 15 | import re |
d6983cb4 | 16 | import sys |
4094b6e3 | 17 | import time |
8f97a15d | 18 | import types |
14f25df2 | 19 | import urllib.parse |
ac668111 | 20 | import urllib.request |
f8271158 | 21 | import xml.etree.ElementTree |
d6983cb4 | 22 | |
6929b41a | 23 | from ..compat import functools # isort: split |
14f25df2 | 24 | from ..compat import compat_etree_fromstring, compat_expanduser, compat_os_name |
8817a80d | 25 | from ..cookies import LenientSimpleCookie |
f8271158 | 26 | from ..downloader.f4m import get_base_url, remove_encrypted_media |
8c25f81b | 27 | from ..utils import ( |
8f97a15d | 28 | IDENTITY, |
f8271158 | 29 | JSON_LD_RE, |
30 | NO_DEFAULT, | |
31 | ExtractorError, | |
d0d74b71 | 32 | FormatSorter, |
f8271158 | 33 | GeoRestrictedError, |
34 | GeoUtils, | |
b7c47b74 | 35 | LenientJSONDecoder, |
f8271158 | 36 | RegexNotFoundError, |
be5c1ae8 | 37 | RetryManager, |
f8271158 | 38 | UnsupportedError, |
05900629 | 39 | age_restricted, |
02dc0a36 | 40 | base_url, |
08f2a92c | 41 | bug_reports_message, |
82d02080 | 42 | classproperty, |
d6983cb4 | 43 | clean_html, |
d0d74b71 | 44 | deprecation_warning, |
70f0f5a8 | 45 | determine_ext, |
d493f15c | 46 | dict_get, |
42676437 | 47 | encode_data_uri, |
9b9c5355 | 48 | error_to_compat_str, |
46b18f23 | 49 | extract_attributes, |
90137ca4 | 50 | filter_dict, |
97f4aecf | 51 | fix_xml_ampersands, |
b14f3a4c | 52 | float_or_none, |
b868936c | 53 | format_field, |
31bb8d3f | 54 | int_or_none, |
34921b43 | 55 | join_nonempty, |
a4a554a7 | 56 | js_to_json, |
46b18f23 | 57 | mimetype2ext, |
3158150c | 58 | network_exceptions, |
46b18f23 | 59 | orderedSet, |
d493f15c | 60 | parse_bitrate, |
46b18f23 JH |
61 | parse_codecs, |
62 | parse_duration, | |
4ca2a3cf | 63 | parse_iso8601, |
46b18f23 | 64 | parse_m3u8_attributes, |
d493f15c | 65 | parse_resolution, |
46b18f23 | 66 | sanitize_filename, |
8f97a15d | 67 | sanitize_url, |
b868936c | 68 | sanitized_Request, |
ade1fa70 | 69 | smuggle_url, |
d493f15c | 70 | str_or_none, |
ce5b9040 | 71 | str_to_int, |
f856816b | 72 | strip_or_none, |
5d3a0e79 | 73 | traverse_obj, |
47046464 | 74 | try_call, |
ffa89477 | 75 | try_get, |
f38de77f | 76 | unescapeHTML, |
647eab45 | 77 | unified_strdate, |
6b3a3098 | 78 | unified_timestamp, |
46b18f23 | 79 | update_Request, |
09d02ea4 | 80 | update_url_query, |
a107193e | 81 | url_basename, |
bebef109 | 82 | url_or_none, |
b868936c | 83 | urljoin, |
6606817a | 84 | variadic, |
a6571f10 | 85 | xpath_element, |
8d6765cf S |
86 | xpath_text, |
87 | xpath_with_ns, | |
d6983cb4 | 88 | ) |
c342041f | 89 | |
d6983cb4 | 90 | |
86e5f3ed | 91 | class InfoExtractor: |
d6983cb4 PH |
92 | """Information Extractor class. |
93 | ||
94 | Information extractors are the classes that, given a URL, extract | |
95 | information about the video (or videos) the URL refers to. This | |
96 | information includes the real video URL, the video title, author and | |
97 | others. The information is stored in a dictionary which is then | |
5d380852 | 98 | passed to the YoutubeDL. The YoutubeDL processes this |
d6983cb4 PH |
99 | information possibly downloading the video to the file system, among |
100 | other possible outcomes. | |
101 | ||
cf0649f8 | 102 | The type field determines the type of the result. |
fed5d032 PH |
103 | By far the most common value (and the default if _type is missing) is |
104 | "video", which indicates a single video. | |
105 | ||
106 | For a video, the dictionaries must include the following fields: | |
d6983cb4 PH |
107 | |
108 | id: Video identifier. | |
d4736fdb | 109 | title: Video title, unescaped. Set to an empty string if video has |
110 | no title as opposed to "None" which signifies that the | |
111 | extractor failed to obtain a title | |
d67b0b15 | 112 | |
f49d89ee | 113 | Additionally, it must contain either a formats entry or a url one: |
d67b0b15 | 114 | |
f49d89ee PH |
115 | formats: A list of dictionaries for each format available, ordered |
116 | from worst to best quality. | |
117 | ||
118 | Potential fields: | |
c790e93a S |
119 | * url The mandatory URL representing the media: |
120 | for plain file media - HTTP URL of this file, | |
121 | for RTMP - RTMP URL, | |
122 | for HLS - URL of the M3U8 media playlist, | |
123 | for HDS - URL of the F4M manifest, | |
79d2077e S |
124 | for DASH |
125 | - HTTP URL to plain file media (in case of | |
126 | unfragmented media) | |
127 | - URL of the MPD manifest or base URL | |
128 | representing the media if MPD manifest | |
8ed7a233 | 129 | is parsed from a string (in case of |
79d2077e | 130 | fragmented media) |
c790e93a | 131 | for MSS - URL of the ISM manifest. |
86f4d14f S |
132 | * manifest_url |
133 | The URL of the manifest file in case of | |
c790e93a S |
134 | fragmented media: |
135 | for HLS - URL of the M3U8 master playlist, | |
136 | for HDS - URL of the F4M manifest, | |
137 | for DASH - URL of the MPD manifest, | |
138 | for MSS - URL of the ISM manifest. | |
a44ca5a4 | 139 | * manifest_stream_number (For internal use only) |
140 | The index of the stream in the manifest file | |
10952eb2 | 141 | * ext Will be calculated from URL if missing |
d67b0b15 PH |
142 | * format A human-readable description of the format |
143 | ("mp4 container with h264/opus"). | |
144 | Calculated from the format_id, width, height. | |
145 | and format_note fields if missing. | |
146 | * format_id A short description of the format | |
5d4f3985 PH |
147 | ("mp4_h264_opus" or "19"). |
148 | Technically optional, but strongly recommended. | |
d67b0b15 PH |
149 | * format_note Additional info about the format |
150 | ("3D" or "DASH video") | |
151 | * width Width of the video, if known | |
152 | * height Height of the video, if known | |
105bfd90 | 153 | * aspect_ratio Aspect ratio of the video, if known |
154 | Automatically calculated from width and height | |
f49d89ee | 155 | * resolution Textual description of width and height |
105bfd90 | 156 | Automatically calculated from width and height |
176f1866 | 157 | * dynamic_range The dynamic range of the video. One of: |
158 | "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV" | |
7217e148 | 159 | * tbr Average bitrate of audio and video in KBit/s |
d67b0b15 PH |
160 | * abr Average audio bitrate in KBit/s |
161 | * acodec Name of the audio codec in use | |
dd27fd17 | 162 | * asr Audio sampling rate in Hertz |
b8ed0f15 | 163 | * audio_channels Number of audio channels |
d67b0b15 | 164 | * vbr Average video bitrate in KBit/s |
fbb21cf5 | 165 | * fps Frame rate |
d67b0b15 | 166 | * vcodec Name of the video codec in use |
1394ce65 | 167 | * container Name of the container format |
d67b0b15 | 168 | * filesize The number of bytes, if known in advance |
9732d77e | 169 | * filesize_approx An estimate for the number of bytes |
d67b0b15 | 170 | * player_url SWF Player URL (used for rtmpdump). |
c7deaa4c | 171 | * protocol The protocol that will be used for the actual |
adbc4ec4 THD |
172 | download, lower-case. One of "http", "https" or |
173 | one of the protocols defined in downloader.PROTOCOL_MAP | |
c58c2d63 S |
174 | * fragment_base_url |
175 | Base URL for fragments. Each fragment's path | |
176 | value (if present) will be relative to | |
177 | this URL. | |
178 | * fragments A list of fragments of a fragmented media. | |
179 | Each fragment entry must contain either an url | |
180 | or a path. If an url is present it should be | |
181 | considered by a client. Otherwise both path and | |
182 | fragment_base_url must be present. Here is | |
183 | the list of all potential fields: | |
184 | * "url" - fragment's URL | |
185 | * "path" - fragment's path relative to | |
186 | fragment_base_url | |
a0d5077c S |
187 | * "duration" (optional, int or float) |
188 | * "filesize" (optional, int) | |
adbc4ec4 THD |
189 | * is_from_start Is a live format that can be downloaded |
190 | from the start. Boolean | |
f49d89ee | 191 | * preference Order number of this format. If this field is |
08d13955 | 192 | present and not None, the formats get sorted |
38d63d84 | 193 | by this field, regardless of all other values. |
f49d89ee PH |
194 | -1 for default (order by other properties), |
195 | -2 or smaller for less than default. | |
e65566a9 PH |
196 | < -1000 to hide the format (if there is |
197 | another one which is strictly better) | |
32f90364 PH |
198 | * language Language code, e.g. "de" or "en-US". |
199 | * language_preference Is this in the language mentioned in | |
200 | the URL? | |
aff2f4f4 PH |
201 | 10 if it's what the URL is about, |
202 | -1 for default (don't know), | |
203 | -10 otherwise, other values reserved for now. | |
5d73273f PH |
204 | * quality Order number of the video quality of this |
205 | format, irrespective of the file format. | |
206 | -1 for default (order by other properties), | |
207 | -2 or smaller for less than default. | |
c64ed2a3 PH |
208 | * source_preference Order number for this video source |
209 | (quality takes higher priority) | |
210 | -1 for default (order by other properties), | |
211 | -2 or smaller for less than default. | |
d769be6c PH |
212 | * http_headers A dictionary of additional HTTP headers |
213 | to add to the request. | |
6271f1ca | 214 | * stretched_ratio If given and not 1, indicates that the |
3dee7826 PH |
215 | video's pixels are not square. |
216 | width : height ratio as float. | |
217 | * no_resume The server does not support resuming the | |
218 | (HTTP or RTMP) download. Boolean. | |
88acdbc2 | 219 | * has_drm The format has DRM and cannot be downloaded. Boolean |
0a5a191a | 220 | * downloader_options A dictionary of downloader options |
221 | (For internal use only) | |
222 | * http_chunk_size Chunk size for HTTP downloads | |
223 | * ffmpeg_args Extra arguments for ffmpeg downloader | |
3b1fe47d | 224 | RTMP formats can also have the additional fields: page_url, |
225 | app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn, | |
226 | rtmp_protocol, rtmp_real_time | |
3dee7826 | 227 | |
c0ba0f48 | 228 | url: Final video URL. |
d6983cb4 | 229 | ext: Video filename extension. |
d67b0b15 PH |
230 | format: The video format, defaults to ext (used for --get-format) |
231 | player_url: SWF Player URL (used for rtmpdump). | |
2f5865cc | 232 | |
d6983cb4 PH |
233 | The following fields are optional: |
234 | ||
08d30158 | 235 | direct: True if a direct video file was given (must only be set by GenericIE) |
f5e43bc6 | 236 | alt_title: A secondary title of the video. |
0afef30b PH |
237 | display_id An alternative identifier for the video, not necessarily |
238 | unique, but available before title. Typically, id is | |
239 | something like "4234987", title "Dancing naked mole rats", | |
240 | and display_id "dancing-naked-mole-rats" | |
d5519808 | 241 | thumbnails: A list of dictionaries, with the following entries: |
cfb56d1a | 242 | * "id" (optional, string) - Thumbnail format ID |
d5519808 | 243 | * "url" |
cfb56d1a | 244 | * "preference" (optional, int) - quality of the image |
d5519808 PH |
245 | * "width" (optional, int) |
246 | * "height" (optional, int) | |
5e1c39ac | 247 | * "resolution" (optional, string "{width}x{height}", |
d5519808 | 248 | deprecated) |
2de624fd | 249 | * "filesize" (optional, int) |
297e9952 | 250 | * "http_headers" (dict) - HTTP headers for the request |
d6983cb4 | 251 | thumbnail: Full URL to a video thumbnail image. |
f5e43bc6 | 252 | description: Full video description. |
d6983cb4 | 253 | uploader: Full name of the video uploader. |
2bc0c46f | 254 | license: License name the video is licensed under. |
8a92e51c | 255 | creator: The creator of the video. |
10db0d2f | 256 | timestamp: UNIX timestamp of the moment the video was uploaded |
ae6a1b95 | 257 | upload_date: Video upload date in UTC (YYYYMMDD). |
f0d785d3 | 258 | If not explicitly set, calculated from timestamp |
259 | release_timestamp: UNIX timestamp of the moment the video was released. | |
260 | If it is not clear whether to use timestamp or this, use the former | |
ae6a1b95 | 261 | release_date: The date (YYYYMMDD) when the video was released in UTC. |
f0d785d3 | 262 | If not explicitly set, calculated from release_timestamp |
263 | modified_timestamp: UNIX timestamp of the moment the video was last modified. | |
ae6a1b95 | 264 | modified_date: The date (YYYYMMDD) when the video was last modified in UTC. |
f0d785d3 | 265 | If not explicitly set, calculated from modified_timestamp |
d6983cb4 | 266 | uploader_id: Nickname or id of the video uploader. |
7bcd2830 | 267 | uploader_url: Full URL to a personal webpage of the video uploader. |
6f1f59f3 | 268 | channel: Full name of the channel the video is uploaded on. |
0e7b8d3e | 269 | Note that channel fields may or may not repeat uploader |
6f1f59f3 S |
270 | fields. This depends on a particular extractor. |
271 | channel_id: Id of the channel. | |
272 | channel_url: Full URL to a channel webpage. | |
6c73052c | 273 | channel_follower_count: Number of followers of the channel. |
da9ec3b9 | 274 | location: Physical location where the video was filmed. |
a504ced0 | 275 | subtitles: The available subtitles as a dictionary in the format |
4606c34e YCH |
276 | {tag: subformats}. "tag" is usually a language code, and |
277 | "subformats" is a list sorted from lower to higher | |
278 | preference, each element is a dictionary with the "ext" | |
279 | entry and one of: | |
a504ced0 | 280 | * "data": The subtitles file contents |
10952eb2 | 281 | * "url": A URL pointing to the subtitles file |
2412044c | 282 | It can optionally also have: |
283 | * "name": Name or description of the subtitles | |
08d30158 | 284 | * "http_headers": A dictionary of additional HTTP headers |
297e9952 | 285 | to add to the request. |
4bba3716 | 286 | "ext" will be calculated from URL if missing |
e167860c | 287 | automatic_captions: Like 'subtitles'; contains automatically generated |
288 | captions instead of normal subtitles | |
62d231c0 | 289 | duration: Length of the video in seconds, as an integer or float. |
f3d29461 | 290 | view_count: How many users have watched the video on the platform. |
867c66ff | 291 | concurrent_view_count: How many users are currently watching the video on the platform. |
19e3dfc9 PH |
292 | like_count: Number of positive ratings of the video |
293 | dislike_count: Number of negative ratings of the video | |
02835c6b | 294 | repost_count: Number of reposts of the video |
2d30521a | 295 | average_rating: Average rating give by users, the scale used depends on the webpage |
19e3dfc9 | 296 | comment_count: Number of comments on the video |
dd622d7c PH |
297 | comments: A list of comments, each with one or more of the following |
298 | properties (all but one of text or html optional): | |
299 | * "author" - human-readable name of the comment author | |
300 | * "author_id" - user ID of the comment author | |
a1c5d2ca | 301 | * "author_thumbnail" - The thumbnail of the comment author |
dd622d7c PH |
302 | * "id" - Comment ID |
303 | * "html" - Comment as HTML | |
304 | * "text" - Plain text of the comment | |
305 | * "timestamp" - UNIX timestamp of comment | |
306 | * "parent" - ID of the comment this one is replying to. | |
307 | Set to "root" to indicate that this is a | |
308 | comment to the original video. | |
a1c5d2ca M |
309 | * "like_count" - Number of positive ratings of the comment |
310 | * "dislike_count" - Number of negative ratings of the comment | |
311 | * "is_favorited" - Whether the comment is marked as | |
312 | favorite by the video uploader | |
313 | * "author_is_uploader" - Whether the comment is made by | |
314 | the video uploader | |
8dbe9899 | 315 | age_limit: Age restriction for the video, as an integer (years) |
7a5c1cfe | 316 | webpage_url: The URL to the video webpage, if given to yt-dlp it |
9103bbc5 JMF |
317 | should allow to get the same result again. (It will be set |
318 | by YoutubeDL if it's missing) | |
ad3bc6ac PH |
319 | categories: A list of categories that the video falls in, for example |
320 | ["Sports", "Berlin"] | |
864f24bd | 321 | tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"] |
d0fb4bd1 | 322 | cast: A list of the video cast |
7267bd53 PH |
323 | is_live: True, False, or None (=unknown). Whether this video is a |
324 | live stream that goes on instead of a fixed-length video. | |
f76ede8e | 325 | was_live: True, False, or None (=unknown). Whether this video was |
326 | originally a live stream. | |
0647d925 | 327 | live_status: None (=unknown), 'is_live', 'is_upcoming', 'was_live', 'not_live', |
e325a21a | 328 | or 'post_live' (was live, but VOD is not yet processed) |
ae30b840 | 329 | If absent, automatically set from is_live, was_live |
7c80519c | 330 | start_time: Time in seconds where the reproduction should start, as |
10952eb2 | 331 | specified in the URL. |
297a564b | 332 | end_time: Time in seconds where the reproduction should end, as |
10952eb2 | 333 | specified in the URL. |
55949fed | 334 | chapters: A list of dictionaries, with the following entries: |
335 | * "start_time" - The start time of the chapter in seconds | |
336 | * "end_time" - The end time of the chapter in seconds | |
337 | * "title" (optional, string) | |
6cfda058 | 338 | playable_in_embed: Whether this video is allowed to play in embedded |
339 | players on other sites. Can be True (=always allowed), | |
340 | False (=never allowed), None (=unknown), or a string | |
62b58c09 | 341 | specifying the criteria for embedability; e.g. 'whitelist' |
c224251a M |
342 | availability: Under what condition the video is available. One of |
343 | 'private', 'premium_only', 'subscriber_only', 'needs_auth', | |
344 | 'unlisted' or 'public'. Use 'InfoExtractor._availability' | |
345 | to set it | |
1e8fe57e | 346 | _old_archive_ids: A list of old archive ids needed for backward compatibility |
784320c9 | 347 | _format_sort_fields: A list of fields to use for sorting formats |
277d6ff5 | 348 | __post_extractor: A function to be called just before the metadata is |
349 | written to either disk, logger or console. The function | |
350 | must return a dict which will be added to the info_dict. | |
351 | This is usefull for additional information that is | |
352 | time-consuming to extract. Note that the fields thus | |
353 | extracted will not be available to output template and | |
354 | match_filter. So, only "comments" and "comment_count" are | |
355 | currently allowed to be extracted via this method. | |
d6983cb4 | 356 | |
7109903e S |
357 | The following fields should only be used when the video belongs to some logical |
358 | chapter or section: | |
359 | ||
360 | chapter: Name or title of the chapter the video belongs to. | |
27bfd4e5 S |
361 | chapter_number: Number of the chapter the video belongs to, as an integer. |
362 | chapter_id: Id of the chapter the video belongs to, as a unicode string. | |
7109903e S |
363 | |
364 | The following fields should only be used when the video is an episode of some | |
8d76bdf1 | 365 | series, programme or podcast: |
7109903e S |
366 | |
367 | series: Title of the series or programme the video episode belongs to. | |
9ac24e23 | 368 | series_id: Id of the series or programme the video episode belongs to, as a unicode string. |
7109903e | 369 | season: Title of the season the video episode belongs to. |
27bfd4e5 S |
370 | season_number: Number of the season the video episode belongs to, as an integer. |
371 | season_id: Id of the season the video episode belongs to, as a unicode string. | |
7109903e S |
372 | episode: Title of the video episode. Unlike mandatory video title field, |
373 | this field should denote the exact title of the video episode | |
374 | without any kind of decoration. | |
27bfd4e5 S |
375 | episode_number: Number of the video episode within a season, as an integer. |
376 | episode_id: Id of the video episode, as a unicode string. | |
7109903e | 377 | |
7a93ab5f S |
378 | The following fields should only be used when the media is a track or a part of |
379 | a music album: | |
380 | ||
381 | track: Title of the track. | |
382 | track_number: Number of the track within an album or a disc, as an integer. | |
383 | track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii), | |
384 | as a unicode string. | |
385 | artist: Artist(s) of the track. | |
386 | genre: Genre(s) of the track. | |
387 | album: Title of the album the track belongs to. | |
388 | album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc). | |
389 | album_artist: List of all artists appeared on the album (e.g. | |
390 | "Ash Borer / Fell Voices" or "Various Artists", useful for splits | |
391 | and compilations). | |
392 | disc_number: Number of the disc or other physical medium the track belongs to, | |
393 | as an integer. | |
394 | release_year: Year (YYYY) when the album was released. | |
8bcd4048 | 395 | composer: Composer of the piece |
7a93ab5f | 396 | |
3975b4d2 | 397 | The following fields should only be set for clips that should be cut from the original video: |
398 | ||
399 | section_start: Start time of the section in seconds | |
400 | section_end: End time of the section in seconds | |
401 | ||
45e8a04e | 402 | The following fields should only be set for storyboards: |
403 | rows: Number of rows in each storyboard fragment, as an integer | |
404 | columns: Number of columns in each storyboard fragment, as an integer | |
405 | ||
deefc05b | 406 | Unless mentioned otherwise, the fields should be Unicode strings. |
d6983cb4 | 407 | |
d838b1bd PH |
408 | Unless mentioned otherwise, None is equivalent to absence of information. |
409 | ||
fed5d032 PH |
410 | |
411 | _type "playlist" indicates multiple videos. | |
b82f815f PH |
412 | There must be a key "entries", which is a list, an iterable, or a PagedList |
413 | object, each element of which is a valid dictionary by this specification. | |
fed5d032 | 414 | |
962ffcf8 | 415 | Additionally, playlists can have "id", "title", and any other relevant |
b60419c5 | 416 | attributes with the same semantics as videos (see above). |
fed5d032 | 417 | |
f0d785d3 | 418 | It can also have the following optional fields: |
419 | ||
420 | playlist_count: The total number of videos in a playlist. If not given, | |
421 | YoutubeDL tries to calculate it from "entries" | |
422 | ||
fed5d032 PH |
423 | |
424 | _type "multi_video" indicates that there are multiple videos that | |
425 | form a single show, for examples multiple acts of an opera or TV episode. | |
426 | It must have an entries key like a playlist and contain all the keys | |
427 | required for a video at the same time. | |
428 | ||
429 | ||
430 | _type "url" indicates that the video must be extracted from another | |
431 | location, possibly by a different extractor. Its only required key is: | |
432 | "url" - the next URL to extract. | |
f58766ce PH |
433 | The key "ie_key" can be set to the class name (minus the trailing "IE", |
434 | e.g. "Youtube") if the extractor class is known in advance. | |
435 | Additionally, the dictionary may have any properties of the resolved entity | |
436 | known in advance, for example "title" if the title of the referred video is | |
fed5d032 PH |
437 | known ahead of time. |
438 | ||
439 | ||
440 | _type "url_transparent" entities have the same specification as "url", but | |
441 | indicate that the given additional information is more precise than the one | |
442 | associated with the resolved URL. | |
443 | This is useful when a site employs a video service that hosts the video and | |
444 | its technical metadata, but that video service does not embed a useful | |
445 | title, description etc. | |
446 | ||
447 | ||
8f97a15d | 448 | Subclasses of this should also be added to the list of extractors and |
449 | should define a _VALID_URL regexp and, re-define the _real_extract() and | |
450 | (optionally) _real_initialize() methods. | |
d6983cb4 | 451 | |
e6f21b3d | 452 | Subclasses may also override suitable() if necessary, but ensure the function |
453 | signature is preserved and that this function imports everything it needs | |
52efa4b3 | 454 | (except other extractors), so that lazy_extractors works correctly. |
455 | ||
8f97a15d | 456 | Subclasses can define a list of _EMBED_REGEX, which will be searched for in |
457 | the HTML of Generic webpages. It may also override _extract_embed_urls | |
458 | or _extract_from_webpage as necessary. While these are normally classmethods, | |
459 | _extract_from_webpage is allowed to be an instance method. | |
460 | ||
461 | _extract_from_webpage may raise self.StopExtraction() to stop further | |
462 | processing of the webpage and obtain exclusive rights to it. This is useful | |
62b58c09 L |
463 | when the extractor cannot reliably be matched using just the URL, |
464 | e.g. invidious/peertube instances | |
8f97a15d | 465 | |
466 | Embed-only extractors can be defined by setting _VALID_URL = False. | |
467 | ||
52efa4b3 | 468 | To support username + password (or netrc) login, the extractor must define a |
469 | _NETRC_MACHINE and re-define _perform_login(username, password) and | |
470 | (optionally) _initialize_pre_login() methods. The _perform_login method will | |
471 | be called between _initialize_pre_login and _real_initialize if credentials | |
472 | are passed by the user. In cases where it is necessary to have the login | |
473 | process as part of the extraction rather than initialization, _perform_login | |
474 | can be left undefined. | |
e6f21b3d | 475 | |
4248dad9 | 476 | _GEO_BYPASS attribute may be set to False in order to disable |
773f291d S |
477 | geo restriction bypass mechanisms for a particular extractor. |
478 | Though it won't disable explicit geo restriction bypass based on | |
504f20dd | 479 | country code provided with geo_bypass_country. |
4248dad9 S |
480 | |
481 | _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted | |
482 | countries for this extractor. One of these countries will be used by | |
483 | geo restriction bypass mechanism right away in order to bypass | |
504f20dd | 484 | geo restriction, of course, if the mechanism is not disabled. |
773f291d | 485 | |
5f95927a S |
486 | _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted |
487 | IP blocks in CIDR notation for this extractor. One of these IP blocks | |
488 | will be used by geo restriction bypass mechanism similarly | |
504f20dd | 489 | to _GEO_COUNTRIES. |
3ccdde8c | 490 | |
fe7866d0 | 491 | The _ENABLED attribute should be set to False for IEs that |
492 | are disabled by default and must be explicitly enabled. | |
493 | ||
e6f21b3d | 494 | The _WORKING attribute should be set to False for broken IEs |
d6983cb4 PH |
495 | in order to warn the users and skip the tests. |
496 | """ | |
497 | ||
498 | _ready = False | |
499 | _downloader = None | |
773f291d | 500 | _x_forwarded_for_ip = None |
4248dad9 S |
501 | _GEO_BYPASS = True |
502 | _GEO_COUNTRIES = None | |
5f95927a | 503 | _GEO_IP_BLOCKS = None |
d6983cb4 | 504 | _WORKING = True |
fe7866d0 | 505 | _ENABLED = True |
52efa4b3 | 506 | _NETRC_MACHINE = None |
231025c4 | 507 | IE_DESC = None |
8dcce6a8 | 508 | SEARCH_KEY = None |
8f97a15d | 509 | _VALID_URL = None |
510 | _EMBED_REGEX = [] | |
d6983cb4 | 511 | |
8dcce6a8 | 512 | def _login_hint(self, method=NO_DEFAULT, netrc=None): |
513 | password_hint = f'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials' | |
514 | return { | |
515 | None: '', | |
516 | 'any': f'Use --cookies, --cookies-from-browser, {password_hint}', | |
517 | 'password': f'Use {password_hint}', | |
518 | 'cookies': ( | |
519 | 'Use --cookies-from-browser or --cookies for the authentication. ' | |
17ffed18 | 520 | 'See https://github.com/yt-dlp/yt-dlp/wiki/FAQ#how-do-i-pass-cookies-to-yt-dlp for how to manually pass cookies'), |
8dcce6a8 | 521 | }[method if method is not NO_DEFAULT else 'any' if self.supports_login() else 'cookies'] |
9d5d4d64 | 522 | |
d6983cb4 | 523 | def __init__(self, downloader=None): |
49a57e70 | 524 | """Constructor. Receives an optional downloader (a YoutubeDL instance). |
525 | If a downloader is not passed during initialization, | |
526 | it must be set using "set_downloader()" before "extract()" is called""" | |
d6983cb4 | 527 | self._ready = False |
773f291d | 528 | self._x_forwarded_for_ip = None |
28f436ba | 529 | self._printed_messages = set() |
d6983cb4 PH |
530 | self.set_downloader(downloader) |
531 | ||
532 | @classmethod | |
5ad28e7f | 533 | def _match_valid_url(cls, url): |
8f97a15d | 534 | if cls._VALID_URL is False: |
535 | return None | |
79cb2577 PH |
536 | # This does not use has/getattr intentionally - we want to know whether |
537 | # we have cached the regexp for *this* class, whereas getattr would also | |
538 | # match the superclass | |
539 | if '_VALID_URL_RE' not in cls.__dict__: | |
540 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) | |
5ad28e7f | 541 | return cls._VALID_URL_RE.match(url) |
542 | ||
543 | @classmethod | |
544 | def suitable(cls, url): | |
545 | """Receives a URL and returns True if suitable for this IE.""" | |
3fb4e21b | 546 | # This function must import everything it needs (except other extractors), |
547 | # so that lazy_extractors works correctly | |
5ad28e7f | 548 | return cls._match_valid_url(url) is not None |
d6983cb4 | 549 | |
ed9266db PH |
550 | @classmethod |
551 | def _match_id(cls, url): | |
5ad28e7f | 552 | return cls._match_valid_url(url).group('id') |
ed9266db | 553 | |
1151c407 | 554 | @classmethod |
555 | def get_temp_id(cls, url): | |
556 | try: | |
557 | return cls._match_id(url) | |
558 | except (IndexError, AttributeError): | |
559 | return None | |
560 | ||
d6983cb4 PH |
561 | @classmethod |
562 | def working(cls): | |
563 | """Getter method for _WORKING.""" | |
564 | return cls._WORKING | |
565 | ||
52efa4b3 | 566 | @classmethod |
567 | def supports_login(cls): | |
568 | return bool(cls._NETRC_MACHINE) | |
569 | ||
d6983cb4 PH |
570 | def initialize(self): |
571 | """Initializes an instance (authentication, etc).""" | |
28f436ba | 572 | self._printed_messages = set() |
5f95927a S |
573 | self._initialize_geo_bypass({ |
574 | 'countries': self._GEO_COUNTRIES, | |
575 | 'ip_blocks': self._GEO_IP_BLOCKS, | |
576 | }) | |
4248dad9 | 577 | if not self._ready: |
52efa4b3 | 578 | self._initialize_pre_login() |
579 | if self.supports_login(): | |
580 | username, password = self._get_login_info() | |
581 | if username: | |
582 | self._perform_login(username, password) | |
583 | elif self.get_param('username') and False not in (self.IE_DESC, self._NETRC_MACHINE): | |
8dcce6a8 | 584 | self.report_warning(f'Login with password is not supported for this website. {self._login_hint("cookies")}') |
4248dad9 S |
585 | self._real_initialize() |
586 | self._ready = True | |
587 | ||
5f95927a | 588 | def _initialize_geo_bypass(self, geo_bypass_context): |
e39b5d4a S |
589 | """ |
590 | Initialize geo restriction bypass mechanism. | |
591 | ||
592 | This method is used to initialize geo bypass mechanism based on faking | |
593 | X-Forwarded-For HTTP header. A random country from provided country list | |
dc0a869e | 594 | is selected and a random IP belonging to this country is generated. This |
e39b5d4a S |
595 | IP will be passed as X-Forwarded-For HTTP header in all subsequent |
596 | HTTP requests. | |
e39b5d4a S |
597 | |
598 | This method will be used for initial geo bypass mechanism initialization | |
5f95927a S |
599 | during the instance initialization with _GEO_COUNTRIES and |
600 | _GEO_IP_BLOCKS. | |
e39b5d4a | 601 | |
5f95927a | 602 | You may also manually call it from extractor's code if geo bypass |
e39b5d4a | 603 | information is not available beforehand (e.g. obtained during |
5f95927a S |
604 | extraction) or due to some other reason. In this case you should pass |
605 | this information in geo bypass context passed as first argument. It may | |
606 | contain following fields: | |
607 | ||
608 | countries: List of geo unrestricted countries (similar | |
609 | to _GEO_COUNTRIES) | |
610 | ip_blocks: List of geo unrestricted IP blocks in CIDR notation | |
611 | (similar to _GEO_IP_BLOCKS) | |
612 | ||
e39b5d4a | 613 | """ |
773f291d | 614 | if not self._x_forwarded_for_ip: |
5f95927a S |
615 | |
616 | # Geo bypass mechanism is explicitly disabled by user | |
a06916d9 | 617 | if not self.get_param('geo_bypass', True): |
5f95927a S |
618 | return |
619 | ||
620 | if not geo_bypass_context: | |
621 | geo_bypass_context = {} | |
622 | ||
623 | # Backward compatibility: previously _initialize_geo_bypass | |
624 | # expected a list of countries, some 3rd party code may still use | |
625 | # it this way | |
626 | if isinstance(geo_bypass_context, (list, tuple)): | |
627 | geo_bypass_context = { | |
628 | 'countries': geo_bypass_context, | |
629 | } | |
630 | ||
631 | # The whole point of geo bypass mechanism is to fake IP | |
632 | # as X-Forwarded-For HTTP header based on some IP block or | |
633 | # country code. | |
634 | ||
635 | # Path 1: bypassing based on IP block in CIDR notation | |
636 | ||
637 | # Explicit IP block specified by user, use it right away | |
638 | # regardless of whether extractor is geo bypassable or not | |
a06916d9 | 639 | ip_block = self.get_param('geo_bypass_ip_block', None) |
5f95927a S |
640 | |
641 | # Otherwise use random IP block from geo bypass context but only | |
642 | # if extractor is known as geo bypassable | |
643 | if not ip_block: | |
644 | ip_blocks = geo_bypass_context.get('ip_blocks') | |
645 | if self._GEO_BYPASS and ip_blocks: | |
646 | ip_block = random.choice(ip_blocks) | |
647 | ||
648 | if ip_block: | |
649 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(ip_block) | |
8a82af35 | 650 | self.write_debug(f'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For') |
5f95927a S |
651 | return |
652 | ||
653 | # Path 2: bypassing based on country code | |
654 | ||
655 | # Explicit country code specified by user, use it right away | |
656 | # regardless of whether extractor is geo bypassable or not | |
a06916d9 | 657 | country = self.get_param('geo_bypass_country', None) |
5f95927a S |
658 | |
659 | # Otherwise use random country code from geo bypass context but | |
660 | # only if extractor is known as geo bypassable | |
661 | if not country: | |
662 | countries = geo_bypass_context.get('countries') | |
663 | if self._GEO_BYPASS and countries: | |
664 | country = random.choice(countries) | |
665 | ||
666 | if country: | |
667 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country) | |
0760b0a7 | 668 | self._downloader.write_debug( |
86e5f3ed | 669 | f'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For') |
d6983cb4 PH |
670 | |
671 | def extract(self, url): | |
672 | """Extracts URL information and returns it in list of dicts.""" | |
3a5bcd03 | 673 | try: |
773f291d S |
674 | for _ in range(2): |
675 | try: | |
676 | self.initialize() | |
a06916d9 | 677 | self.write_debug('Extracting URL: %s' % url) |
0016b84e | 678 | ie_result = self._real_extract(url) |
07cce701 | 679 | if ie_result is None: |
680 | return None | |
0016b84e S |
681 | if self._x_forwarded_for_ip: |
682 | ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip | |
b79f9e30 | 683 | subtitles = ie_result.get('subtitles') or {} |
684 | if 'no-live-chat' in self.get_param('compat_opts'): | |
685 | for lang in ('live_chat', 'comments', 'danmaku'): | |
686 | subtitles.pop(lang, None) | |
0016b84e | 687 | return ie_result |
773f291d | 688 | except GeoRestrictedError as e: |
4248dad9 S |
689 | if self.__maybe_fake_ip_and_retry(e.countries): |
690 | continue | |
773f291d | 691 | raise |
0db3bae8 | 692 | except UnsupportedError: |
693 | raise | |
1151c407 | 694 | except ExtractorError as e: |
0db3bae8 | 695 | kwargs = { |
696 | 'video_id': e.video_id or self.get_temp_id(url), | |
697 | 'ie': self.IE_NAME, | |
b69fd25c | 698 | 'tb': e.traceback or sys.exc_info()[2], |
0db3bae8 | 699 | 'expected': e.expected, |
700 | 'cause': e.cause | |
701 | } | |
702 | if hasattr(e, 'countries'): | |
703 | kwargs['countries'] = e.countries | |
7265a219 | 704 | raise type(e)(e.orig_msg, **kwargs) |
ac668111 | 705 | except http.client.IncompleteRead as e: |
1151c407 | 706 | raise ExtractorError('A network error has occurred.', cause=e, expected=True, video_id=self.get_temp_id(url)) |
9650885b | 707 | except (KeyError, StopIteration) as e: |
1151c407 | 708 | raise ExtractorError('An extractor error has occurred.', cause=e, video_id=self.get_temp_id(url)) |
d6983cb4 | 709 | |
4248dad9 | 710 | def __maybe_fake_ip_and_retry(self, countries): |
a06916d9 | 711 | if (not self.get_param('geo_bypass_country', None) |
3089bc74 | 712 | and self._GEO_BYPASS |
a06916d9 | 713 | and self.get_param('geo_bypass', True) |
3089bc74 S |
714 | and not self._x_forwarded_for_ip |
715 | and countries): | |
eea0716c S |
716 | country_code = random.choice(countries) |
717 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) | |
4248dad9 S |
718 | if self._x_forwarded_for_ip: |
719 | self.report_warning( | |
eea0716c S |
720 | 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.' |
721 | % (self._x_forwarded_for_ip, country_code.upper())) | |
4248dad9 S |
722 | return True |
723 | return False | |
724 | ||
d6983cb4 | 725 | def set_downloader(self, downloader): |
08d30158 | 726 | """Sets a YoutubeDL instance as the downloader for this IE.""" |
d6983cb4 PH |
727 | self._downloader = downloader |
728 | ||
9809740b | 729 | @property |
730 | def cache(self): | |
731 | return self._downloader.cache | |
732 | ||
733 | @property | |
734 | def cookiejar(self): | |
735 | return self._downloader.cookiejar | |
736 | ||
52efa4b3 | 737 | def _initialize_pre_login(self): |
962ffcf8 | 738 | """ Initialization before login. Redefine in subclasses.""" |
52efa4b3 | 739 | pass |
740 | ||
741 | def _perform_login(self, username, password): | |
742 | """ Login with username and password. Redefine in subclasses.""" | |
743 | pass | |
744 | ||
d6983cb4 PH |
745 | def _real_initialize(self): |
746 | """Real initialization process. Redefine in subclasses.""" | |
747 | pass | |
748 | ||
749 | def _real_extract(self, url): | |
750 | """Real extraction process. Redefine in subclasses.""" | |
08d30158 | 751 | raise NotImplementedError('This method must be implemented by subclasses') |
d6983cb4 | 752 | |
56c73665 JMF |
753 | @classmethod |
754 | def ie_key(cls): | |
755 | """A string for getting the InfoExtractor with get_info_extractor""" | |
3fb4e21b | 756 | return cls.__name__[:-2] |
56c73665 | 757 | |
82d02080 | 758 | @classproperty |
759 | def IE_NAME(cls): | |
760 | return cls.__name__[:-2] | |
d6983cb4 | 761 | |
d391b7e2 S |
762 | @staticmethod |
763 | def __can_accept_status_code(err, expected_status): | |
ac668111 | 764 | assert isinstance(err, urllib.error.HTTPError) |
d391b7e2 S |
765 | if expected_status is None: |
766 | return False | |
d391b7e2 S |
767 | elif callable(expected_status): |
768 | return expected_status(err.code) is True | |
769 | else: | |
6606817a | 770 | return err.code in variadic(expected_status) |
d391b7e2 | 771 | |
c043c246 | 772 | def _create_request(self, url_or_request, data=None, headers=None, query=None): |
ac668111 | 773 | if isinstance(url_or_request, urllib.request.Request): |
09d02ea4 | 774 | return update_Request(url_or_request, data=data, headers=headers, query=query) |
775 | if query: | |
776 | url_or_request = update_url_query(url_or_request, query) | |
c043c246 | 777 | return sanitized_Request(url_or_request, data, headers or {}) |
f95b9dee | 778 | |
c043c246 | 779 | def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers=None, query=None, expected_status=None): |
d391b7e2 S |
780 | """ |
781 | Return the response handle. | |
782 | ||
783 | See _download_webpage docstring for arguments specification. | |
784 | """ | |
1cf376f5 | 785 | if not self._downloader._first_webpage_request: |
49a57e70 | 786 | sleep_interval = self.get_param('sleep_interval_requests') or 0 |
1cf376f5 | 787 | if sleep_interval > 0: |
5ef7d9bd | 788 | self.to_screen('Sleeping %s seconds ...' % sleep_interval) |
1cf376f5 | 789 | time.sleep(sleep_interval) |
790 | else: | |
791 | self._downloader._first_webpage_request = False | |
792 | ||
d6983cb4 PH |
793 | if note is None: |
794 | self.report_download_webpage(video_id) | |
795 | elif note is not False: | |
7cc3570e | 796 | if video_id is None: |
86e5f3ed | 797 | self.to_screen(str(note)) |
7cc3570e | 798 | else: |
86e5f3ed | 799 | self.to_screen(f'{video_id}: {note}') |
2132edaa S |
800 | |
801 | # Some sites check X-Forwarded-For HTTP header in order to figure out | |
802 | # the origin of the client behind proxy. This allows bypassing geo | |
803 | # restriction by faking this header's value to IP that belongs to some | |
804 | # geo unrestricted country. We will do so once we encounter any | |
805 | # geo restriction error. | |
806 | if self._x_forwarded_for_ip: | |
c043c246 | 807 | headers = (headers or {}).copy() |
808 | headers.setdefault('X-Forwarded-For', self._x_forwarded_for_ip) | |
2132edaa | 809 | |
d6983cb4 | 810 | try: |
f95b9dee | 811 | return self._downloader.urlopen(self._create_request(url_or_request, data, headers, query)) |
3158150c | 812 | except network_exceptions as err: |
ac668111 | 813 | if isinstance(err, urllib.error.HTTPError): |
d391b7e2 | 814 | if self.__can_accept_status_code(err, expected_status): |
95e42d73 XDG |
815 | # Retain reference to error to prevent file object from |
816 | # being closed before it can be read. Works around the | |
817 | # effects of <https://bugs.python.org/issue15002> | |
818 | # introduced in Python 3.4.1. | |
819 | err.fp._error = err | |
d391b7e2 S |
820 | return err.fp |
821 | ||
aa94a6d3 PH |
822 | if errnote is False: |
823 | return False | |
d6983cb4 | 824 | if errnote is None: |
f1a9d64e | 825 | errnote = 'Unable to download webpage' |
7f8b2714 | 826 | |
86e5f3ed | 827 | errmsg = f'{errnote}: {error_to_compat_str(err)}' |
7cc3570e | 828 | if fatal: |
497d2fab | 829 | raise ExtractorError(errmsg, cause=err) |
7cc3570e | 830 | else: |
6a39ee13 | 831 | self.report_warning(errmsg) |
7cc3570e | 832 | return False |
d6983cb4 | 833 | |
1890fc63 | 834 | def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, |
835 | encoding=None, data=None, headers={}, query={}, expected_status=None): | |
d391b7e2 S |
836 | """ |
837 | Return a tuple (page content as string, URL handle). | |
838 | ||
617f658b | 839 | Arguments: |
840 | url_or_request -- plain text URL as a string or | |
ac668111 | 841 | a urllib.request.Request object |
617f658b | 842 | video_id -- Video/playlist/item identifier (string) |
843 | ||
844 | Keyword arguments: | |
845 | note -- note printed before downloading (string) | |
846 | errnote -- note printed in case of an error (string) | |
847 | fatal -- flag denoting whether error should be considered fatal, | |
848 | i.e. whether it should cause ExtractionError to be raised, | |
849 | otherwise a warning will be reported and extraction continued | |
850 | encoding -- encoding for a page content decoding, guessed automatically | |
851 | when not explicitly specified | |
852 | data -- POST data (bytes) | |
853 | headers -- HTTP headers (dict) | |
854 | query -- URL query (dict) | |
855 | expected_status -- allows to accept failed HTTP requests (non 2xx | |
856 | status code) by explicitly specifying a set of accepted status | |
857 | codes. Can be any of the following entities: | |
858 | - an integer type specifying an exact failed status code to | |
859 | accept | |
860 | - a list or a tuple of integer types specifying a list of | |
861 | failed status codes to accept | |
862 | - a callable accepting an actual failed status code and | |
863 | returning True if it should be accepted | |
864 | Note that this argument does not affect success status codes (2xx) | |
865 | which are always accepted. | |
d391b7e2 | 866 | """ |
617f658b | 867 | |
b9d3e163 | 868 | # Strip hashes from the URL (#1038) |
14f25df2 | 869 | if isinstance(url_or_request, str): |
b9d3e163 PH |
870 | url_or_request = url_or_request.partition('#')[0] |
871 | ||
d391b7e2 | 872 | urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query, expected_status=expected_status) |
7cc3570e PH |
873 | if urlh is False: |
874 | assert not fatal | |
875 | return False | |
c9a77969 | 876 | content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding) |
23be51d8 PH |
877 | return (content, urlh) |
878 | ||
c9a77969 YCH |
879 | @staticmethod |
880 | def _guess_encoding_from_content(content_type, webpage_bytes): | |
d6983cb4 PH |
881 | m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type) |
882 | if m: | |
883 | encoding = m.group(1) | |
884 | else: | |
0d75ae2c | 885 | m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]', |
f143d86a PH |
886 | webpage_bytes[:1024]) |
887 | if m: | |
888 | encoding = m.group(1).decode('ascii') | |
b60016e8 PH |
889 | elif webpage_bytes.startswith(b'\xff\xfe'): |
890 | encoding = 'utf-16' | |
f143d86a PH |
891 | else: |
892 | encoding = 'utf-8' | |
c9a77969 YCH |
893 | |
894 | return encoding | |
895 | ||
4457823d S |
896 | def __check_blocked(self, content): |
897 | first_block = content[:512] | |
3089bc74 S |
898 | if ('<title>Access to this site is blocked</title>' in content |
899 | and 'Websense' in first_block): | |
4457823d S |
900 | msg = 'Access to this webpage has been blocked by Websense filtering software in your network.' |
901 | blocked_iframe = self._html_search_regex( | |
902 | r'<iframe src="([^"]+)"', content, | |
903 | 'Websense information URL', default=None) | |
904 | if blocked_iframe: | |
905 | msg += ' Visit %s for more details' % blocked_iframe | |
906 | raise ExtractorError(msg, expected=True) | |
907 | if '<title>The URL you requested has been blocked</title>' in first_block: | |
908 | msg = ( | |
909 | 'Access to this webpage has been blocked by Indian censorship. ' | |
910 | 'Use a VPN or proxy server (with --proxy) to route around it.') | |
911 | block_msg = self._html_search_regex( | |
912 | r'</h1><p>(.*?)</p>', | |
913 | content, 'block message', default=None) | |
914 | if block_msg: | |
915 | msg += ' (Message: "%s")' % block_msg.replace('\n', ' ') | |
916 | raise ExtractorError(msg, expected=True) | |
3089bc74 S |
917 | if ('<title>TTK :: Доступ к ресурсу ограничен</title>' in content |
918 | and 'blocklist.rkn.gov.ru' in content): | |
4457823d S |
919 | raise ExtractorError( |
920 | 'Access to this webpage has been blocked by decision of the Russian government. ' | |
921 | 'Visit http://blocklist.rkn.gov.ru/ for a block reason.', | |
922 | expected=True) | |
923 | ||
f95b9dee | 924 | def _request_dump_filename(self, url, video_id): |
925 | basen = f'{video_id}_{url}' | |
926 | trim_length = self.get_param('trim_file_name') or 240 | |
927 | if len(basen) > trim_length: | |
928 | h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest() | |
929 | basen = basen[:trim_length - len(h)] + h | |
930 | filename = sanitize_filename(f'{basen}.dump', restricted=True) | |
931 | # Working around MAX_PATH limitation on Windows (see | |
932 | # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx) | |
933 | if compat_os_name == 'nt': | |
934 | absfilepath = os.path.abspath(filename) | |
935 | if len(absfilepath) > 259: | |
936 | filename = fR'\\?\{absfilepath}' | |
937 | return filename | |
938 | ||
939 | def __decode_webpage(self, webpage_bytes, encoding, headers): | |
940 | if not encoding: | |
941 | encoding = self._guess_encoding_from_content(headers.get('Content-Type', ''), webpage_bytes) | |
942 | try: | |
943 | return webpage_bytes.decode(encoding, 'replace') | |
944 | except LookupError: | |
945 | return webpage_bytes.decode('utf-8', 'replace') | |
946 | ||
c9a77969 | 947 | def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None): |
c9a77969 YCH |
948 | webpage_bytes = urlh.read() |
949 | if prefix is not None: | |
950 | webpage_bytes = prefix + webpage_bytes | |
a06916d9 | 951 | if self.get_param('dump_intermediate_pages', False): |
f610dbb0 | 952 | self.to_screen('Dumping request to ' + urlh.geturl()) |
d6983cb4 PH |
953 | dump = base64.b64encode(webpage_bytes).decode('ascii') |
954 | self._downloader.to_screen(dump) | |
f95b9dee | 955 | if self.get_param('write_pages'): |
e121e3ce | 956 | filename = self._request_dump_filename(urlh.geturl(), video_id) |
f95b9dee | 957 | self.to_screen(f'Saving request to {filename}') |
d41e6efc PH |
958 | with open(filename, 'wb') as outf: |
959 | outf.write(webpage_bytes) | |
960 | ||
f95b9dee | 961 | content = self.__decode_webpage(webpage_bytes, encoding, urlh.headers) |
4457823d | 962 | self.__check_blocked(content) |
2410c43d | 963 | |
23be51d8 | 964 | return content |
d6983cb4 | 965 | |
6edf2808 | 966 | def __print_error(self, errnote, fatal, video_id, err): |
967 | if fatal: | |
c6e07cf1 | 968 | raise ExtractorError(f'{video_id}: {errnote}', cause=err) |
6edf2808 | 969 | elif errnote: |
c6e07cf1 | 970 | self.report_warning(f'{video_id}: {errnote}: {err}') |
6edf2808 | 971 | |
972 | def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True, errnote=None): | |
e2b38da9 PH |
973 | if transform_source: |
974 | xml_string = transform_source(xml_string) | |
e01c3d2e S |
975 | try: |
976 | return compat_etree_fromstring(xml_string.encode('utf-8')) | |
f9934b96 | 977 | except xml.etree.ElementTree.ParseError as ve: |
6edf2808 | 978 | self.__print_error('Failed to parse XML' if errnote is None else errnote, fatal, video_id, ve) |
267ed0c5 | 979 | |
6edf2808 | 980 | def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, errnote=None, **parser_kwargs): |
3d3538e4 | 981 | try: |
b7c47b74 | 982 | return json.loads( |
983 | json_string, cls=LenientJSONDecoder, strict=False, transform_source=transform_source, **parser_kwargs) | |
3d3538e4 | 984 | except ValueError as ve: |
6edf2808 | 985 | self.__print_error('Failed to parse JSON' if errnote is None else errnote, fatal, video_id, ve) |
3d3538e4 | 986 | |
6edf2808 | 987 | def _parse_socket_response_as_json(self, data, *args, **kwargs): |
988 | return self._parse_json(data[data.find('{'):data.rfind('}') + 1], *args, **kwargs) | |
adddc50c | 989 | |
617f658b | 990 | def __create_download_methods(name, parser, note, errnote, return_value): |
991 | ||
6edf2808 | 992 | def parse(ie, content, *args, errnote=errnote, **kwargs): |
617f658b | 993 | if parser is None: |
994 | return content | |
6edf2808 | 995 | if errnote is False: |
996 | kwargs['errnote'] = errnote | |
617f658b | 997 | # parser is fetched by name so subclasses can override it |
998 | return getattr(ie, parser)(content, *args, **kwargs) | |
999 | ||
c4910024 | 1000 | def download_handle(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None, |
1001 | fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None): | |
1002 | res = self._download_webpage_handle( | |
1003 | url_or_request, video_id, note=note, errnote=errnote, fatal=fatal, encoding=encoding, | |
1004 | data=data, headers=headers, query=query, expected_status=expected_status) | |
617f658b | 1005 | if res is False: |
1006 | return res | |
1007 | content, urlh = res | |
6edf2808 | 1008 | return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote), urlh |
617f658b | 1009 | |
f95b9dee | 1010 | def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None, |
c4910024 | 1011 | fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None): |
f95b9dee | 1012 | if self.get_param('load_pages'): |
1013 | url_or_request = self._create_request(url_or_request, data, headers, query) | |
1014 | filename = self._request_dump_filename(url_or_request.full_url, video_id) | |
1015 | self.to_screen(f'Loading request from {filename}') | |
1016 | try: | |
1017 | with open(filename, 'rb') as dumpf: | |
1018 | webpage_bytes = dumpf.read() | |
1019 | except OSError as e: | |
1020 | self.report_warning(f'Unable to load request from disk: {e}') | |
1021 | else: | |
1022 | content = self.__decode_webpage(webpage_bytes, encoding, url_or_request.headers) | |
6edf2808 | 1023 | return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote) |
c4910024 | 1024 | kwargs = { |
1025 | 'note': note, | |
1026 | 'errnote': errnote, | |
1027 | 'transform_source': transform_source, | |
1028 | 'fatal': fatal, | |
1029 | 'encoding': encoding, | |
1030 | 'data': data, | |
1031 | 'headers': headers, | |
1032 | 'query': query, | |
1033 | 'expected_status': expected_status, | |
1034 | } | |
617f658b | 1035 | if parser is None: |
c4910024 | 1036 | kwargs.pop('transform_source') |
617f658b | 1037 | # The method is fetched by name so subclasses can override _download_..._handle |
c4910024 | 1038 | res = getattr(self, download_handle.__name__)(url_or_request, video_id, **kwargs) |
617f658b | 1039 | return res if res is False else res[0] |
1040 | ||
1041 | def impersonate(func, name, return_value): | |
1042 | func.__name__, func.__qualname__ = name, f'InfoExtractor.{name}' | |
1043 | func.__doc__ = f''' | |
1044 | @param transform_source Apply this transformation before parsing | |
1045 | @returns {return_value} | |
1046 | ||
1047 | See _download_webpage_handle docstring for other arguments specification | |
1048 | ''' | |
1049 | ||
1050 | impersonate(download_handle, f'_download_{name}_handle', f'({return_value}, URL handle)') | |
1051 | impersonate(download_content, f'_download_{name}', f'{return_value}') | |
1052 | return download_handle, download_content | |
1053 | ||
1054 | _download_xml_handle, _download_xml = __create_download_methods( | |
1055 | 'xml', '_parse_xml', 'Downloading XML', 'Unable to download XML', 'xml as an xml.etree.ElementTree.Element') | |
1056 | _download_json_handle, _download_json = __create_download_methods( | |
1057 | 'json', '_parse_json', 'Downloading JSON metadata', 'Unable to download JSON metadata', 'JSON object as a dict') | |
1058 | _download_socket_json_handle, _download_socket_json = __create_download_methods( | |
1059 | 'socket_json', '_parse_socket_response_as_json', 'Polling socket', 'Unable to poll socket', 'JSON object as a dict') | |
1060 | __download_webpage = __create_download_methods('webpage', None, None, None, 'data of the page as a string')[1] | |
adddc50c | 1061 | |
617f658b | 1062 | def _download_webpage( |
1063 | self, url_or_request, video_id, note=None, errnote=None, | |
1064 | fatal=True, tries=1, timeout=NO_DEFAULT, *args, **kwargs): | |
adddc50c | 1065 | """ |
617f658b | 1066 | Return the data of the page as a string. |
adddc50c | 1067 | |
617f658b | 1068 | Keyword arguments: |
1069 | tries -- number of tries | |
1070 | timeout -- sleep interval between tries | |
1071 | ||
1072 | See _download_webpage_handle docstring for other arguments specification. | |
adddc50c | 1073 | """ |
617f658b | 1074 | |
1075 | R''' # NB: These are unused; should they be deprecated? | |
1076 | if tries != 1: | |
1077 | self._downloader.deprecation_warning('tries argument is deprecated in InfoExtractor._download_webpage') | |
1078 | if timeout is NO_DEFAULT: | |
1079 | timeout = 5 | |
1080 | else: | |
1081 | self._downloader.deprecation_warning('timeout argument is deprecated in InfoExtractor._download_webpage') | |
1082 | ''' | |
1083 | ||
1084 | try_count = 0 | |
1085 | while True: | |
1086 | try: | |
1087 | return self.__download_webpage(url_or_request, video_id, note, errnote, None, fatal, *args, **kwargs) | |
ac668111 | 1088 | except http.client.IncompleteRead as e: |
617f658b | 1089 | try_count += 1 |
1090 | if try_count >= tries: | |
1091 | raise e | |
1092 | self._sleep(timeout, video_id) | |
adddc50c | 1093 | |
28f436ba | 1094 | def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs): |
a70635b8 | 1095 | idstr = format_field(video_id, None, '%s: ') |
28f436ba | 1096 | msg = f'[{self.IE_NAME}] {idstr}{msg}' |
1097 | if only_once: | |
1098 | if f'WARNING: {msg}' in self._printed_messages: | |
1099 | return | |
1100 | self._printed_messages.add(f'WARNING: {msg}') | |
1101 | self._downloader.report_warning(msg, *args, **kwargs) | |
f45f96f8 | 1102 | |
a06916d9 | 1103 | def to_screen(self, msg, *args, **kwargs): |
d6983cb4 | 1104 | """Print msg to screen, prefixing it with '[ie_name]'""" |
86e5f3ed | 1105 | self._downloader.to_screen(f'[{self.IE_NAME}] {msg}', *args, **kwargs) |
a06916d9 | 1106 | |
1107 | def write_debug(self, msg, *args, **kwargs): | |
86e5f3ed | 1108 | self._downloader.write_debug(f'[{self.IE_NAME}] {msg}', *args, **kwargs) |
a06916d9 | 1109 | |
1110 | def get_param(self, name, default=None, *args, **kwargs): | |
1111 | if self._downloader: | |
1112 | return self._downloader.params.get(name, default, *args, **kwargs) | |
1113 | return default | |
d6983cb4 | 1114 | |
d5d1df8a | 1115 | def report_drm(self, video_id, partial=NO_DEFAULT): |
1116 | if partial is not NO_DEFAULT: | |
1117 | self._downloader.deprecation_warning('InfoExtractor.report_drm no longer accepts the argument partial') | |
88acdbc2 | 1118 | self.raise_no_formats('This video is DRM protected', expected=True, video_id=video_id) |
1119 | ||
d6983cb4 PH |
1120 | def report_extraction(self, id_or_name): |
1121 | """Report information extraction.""" | |
f1a9d64e | 1122 | self.to_screen('%s: Extracting information' % id_or_name) |
d6983cb4 PH |
1123 | |
1124 | def report_download_webpage(self, video_id): | |
1125 | """Report webpage download.""" | |
f1a9d64e | 1126 | self.to_screen('%s: Downloading webpage' % video_id) |
d6983cb4 PH |
1127 | |
1128 | def report_age_confirmation(self): | |
1129 | """Report attempt to confirm age.""" | |
f1a9d64e | 1130 | self.to_screen('Confirming age') |
d6983cb4 | 1131 | |
fc79158d JMF |
1132 | def report_login(self): |
1133 | """Report attempt to log in.""" | |
f1a9d64e | 1134 | self.to_screen('Logging in') |
fc79158d | 1135 | |
b7da73eb | 1136 | def raise_login_required( |
9d5d4d64 | 1137 | self, msg='This video is only available for registered users', |
52efa4b3 | 1138 | metadata_available=False, method=NO_DEFAULT): |
f2ebc5c7 | 1139 | if metadata_available and ( |
1140 | self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')): | |
b7da73eb | 1141 | self.report_warning(msg) |
7265a219 | 1142 | return |
a70635b8 | 1143 | msg += format_field(self._login_hint(method), None, '. %s') |
46890374 | 1144 | raise ExtractorError(msg, expected=True) |
43e7d3c9 | 1145 | |
b7da73eb | 1146 | def raise_geo_restricted( |
1147 | self, msg='This video is not available from your location due to geo restriction', | |
1148 | countries=None, metadata_available=False): | |
f2ebc5c7 | 1149 | if metadata_available and ( |
1150 | self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')): | |
b7da73eb | 1151 | self.report_warning(msg) |
1152 | else: | |
1153 | raise GeoRestrictedError(msg, countries=countries) | |
1154 | ||
1155 | def raise_no_formats(self, msg, expected=False, video_id=None): | |
f2ebc5c7 | 1156 | if expected and ( |
1157 | self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')): | |
b7da73eb | 1158 | self.report_warning(msg, video_id) |
68f5867c L |
1159 | elif isinstance(msg, ExtractorError): |
1160 | raise msg | |
b7da73eb | 1161 | else: |
1162 | raise ExtractorError(msg, expected=expected, video_id=video_id) | |
c430802e | 1163 | |
5f6a1245 | 1164 | # Methods for following #608 |
c0d0b01f | 1165 | @staticmethod |
311b6615 | 1166 | def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs): |
10952eb2 | 1167 | """Returns a URL that points to a page that should be processed""" |
311b6615 | 1168 | if ie is not None: |
1169 | kwargs['ie_key'] = ie if isinstance(ie, str) else ie.ie_key() | |
7012b23c | 1170 | if video_id is not None: |
311b6615 | 1171 | kwargs['id'] = video_id |
830d53bf | 1172 | if video_title is not None: |
311b6615 | 1173 | kwargs['title'] = video_title |
1174 | return { | |
1175 | **kwargs, | |
1176 | '_type': 'url_transparent' if url_transparent else 'url', | |
1177 | 'url': url, | |
1178 | } | |
1179 | ||
8f97a15d | 1180 | @classmethod |
1181 | def playlist_from_matches(cls, matches, playlist_id=None, playlist_title=None, | |
1182 | getter=IDENTITY, ie=None, video_kwargs=None, **kwargs): | |
1183 | return cls.playlist_result( | |
1184 | (cls.url_result(m, ie, **(video_kwargs or {})) for m in orderedSet(map(getter, matches), lazy=True)), | |
1185 | playlist_id, playlist_title, **kwargs) | |
46b18f23 | 1186 | |
c0d0b01f | 1187 | @staticmethod |
311b6615 | 1188 | def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs): |
d6983cb4 | 1189 | """Returns a playlist""" |
d6983cb4 | 1190 | if playlist_id: |
311b6615 | 1191 | kwargs['id'] = playlist_id |
d6983cb4 | 1192 | if playlist_title: |
311b6615 | 1193 | kwargs['title'] = playlist_title |
ecc97af3 | 1194 | if playlist_description is not None: |
311b6615 | 1195 | kwargs['description'] = playlist_description |
1196 | return { | |
1197 | **kwargs, | |
1198 | '_type': 'multi_video' if multi_video else 'playlist', | |
1199 | 'entries': entries, | |
1200 | } | |
d6983cb4 | 1201 | |
c342041f | 1202 | def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
1203 | """ |
1204 | Perform a regex search on the given string, using a single or a list of | |
1205 | patterns returning the first matching group. | |
1206 | In case of failure return a default value or raise a WARNING or a | |
55b3e45b | 1207 | RegexNotFoundError, depending on fatal, specifying the field name. |
d6983cb4 | 1208 | """ |
61d3665d | 1209 | if string is None: |
1210 | mobj = None | |
77f90330 | 1211 | elif isinstance(pattern, (str, re.Pattern)): |
d6983cb4 PH |
1212 | mobj = re.search(pattern, string, flags) |
1213 | else: | |
1214 | for p in pattern: | |
1215 | mobj = re.search(p, string, flags) | |
c3415d1b PH |
1216 | if mobj: |
1217 | break | |
d6983cb4 | 1218 | |
ec11a9f4 | 1219 | _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS) |
d6983cb4 PH |
1220 | |
1221 | if mobj: | |
711ede6e PH |
1222 | if group is None: |
1223 | # return the first matching group | |
1224 | return next(g for g in mobj.groups() if g is not None) | |
198f7ea8 | 1225 | elif isinstance(group, (list, tuple)): |
1226 | return tuple(mobj.group(g) for g in group) | |
711ede6e PH |
1227 | else: |
1228 | return mobj.group(group) | |
c342041f | 1229 | elif default is not NO_DEFAULT: |
d6983cb4 PH |
1230 | return default |
1231 | elif fatal: | |
f1a9d64e | 1232 | raise RegexNotFoundError('Unable to extract %s' % _name) |
d6983cb4 | 1233 | else: |
6a39ee13 | 1234 | self.report_warning('unable to extract %s' % _name + bug_reports_message()) |
d6983cb4 PH |
1235 | return None |
1236 | ||
f0bc6e20 | 1237 | def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='', |
8b7fb8b6 | 1238 | contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT, **kwargs): |
b7c47b74 | 1239 | """Searches string for the JSON object specified by start_pattern""" |
1240 | # NB: end_pattern is only used to reduce the size of the initial match | |
f0bc6e20 | 1241 | if default is NO_DEFAULT: |
1242 | default, has_default = {}, False | |
1243 | else: | |
1244 | fatal, has_default = False, True | |
1245 | ||
1246 | json_string = self._search_regex( | |
8b7fb8b6 | 1247 | rf'(?:{start_pattern})\s*(?P<json>{contains_pattern})\s*(?:{end_pattern})', |
f0bc6e20 | 1248 | string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT) |
1249 | if not json_string: | |
1250 | return default | |
1251 | ||
1252 | _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS) | |
1253 | try: | |
1254 | return self._parse_json(json_string, video_id, ignore_extra=True, **kwargs) | |
1255 | except ExtractorError as e: | |
1256 | if fatal: | |
1257 | raise ExtractorError( | |
1258 | f'Unable to extract {_name} - Failed to parse JSON', cause=e.cause, video_id=video_id) | |
1259 | elif not has_default: | |
1260 | self.report_warning( | |
1261 | f'Unable to extract {_name} - Failed to parse JSON: {e}', video_id=video_id) | |
1262 | return default | |
b7c47b74 | 1263 | |
c342041f | 1264 | def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
1265 | """ |
1266 | Like _search_regex, but strips HTML tags and unescapes entities. | |
1267 | """ | |
711ede6e | 1268 | res = self._search_regex(pattern, string, name, default, fatal, flags, group) |
d6983cb4 PH |
1269 | if res: |
1270 | return clean_html(res).strip() | |
1271 | else: | |
1272 | return res | |
1273 | ||
2118fdd1 RA |
1274 | def _get_netrc_login_info(self, netrc_machine=None): |
1275 | username = None | |
1276 | password = None | |
1277 | netrc_machine = netrc_machine or self._NETRC_MACHINE | |
1278 | ||
a06916d9 | 1279 | if self.get_param('usenetrc', False): |
2118fdd1 | 1280 | try: |
0001fcb5 | 1281 | netrc_file = compat_expanduser(self.get_param('netrc_location') or '~') |
1282 | if os.path.isdir(netrc_file): | |
1283 | netrc_file = os.path.join(netrc_file, '.netrc') | |
1284 | info = netrc.netrc(file=netrc_file).authenticators(netrc_machine) | |
2118fdd1 RA |
1285 | if info is not None: |
1286 | username = info[0] | |
1287 | password = info[2] | |
1288 | else: | |
dcce092e S |
1289 | raise netrc.NetrcParseError( |
1290 | 'No authenticators for %s' % netrc_machine) | |
86e5f3ed | 1291 | except (OSError, netrc.NetrcParseError) as err: |
6a39ee13 | 1292 | self.report_warning( |
dcce092e | 1293 | 'parsing .netrc: %s' % error_to_compat_str(err)) |
2118fdd1 | 1294 | |
dcce092e | 1295 | return username, password |
2118fdd1 | 1296 | |
1b6712ab | 1297 | def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None): |
fc79158d | 1298 | """ |
cf0649f8 | 1299 | Get the login info as (username, password) |
32443dd3 S |
1300 | First look for the manually specified credentials using username_option |
1301 | and password_option as keys in params dictionary. If no such credentials | |
1302 | available look in the netrc file using the netrc_machine or _NETRC_MACHINE | |
1303 | value. | |
fc79158d JMF |
1304 | If there's no info available, return (None, None) |
1305 | """ | |
fc79158d JMF |
1306 | |
1307 | # Attempt to use provided username and password or .netrc data | |
a06916d9 | 1308 | username = self.get_param(username_option) |
1309 | if username is not None: | |
1310 | password = self.get_param(password_option) | |
2118fdd1 | 1311 | else: |
1b6712ab | 1312 | username, password = self._get_netrc_login_info(netrc_machine) |
5f6a1245 | 1313 | |
2133565c | 1314 | return username, password |
fc79158d | 1315 | |
e64b7569 | 1316 | def _get_tfa_info(self, note='two-factor verification code'): |
83317f69 | 1317 | """ |
1318 | Get the two-factor authentication info | |
1319 | TODO - asking the user will be required for sms/phone verify | |
1320 | currently just uses the command line option | |
1321 | If there's no info available, return None | |
1322 | """ | |
83317f69 | 1323 | |
a06916d9 | 1324 | tfa = self.get_param('twofactor') |
1325 | if tfa is not None: | |
1326 | return tfa | |
83317f69 | 1327 | |
ac668111 | 1328 | return getpass.getpass('Type %s and press [Return]: ' % note) |
83317f69 | 1329 | |
46720279 JMF |
1330 | # Helper functions for extracting OpenGraph info |
1331 | @staticmethod | |
ab2d5247 | 1332 | def _og_regexes(prop): |
448ef1f3 | 1333 | content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))' |
fbfde1c3 F |
1334 | property_re = (r'(?:name|property)=(?:\'og%(sep)s%(prop)s\'|"og%(sep)s%(prop)s"|\s*og%(sep)s%(prop)s\b)' |
1335 | % {'prop': re.escape(prop), 'sep': '(?::|[:-])'}) | |
78fb87b2 | 1336 | template = r'<meta[^>]+?%s[^>]+?%s' |
ab2d5247 | 1337 | return [ |
78fb87b2 JMF |
1338 | template % (property_re, content_re), |
1339 | template % (content_re, property_re), | |
ab2d5247 | 1340 | ] |
46720279 | 1341 | |
864f24bd S |
1342 | @staticmethod |
1343 | def _meta_regex(prop): | |
1344 | return r'''(?isx)<meta | |
8b9848ac | 1345 | (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1) |
864f24bd S |
1346 | [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop) |
1347 | ||
3c4e6d83 | 1348 | def _og_search_property(self, prop, html, name=None, **kargs): |
6606817a | 1349 | prop = variadic(prop) |
46720279 | 1350 | if name is None: |
b070564e S |
1351 | name = 'OpenGraph %s' % prop[0] |
1352 | og_regexes = [] | |
1353 | for p in prop: | |
1354 | og_regexes.extend(self._og_regexes(p)) | |
1355 | escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs) | |
eb0a8398 PH |
1356 | if escaped is None: |
1357 | return None | |
1358 | return unescapeHTML(escaped) | |
46720279 JMF |
1359 | |
1360 | def _og_search_thumbnail(self, html, **kargs): | |
10952eb2 | 1361 | return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs) |
46720279 JMF |
1362 | |
1363 | def _og_search_description(self, html, **kargs): | |
1364 | return self._og_search_property('description', html, fatal=False, **kargs) | |
1365 | ||
04f3fd2c | 1366 | def _og_search_title(self, html, *, fatal=False, **kargs): |
1367 | return self._og_search_property('title', html, fatal=fatal, **kargs) | |
46720279 | 1368 | |
8ffa13e0 | 1369 | def _og_search_video_url(self, html, name='video url', secure=True, **kargs): |
a3681973 PH |
1370 | regexes = self._og_regexes('video') + self._og_regexes('video:url') |
1371 | if secure: | |
1372 | regexes = self._og_regexes('video:secure_url') + regexes | |
8ffa13e0 | 1373 | return self._html_search_regex(regexes, html, name, **kargs) |
46720279 | 1374 | |
78338f71 JMF |
1375 | def _og_search_url(self, html, **kargs): |
1376 | return self._og_search_property('url', html, **kargs) | |
1377 | ||
04f3fd2c | 1378 | def _html_extract_title(self, html, name='title', *, fatal=False, **kwargs): |
21633673 | 1379 | return self._html_search_regex(r'(?s)<title\b[^>]*>([^<]+)</title>', html, name, fatal=fatal, **kwargs) |
77cc7c6e | 1380 | |
40c696e5 | 1381 | def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs): |
6606817a | 1382 | name = variadic(name) |
59040888 | 1383 | if display_name is None: |
88d9f6c0 | 1384 | display_name = name[0] |
59040888 | 1385 | return self._html_search_regex( |
88d9f6c0 | 1386 | [self._meta_regex(n) for n in name], |
711ede6e | 1387 | html, display_name, fatal=fatal, group='content', **kwargs) |
59040888 PH |
1388 | |
1389 | def _dc_search_uploader(self, html): | |
1390 | return self._html_search_meta('dc.creator', html, 'uploader') | |
1391 | ||
8f97a15d | 1392 | @staticmethod |
1393 | def _rta_search(html): | |
8dbe9899 PH |
1394 | # See http://www.rtalabel.org/index.php?content=howtofaq#single |
1395 | if re.search(r'(?ix)<meta\s+name="rating"\s+' | |
1396 | r' content="RTA-5042-1996-1400-1577-RTA"', | |
1397 | html): | |
1398 | return 18 | |
8f97a15d | 1399 | |
1400 | # And then there are the jokers who advertise that they use RTA, but actually don't. | |
1401 | AGE_LIMIT_MARKERS = [ | |
1402 | r'Proudly Labeled <a href="http://www\.rtalabel\.org/" title="Restricted to Adults">RTA</a>', | |
1403 | ] | |
1404 | if any(re.search(marker, html) for marker in AGE_LIMIT_MARKERS): | |
1405 | return 18 | |
8dbe9899 PH |
1406 | return 0 |
1407 | ||
59040888 PH |
1408 | def _media_rating_search(self, html): |
1409 | # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/ | |
1410 | rating = self._html_search_meta('rating', html) | |
1411 | ||
1412 | if not rating: | |
1413 | return None | |
1414 | ||
1415 | RATING_TABLE = { | |
1416 | 'safe for kids': 0, | |
1417 | 'general': 8, | |
1418 | '14 years': 14, | |
1419 | 'mature': 17, | |
1420 | 'restricted': 19, | |
1421 | } | |
d800609c | 1422 | return RATING_TABLE.get(rating.lower()) |
59040888 | 1423 | |
69319969 | 1424 | def _family_friendly_search(self, html): |
6ca7732d | 1425 | # See http://schema.org/VideoObject |
ac8491fc S |
1426 | family_friendly = self._html_search_meta( |
1427 | 'isFamilyFriendly', html, default=None) | |
69319969 NJ |
1428 | |
1429 | if not family_friendly: | |
1430 | return None | |
1431 | ||
1432 | RATING_TABLE = { | |
1433 | '1': 0, | |
1434 | 'true': 0, | |
1435 | '0': 18, | |
1436 | 'false': 18, | |
1437 | } | |
d800609c | 1438 | return RATING_TABLE.get(family_friendly.lower()) |
69319969 | 1439 | |
0c708f11 JMF |
1440 | def _twitter_search_player(self, html): |
1441 | return self._html_search_meta('twitter:player', html, | |
9e1a5b84 | 1442 | 'twitter card player') |
0c708f11 | 1443 | |
0c36dc00 | 1444 | def _yield_json_ld(self, html, video_id, *, fatal=True, default=NO_DEFAULT): |
1445 | """Yield all json ld objects in the html""" | |
1446 | if default is not NO_DEFAULT: | |
1447 | fatal = False | |
1448 | for mobj in re.finditer(JSON_LD_RE, html): | |
1449 | json_ld_item = self._parse_json(mobj.group('json_ld'), video_id, fatal=fatal) | |
1450 | for json_ld in variadic(json_ld_item): | |
1451 | if isinstance(json_ld, dict): | |
1452 | yield json_ld | |
1453 | ||
1454 | def _search_json_ld(self, html, video_id, expected_type=None, *, fatal=True, default=NO_DEFAULT): | |
1455 | """Search for a video in any json ld in the html""" | |
1456 | if default is not NO_DEFAULT: | |
1457 | fatal = False | |
1458 | info = self._json_ld( | |
1459 | list(self._yield_json_ld(html, video_id, fatal=fatal, default=default)), | |
1460 | video_id, fatal=fatal, expected_type=expected_type) | |
1461 | if info: | |
1462 | return info | |
4433bb02 S |
1463 | if default is not NO_DEFAULT: |
1464 | return default | |
1465 | elif fatal: | |
1466 | raise RegexNotFoundError('Unable to extract JSON-LD') | |
1467 | else: | |
6a39ee13 | 1468 | self.report_warning('unable to extract JSON-LD %s' % bug_reports_message()) |
4433bb02 | 1469 | return {} |
4ca2a3cf | 1470 | |
95b31e26 | 1471 | def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): |
14f25df2 | 1472 | if isinstance(json_ld, str): |
4ca2a3cf S |
1473 | json_ld = self._parse_json(json_ld, video_id, fatal=fatal) |
1474 | if not json_ld: | |
1475 | return {} | |
1476 | info = {} | |
bae14048 | 1477 | |
e7e4a6e0 S |
1478 | INTERACTION_TYPE_MAP = { |
1479 | 'CommentAction': 'comment', | |
1480 | 'AgreeAction': 'like', | |
1481 | 'DisagreeAction': 'dislike', | |
1482 | 'LikeAction': 'like', | |
1483 | 'DislikeAction': 'dislike', | |
1484 | 'ListenAction': 'view', | |
1485 | 'WatchAction': 'view', | |
1486 | 'ViewAction': 'view', | |
1487 | } | |
1488 | ||
f3c0c773 | 1489 | def is_type(e, *expected_types): |
1490 | type = variadic(traverse_obj(e, '@type')) | |
1491 | return any(x in type for x in expected_types) | |
1492 | ||
29f7c58a | 1493 | def extract_interaction_type(e): |
1494 | interaction_type = e.get('interactionType') | |
1495 | if isinstance(interaction_type, dict): | |
1496 | interaction_type = interaction_type.get('@type') | |
1497 | return str_or_none(interaction_type) | |
1498 | ||
e7e4a6e0 S |
1499 | def extract_interaction_statistic(e): |
1500 | interaction_statistic = e.get('interactionStatistic') | |
29f7c58a | 1501 | if isinstance(interaction_statistic, dict): |
1502 | interaction_statistic = [interaction_statistic] | |
e7e4a6e0 S |
1503 | if not isinstance(interaction_statistic, list): |
1504 | return | |
1505 | for is_e in interaction_statistic: | |
f3c0c773 | 1506 | if not is_type(is_e, 'InteractionCounter'): |
e7e4a6e0 | 1507 | continue |
29f7c58a | 1508 | interaction_type = extract_interaction_type(is_e) |
1509 | if not interaction_type: | |
e7e4a6e0 | 1510 | continue |
ce5b9040 S |
1511 | # For interaction count some sites provide string instead of |
1512 | # an integer (as per spec) with non digit characters (e.g. ",") | |
1513 | # so extracting count with more relaxed str_to_int | |
1514 | interaction_count = str_to_int(is_e.get('userInteractionCount')) | |
e7e4a6e0 S |
1515 | if interaction_count is None: |
1516 | continue | |
1517 | count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1]) | |
1518 | if not count_kind: | |
1519 | continue | |
1520 | count_key = '%s_count' % count_kind | |
1521 | if info.get(count_key) is not None: | |
1522 | continue | |
1523 | info[count_key] = interaction_count | |
1524 | ||
f5225737 | 1525 | def extract_chapter_information(e): |
1526 | chapters = [{ | |
1527 | 'title': part.get('name'), | |
1528 | 'start_time': part.get('startOffset'), | |
1529 | 'end_time': part.get('endOffset'), | |
85553414 | 1530 | } for part in variadic(e.get('hasPart') or []) if part.get('@type') == 'Clip'] |
f5225737 | 1531 | for idx, (last_c, current_c, next_c) in enumerate(zip( |
1532 | [{'end_time': 0}] + chapters, chapters, chapters[1:])): | |
1533 | current_c['end_time'] = current_c['end_time'] or next_c['start_time'] | |
1534 | current_c['start_time'] = current_c['start_time'] or last_c['end_time'] | |
1535 | if None in current_c.values(): | |
1536 | self.report_warning(f'Chapter {idx} contains broken data. Not extracting chapters') | |
1537 | return | |
1538 | if chapters: | |
1539 | chapters[-1]['end_time'] = chapters[-1]['end_time'] or info['duration'] | |
1540 | info['chapters'] = chapters | |
1541 | ||
bae14048 | 1542 | def extract_video_object(e): |
f7ad7160 | 1543 | author = e.get('author') |
bae14048 | 1544 | info.update({ |
0c36dc00 | 1545 | 'url': url_or_none(e.get('contentUrl')), |
0f60ba6e | 1546 | 'ext': mimetype2ext(e.get('encodingFormat')), |
bae14048 S |
1547 | 'title': unescapeHTML(e.get('name')), |
1548 | 'description': unescapeHTML(e.get('description')), | |
eb2333bc | 1549 | 'thumbnails': [{'url': unescapeHTML(url)} |
21633673 | 1550 | for url in variadic(traverse_obj(e, 'thumbnailUrl', 'thumbnailURL')) |
1551 | if url_or_none(url)], | |
bae14048 S |
1552 | 'duration': parse_duration(e.get('duration')), |
1553 | 'timestamp': unified_timestamp(e.get('uploadDate')), | |
f7ad7160 | 1554 | # author can be an instance of 'Organization' or 'Person' types. |
1555 | # both types can have 'name' property(inherited from 'Thing' type). [1] | |
1556 | # however some websites are using 'Text' type instead. | |
1557 | # 1. https://schema.org/VideoObject | |
14f25df2 | 1558 | 'uploader': author.get('name') if isinstance(author, dict) else author if isinstance(author, str) else None, |
0f60ba6e | 1559 | 'artist': traverse_obj(e, ('byArtist', 'name'), expected_type=str), |
56ba69e4 | 1560 | 'filesize': int_or_none(float_or_none(e.get('contentSize'))), |
bae14048 S |
1561 | 'tbr': int_or_none(e.get('bitrate')), |
1562 | 'width': int_or_none(e.get('width')), | |
1563 | 'height': int_or_none(e.get('height')), | |
33a81c2c | 1564 | 'view_count': int_or_none(e.get('interactionCount')), |
0f60ba6e | 1565 | 'tags': try_call(lambda: e.get('keywords').split(',')), |
bae14048 | 1566 | }) |
0f60ba6e | 1567 | if is_type(e, 'AudioObject'): |
1568 | info.update({ | |
1569 | 'vcodec': 'none', | |
1570 | 'abr': int_or_none(e.get('bitrate')), | |
1571 | }) | |
e7e4a6e0 | 1572 | extract_interaction_statistic(e) |
f5225737 | 1573 | extract_chapter_information(e) |
bae14048 | 1574 | |
d5c32548 | 1575 | def traverse_json_ld(json_ld, at_top_level=True): |
1d55ebab SS |
1576 | for e in variadic(json_ld): |
1577 | if not isinstance(e, dict): | |
1578 | continue | |
d5c32548 ZM |
1579 | if at_top_level and '@context' not in e: |
1580 | continue | |
1581 | if at_top_level and set(e.keys()) == {'@context', '@graph'}: | |
1d55ebab | 1582 | traverse_json_ld(e['@graph'], at_top_level=False) |
c13a301a | 1583 | continue |
f3c0c773 | 1584 | if expected_type is not None and not is_type(e, expected_type): |
4433bb02 | 1585 | continue |
8f122fa0 | 1586 | rating = traverse_obj(e, ('aggregateRating', 'ratingValue'), expected_type=float_or_none) |
1587 | if rating is not None: | |
1588 | info['average_rating'] = rating | |
f3c0c773 | 1589 | if is_type(e, 'TVEpisode', 'Episode'): |
440863ad | 1590 | episode_name = unescapeHTML(e.get('name')) |
46933a15 | 1591 | info.update({ |
440863ad | 1592 | 'episode': episode_name, |
46933a15 S |
1593 | 'episode_number': int_or_none(e.get('episodeNumber')), |
1594 | 'description': unescapeHTML(e.get('description')), | |
1595 | }) | |
440863ad S |
1596 | if not info.get('title') and episode_name: |
1597 | info['title'] = episode_name | |
46933a15 | 1598 | part_of_season = e.get('partOfSeason') |
f3c0c773 | 1599 | if is_type(part_of_season, 'TVSeason', 'Season', 'CreativeWorkSeason'): |
458fd30f S |
1600 | info.update({ |
1601 | 'season': unescapeHTML(part_of_season.get('name')), | |
1602 | 'season_number': int_or_none(part_of_season.get('seasonNumber')), | |
1603 | }) | |
d16b3c66 | 1604 | part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries') |
f3c0c773 | 1605 | if is_type(part_of_series, 'TVSeries', 'Series', 'CreativeWorkSeries'): |
46933a15 | 1606 | info['series'] = unescapeHTML(part_of_series.get('name')) |
f3c0c773 | 1607 | elif is_type(e, 'Movie'): |
391256dc S |
1608 | info.update({ |
1609 | 'title': unescapeHTML(e.get('name')), | |
1610 | 'description': unescapeHTML(e.get('description')), | |
1611 | 'duration': parse_duration(e.get('duration')), | |
1612 | 'timestamp': unified_timestamp(e.get('dateCreated')), | |
1613 | }) | |
f3c0c773 | 1614 | elif is_type(e, 'Article', 'NewsArticle'): |
46933a15 S |
1615 | info.update({ |
1616 | 'timestamp': parse_iso8601(e.get('datePublished')), | |
1617 | 'title': unescapeHTML(e.get('headline')), | |
d5c32548 | 1618 | 'description': unescapeHTML(e.get('articleBody') or e.get('description')), |
46933a15 | 1619 | }) |
f3c0c773 | 1620 | if is_type(traverse_obj(e, ('video', 0)), 'VideoObject'): |
2edb38e8 | 1621 | extract_video_object(e['video'][0]) |
f3c0c773 | 1622 | elif is_type(traverse_obj(e, ('subjectOf', 0)), 'VideoObject'): |
e50c3500 | 1623 | extract_video_object(e['subjectOf'][0]) |
0f60ba6e | 1624 | elif is_type(e, 'VideoObject', 'AudioObject'): |
bae14048 | 1625 | extract_video_object(e) |
4433bb02 S |
1626 | if expected_type is None: |
1627 | continue | |
1628 | else: | |
1629 | break | |
c69701c6 | 1630 | video = e.get('video') |
f3c0c773 | 1631 | if is_type(video, 'VideoObject'): |
c69701c6 | 1632 | extract_video_object(video) |
4433bb02 S |
1633 | if expected_type is None: |
1634 | continue | |
1635 | else: | |
1636 | break | |
d5c32548 | 1637 | |
1d55ebab | 1638 | traverse_json_ld(json_ld) |
90137ca4 | 1639 | return filter_dict(info) |
4ca2a3cf | 1640 | |
135dfa2c | 1641 | def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw): |
f98709af LL |
1642 | return self._parse_json( |
1643 | self._search_regex( | |
1644 | r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>', | |
135dfa2c | 1645 | webpage, 'next.js data', fatal=fatal, **kw), |
1646 | video_id, transform_source=transform_source, fatal=fatal) | |
f98709af | 1647 | |
8072ef2b | 1648 | def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__', *, fatal=True, traverse=('data', 0)): |
1649 | """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function""" | |
66f4c04e | 1650 | rectx = re.escape(context_name) |
8072ef2b | 1651 | FUNCTION_RE = r'\(function\((?P<arg_keys>.*?)\){return\s+(?P<js>{.*?})\s*;?\s*}\((?P<arg_vals>.*?)\)' |
66f4c04e | 1652 | js, arg_keys, arg_vals = self._search_regex( |
8072ef2b | 1653 | (rf'<script>\s*window\.{rectx}={FUNCTION_RE}\s*\)\s*;?\s*</script>', rf'{rectx}\(.*?{FUNCTION_RE}'), |
f7fc8d39 | 1654 | webpage, context_name, group=('js', 'arg_keys', 'arg_vals'), |
1655 | default=NO_DEFAULT if fatal else (None, None, None)) | |
1656 | if js is None: | |
1657 | return {} | |
66f4c04e THD |
1658 | |
1659 | args = dict(zip(arg_keys.split(','), arg_vals.split(','))) | |
1660 | ||
1661 | for key, val in args.items(): | |
1662 | if val in ('undefined', 'void 0'): | |
1663 | args[key] = 'null' | |
1664 | ||
8072ef2b | 1665 | ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal) |
1666 | return traverse_obj(ret, traverse) or {} | |
66f4c04e | 1667 | |
27713812 | 1668 | @staticmethod |
f8da79f8 | 1669 | def _hidden_inputs(html): |
586f1cc5 | 1670 | html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html) |
201ea3ee | 1671 | hidden_inputs = {} |
c8498368 S |
1672 | for input in re.findall(r'(?i)(<input[^>]+>)', html): |
1673 | attrs = extract_attributes(input) | |
1674 | if not input: | |
201ea3ee | 1675 | continue |
c8498368 | 1676 | if attrs.get('type') not in ('hidden', 'submit'): |
201ea3ee | 1677 | continue |
c8498368 S |
1678 | name = attrs.get('name') or attrs.get('id') |
1679 | value = attrs.get('value') | |
1680 | if name and value is not None: | |
1681 | hidden_inputs[name] = value | |
201ea3ee | 1682 | return hidden_inputs |
27713812 | 1683 | |
cf61d96d S |
1684 | def _form_hidden_inputs(self, form_id, html): |
1685 | form = self._search_regex( | |
73eb13df | 1686 | r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id, |
cf61d96d S |
1687 | html, '%s form' % form_id, group='form') |
1688 | return self._hidden_inputs(form) | |
1689 | ||
d0d74b71 | 1690 | @classproperty(cache=True) |
1691 | def FormatSort(cls): | |
1692 | class FormatSort(FormatSorter): | |
1693 | def __init__(ie, *args, **kwargs): | |
1694 | super().__init__(ie._downloader, *args, **kwargs) | |
eb8a4433 | 1695 | |
d0d74b71 | 1696 | deprecation_warning( |
1697 | 'yt_dlp.InfoExtractor.FormatSort is deprecated and may be removed in the future. ' | |
1698 | 'Use yt_dlp.utils.FormatSorter instead') | |
1699 | return FormatSort | |
eb8a4433 | 1700 | |
1701 | def _sort_formats(self, formats, field_preference=[]): | |
9f14daf2 | 1702 | if not field_preference: |
1703 | self._downloader.deprecation_warning( | |
1704 | 'yt_dlp.InfoExtractor._sort_formats is deprecated and is no longer required') | |
1705 | return | |
1706 | self._downloader.deprecation_warning( | |
1707 | 'yt_dlp.InfoExtractor._sort_formats is deprecated and no longer works as expected. ' | |
1708 | 'Return _format_sort_fields in the info_dict instead') | |
1709 | if formats: | |
784320c9 | 1710 | formats[0]['__sort_fields'] = field_preference |
59040888 | 1711 | |
96a53167 S |
1712 | def _check_formats(self, formats, video_id): |
1713 | if formats: | |
1714 | formats[:] = filter( | |
1715 | lambda f: self._is_valid_url( | |
1716 | f['url'], video_id, | |
1717 | item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'), | |
1718 | formats) | |
1719 | ||
f5bdb444 S |
1720 | @staticmethod |
1721 | def _remove_duplicate_formats(formats): | |
1722 | format_urls = set() | |
1723 | unique_formats = [] | |
1724 | for f in formats: | |
1725 | if f['url'] not in format_urls: | |
1726 | format_urls.add(f['url']) | |
1727 | unique_formats.append(f) | |
1728 | formats[:] = unique_formats | |
1729 | ||
45024183 | 1730 | def _is_valid_url(self, url, video_id, item='video', headers={}): |
2f0f6578 S |
1731 | url = self._proto_relative_url(url, scheme='http:') |
1732 | # For now assume non HTTP(S) URLs always valid | |
1733 | if not (url.startswith('http://') or url.startswith('https://')): | |
1734 | return True | |
96a53167 | 1735 | try: |
45024183 | 1736 | self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers) |
96a53167 | 1737 | return True |
8bdd16b4 | 1738 | except ExtractorError as e: |
25e911a9 | 1739 | self.to_screen( |
8bdd16b4 | 1740 | '%s: %s URL is invalid, skipping: %s' |
1741 | % (video_id, item, error_to_compat_str(e.cause))) | |
25e911a9 | 1742 | return False |
96a53167 | 1743 | |
20991253 | 1744 | def http_scheme(self): |
1ede5b24 | 1745 | """ Either "http:" or "https:", depending on the user's preferences """ |
20991253 PH |
1746 | return ( |
1747 | 'http:' | |
a06916d9 | 1748 | if self.get_param('prefer_insecure', False) |
20991253 PH |
1749 | else 'https:') |
1750 | ||
57c7411f | 1751 | def _proto_relative_url(self, url, scheme=None): |
8f97a15d | 1752 | scheme = scheme or self.http_scheme() |
1753 | assert scheme.endswith(':') | |
1754 | return sanitize_url(url, scheme=scheme[:-1]) | |
57c7411f | 1755 | |
4094b6e3 PH |
1756 | def _sleep(self, timeout, video_id, msg_template=None): |
1757 | if msg_template is None: | |
f1a9d64e | 1758 | msg_template = '%(video_id)s: Waiting for %(timeout)s seconds' |
4094b6e3 PH |
1759 | msg = msg_template % {'video_id': video_id, 'timeout': timeout} |
1760 | self.to_screen(msg) | |
1761 | time.sleep(timeout) | |
1762 | ||
f983b875 | 1763 | def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None, |
4de61310 | 1764 | transform_source=lambda s: fix_xml_ampersands(s).strip(), |
7360c06f | 1765 | fatal=True, m3u8_id=None, data=None, headers={}, query={}): |
a076c1f9 | 1766 | res = self._download_xml_handle( |
f036a632 | 1767 | manifest_url, video_id, 'Downloading f4m manifest', |
97f4aecf S |
1768 | 'Unable to download f4m manifest', |
1769 | # Some manifests may be malformed, e.g. prosiebensat1 generated manifests | |
067aa17e | 1770 | # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244) |
4de61310 | 1771 | transform_source=transform_source, |
7360c06f | 1772 | fatal=fatal, data=data, headers=headers, query=query) |
a076c1f9 | 1773 | if res is False: |
8d29e47f | 1774 | return [] |
31bb8d3f | 1775 | |
a076c1f9 E |
1776 | manifest, urlh = res |
1777 | manifest_url = urlh.geturl() | |
1778 | ||
0fdbb332 | 1779 | return self._parse_f4m_formats( |
f983b875 | 1780 | manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id, |
448bb5f3 | 1781 | transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id) |
0fdbb332 | 1782 | |
f983b875 | 1783 | def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None, |
0fdbb332 | 1784 | transform_source=lambda s: fix_xml_ampersands(s).strip(), |
448bb5f3 | 1785 | fatal=True, m3u8_id=None): |
f9934b96 | 1786 | if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal: |
d9eb580a S |
1787 | return [] |
1788 | ||
7a5c1cfe | 1789 | # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy |
fb72ec58 | 1790 | akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0') |
1791 | if akamai_pv is not None and ';' in akamai_pv.text: | |
1792 | playerVerificationChallenge = akamai_pv.text.split(';')[0] | |
1793 | if playerVerificationChallenge.strip() != '': | |
1794 | return [] | |
1795 | ||
31bb8d3f | 1796 | formats = [] |
7a47d07c | 1797 | manifest_version = '1.0' |
b2527359 | 1798 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media') |
34e48bed | 1799 | if not media_nodes: |
7a47d07c | 1800 | manifest_version = '2.0' |
34e48bed | 1801 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') |
b22ca762 | 1802 | # Remove unsupported DRM protected media from final formats |
067aa17e | 1803 | # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573). |
b22ca762 S |
1804 | media_nodes = remove_encrypted_media(media_nodes) |
1805 | if not media_nodes: | |
1806 | return formats | |
48107c19 S |
1807 | |
1808 | manifest_base_url = get_base_url(manifest) | |
0a5685b2 | 1809 | |
a6571f10 | 1810 | bootstrap_info = xpath_element( |
0a5685b2 YCH |
1811 | manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'], |
1812 | 'bootstrap info', default=None) | |
1813 | ||
edd6074c RA |
1814 | vcodec = None |
1815 | mime_type = xpath_text( | |
1816 | manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'], | |
1817 | 'base URL', default=None) | |
1818 | if mime_type and mime_type.startswith('audio/'): | |
1819 | vcodec = 'none' | |
1820 | ||
b2527359 | 1821 | for i, media_el in enumerate(media_nodes): |
77b8b4e6 S |
1822 | tbr = int_or_none(media_el.attrib.get('bitrate')) |
1823 | width = int_or_none(media_el.attrib.get('width')) | |
1824 | height = int_or_none(media_el.attrib.get('height')) | |
34921b43 | 1825 | format_id = join_nonempty(f4m_id, tbr or i) |
448bb5f3 YCH |
1826 | # If <bootstrapInfo> is present, the specified f4m is a |
1827 | # stream-level manifest, and only set-level manifests may refer to | |
1828 | # external resources. See section 11.4 and section 4 of F4M spec | |
1829 | if bootstrap_info is None: | |
1830 | media_url = None | |
1831 | # @href is introduced in 2.0, see section 11.6 of F4M spec | |
1832 | if manifest_version == '2.0': | |
1833 | media_url = media_el.attrib.get('href') | |
1834 | if media_url is None: | |
1835 | media_url = media_el.attrib.get('url') | |
31c746e5 S |
1836 | if not media_url: |
1837 | continue | |
cc357c4d S |
1838 | manifest_url = ( |
1839 | media_url if media_url.startswith('http://') or media_url.startswith('https://') | |
48107c19 | 1840 | else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url)) |
70f0f5a8 S |
1841 | # If media_url is itself a f4m manifest do the recursive extraction |
1842 | # since bitrates in parent manifest (this one) and media_url manifest | |
1843 | # may differ leading to inability to resolve the format by requested | |
1844 | # bitrate in f4m downloader | |
240b6045 YCH |
1845 | ext = determine_ext(manifest_url) |
1846 | if ext == 'f4m': | |
77b8b4e6 | 1847 | f4m_formats = self._extract_f4m_formats( |
f983b875 | 1848 | manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id, |
77b8b4e6 S |
1849 | transform_source=transform_source, fatal=fatal) |
1850 | # Sometimes stream-level manifest contains single media entry that | |
1851 | # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player). | |
1852 | # At the same time parent's media entry in set-level manifest may | |
1853 | # contain it. We will copy it from parent in such cases. | |
1854 | if len(f4m_formats) == 1: | |
1855 | f = f4m_formats[0] | |
1856 | f.update({ | |
1857 | 'tbr': f.get('tbr') or tbr, | |
1858 | 'width': f.get('width') or width, | |
1859 | 'height': f.get('height') or height, | |
1860 | 'format_id': f.get('format_id') if not tbr else format_id, | |
edd6074c | 1861 | 'vcodec': vcodec, |
77b8b4e6 S |
1862 | }) |
1863 | formats.extend(f4m_formats) | |
70f0f5a8 | 1864 | continue |
240b6045 YCH |
1865 | elif ext == 'm3u8': |
1866 | formats.extend(self._extract_m3u8_formats( | |
1867 | manifest_url, video_id, 'mp4', preference=preference, | |
f983b875 | 1868 | quality=quality, m3u8_id=m3u8_id, fatal=fatal)) |
240b6045 | 1869 | continue |
31bb8d3f | 1870 | formats.append({ |
77b8b4e6 | 1871 | 'format_id': format_id, |
31bb8d3f | 1872 | 'url': manifest_url, |
30d0b549 | 1873 | 'manifest_url': manifest_url, |
a6571f10 | 1874 | 'ext': 'flv' if bootstrap_info is not None else None, |
187ee66c | 1875 | 'protocol': 'f4m', |
b2527359 | 1876 | 'tbr': tbr, |
77b8b4e6 S |
1877 | 'width': width, |
1878 | 'height': height, | |
edd6074c | 1879 | 'vcodec': vcodec, |
60ca389c | 1880 | 'preference': preference, |
f983b875 | 1881 | 'quality': quality, |
31bb8d3f | 1882 | }) |
31bb8d3f JMF |
1883 | return formats |
1884 | ||
f983b875 | 1885 | def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None): |
16da9bbc | 1886 | return { |
34921b43 | 1887 | 'format_id': join_nonempty(m3u8_id, 'meta'), |
704df56d PH |
1888 | 'url': m3u8_url, |
1889 | 'ext': ext, | |
1890 | 'protocol': 'm3u8', | |
37768f92 | 1891 | 'preference': preference - 100 if preference else -100, |
f983b875 | 1892 | 'quality': quality, |
704df56d PH |
1893 | 'resolution': 'multiple', |
1894 | 'format_note': 'Quality selection URL', | |
16da9bbc YCH |
1895 | } |
1896 | ||
b5ae35ee | 1897 | def _report_ignoring_subs(self, name): |
1898 | self.report_warning(bug_reports_message( | |
1899 | f'Ignoring subtitle tracks found in the {name} manifest; ' | |
1900 | 'if any subtitle tracks are missing,' | |
1901 | ), only_once=True) | |
1902 | ||
a0c3b2d5 F |
1903 | def _extract_m3u8_formats(self, *args, **kwargs): |
1904 | fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs) | |
1905 | if subs: | |
b5ae35ee | 1906 | self._report_ignoring_subs('HLS') |
a0c3b2d5 F |
1907 | return fmts |
1908 | ||
1909 | def _extract_m3u8_formats_and_subtitles( | |
177877c5 | 1910 | self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native', |
a0c3b2d5 F |
1911 | preference=None, quality=None, m3u8_id=None, note=None, |
1912 | errnote=None, fatal=True, live=False, data=None, headers={}, | |
1913 | query={}): | |
1914 | ||
dbd82a1d | 1915 | res = self._download_webpage_handle( |
81515ad9 | 1916 | m3u8_url, video_id, |
37a3bb66 | 1917 | note='Downloading m3u8 information' if note is None else note, |
1918 | errnote='Failed to download m3u8 information' if errnote is None else errnote, | |
7360c06f | 1919 | fatal=fatal, data=data, headers=headers, query=query) |
cb252080 | 1920 | |
dbd82a1d | 1921 | if res is False: |
a0c3b2d5 | 1922 | return [], {} |
cb252080 | 1923 | |
dbd82a1d | 1924 | m3u8_doc, urlh = res |
37113045 | 1925 | m3u8_url = urlh.geturl() |
9cdffeeb | 1926 | |
a0c3b2d5 | 1927 | return self._parse_m3u8_formats_and_subtitles( |
cb252080 | 1928 | m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol, |
310c2ed2 | 1929 | preference=preference, quality=quality, m3u8_id=m3u8_id, |
1930 | note=note, errnote=errnote, fatal=fatal, live=live, data=data, | |
1931 | headers=headers, query=query, video_id=video_id) | |
cb252080 | 1932 | |
a0c3b2d5 | 1933 | def _parse_m3u8_formats_and_subtitles( |
42676437 | 1934 | self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native', |
a0c3b2d5 F |
1935 | preference=None, quality=None, m3u8_id=None, live=False, note=None, |
1936 | errnote=None, fatal=True, data=None, headers={}, query={}, | |
1937 | video_id=None): | |
60755938 | 1938 | formats, subtitles = [], {} |
a0c3b2d5 | 1939 | |
6b993ca7 | 1940 | has_drm = re.search('|'.join([ |
1941 | r'#EXT-X-FAXS-CM:', # Adobe Flash Access | |
1942 | r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd://', # Apple FairPlay | |
1943 | ]), m3u8_doc) | |
a0c3b2d5 | 1944 | |
60755938 | 1945 | def format_url(url): |
14f25df2 | 1946 | return url if re.match(r'^https?://', url) else urllib.parse.urljoin(m3u8_url, url) |
60755938 | 1947 | |
1948 | if self.get_param('hls_split_discontinuity', False): | |
1949 | def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None): | |
1950 | if not m3u8_doc: | |
1951 | if not manifest_url: | |
1952 | return [] | |
1953 | m3u8_doc = self._download_webpage( | |
1954 | manifest_url, video_id, fatal=fatal, data=data, headers=headers, | |
1955 | note=False, errnote='Failed to download m3u8 playlist information') | |
1956 | if m3u8_doc is False: | |
1957 | return [] | |
1958 | return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines())) | |
0def7587 | 1959 | |
60755938 | 1960 | else: |
1961 | def _extract_m3u8_playlist_indices(*args, **kwargs): | |
1962 | return [None] | |
310c2ed2 | 1963 | |
cb252080 S |
1964 | # References: |
1965 | # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21 | |
067aa17e S |
1966 | # 2. https://github.com/ytdl-org/youtube-dl/issues/12211 |
1967 | # 3. https://github.com/ytdl-org/youtube-dl/issues/18923 | |
cb252080 S |
1968 | |
1969 | # We should try extracting formats only from master playlists [1, 4.3.4], | |
1970 | # i.e. playlists that describe available qualities. On the other hand | |
1971 | # media playlists [1, 4.3.3] should be returned as is since they contain | |
1972 | # just the media without qualities renditions. | |
9cdffeeb | 1973 | # Fortunately, master playlist can be easily distinguished from media |
cb252080 | 1974 | # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4] |
a0566bbf | 1975 | # master playlist tags MUST NOT appear in a media playlist and vice versa. |
cb252080 S |
1976 | # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every |
1977 | # media playlist and MUST NOT appear in master playlist thus we can | |
1978 | # clearly detect media playlist with this criterion. | |
1979 | ||
9cdffeeb | 1980 | if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is |
60755938 | 1981 | formats = [{ |
34921b43 | 1982 | 'format_id': join_nonempty(m3u8_id, idx), |
60755938 | 1983 | 'format_index': idx, |
42676437 | 1984 | 'url': m3u8_url or encode_data_uri(m3u8_doc.encode('utf-8'), 'application/x-mpegurl'), |
60755938 | 1985 | 'ext': ext, |
1986 | 'protocol': entry_protocol, | |
1987 | 'preference': preference, | |
1988 | 'quality': quality, | |
88acdbc2 | 1989 | 'has_drm': has_drm, |
60755938 | 1990 | } for idx in _extract_m3u8_playlist_indices(m3u8_doc=m3u8_doc)] |
310c2ed2 | 1991 | |
a0c3b2d5 | 1992 | return formats, subtitles |
cb252080 S |
1993 | |
1994 | groups = {} | |
1995 | last_stream_inf = {} | |
1996 | ||
1997 | def extract_media(x_media_line): | |
1998 | media = parse_m3u8_attributes(x_media_line) | |
1999 | # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED | |
2000 | media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME') | |
2001 | if not (media_type and group_id and name): | |
2002 | return | |
2003 | groups.setdefault(group_id, []).append(media) | |
a0c3b2d5 F |
2004 | # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1> |
2005 | if media_type == 'SUBTITLES': | |
3907333c | 2006 | # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the |
2007 | # EXT-X-MEDIA tag if the media type is SUBTITLES. | |
2008 | # However, lack of URI has been spotted in the wild. | |
2009 | # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339 | |
2010 | if not media.get('URI'): | |
2011 | return | |
a0c3b2d5 F |
2012 | url = format_url(media['URI']) |
2013 | sub_info = { | |
2014 | 'url': url, | |
2015 | 'ext': determine_ext(url), | |
2016 | } | |
4a2f19ab F |
2017 | if sub_info['ext'] == 'm3u8': |
2018 | # Per RFC 8216 §3.1, the only possible subtitle format m3u8 | |
2019 | # files may contain is WebVTT: | |
2020 | # <https://tools.ietf.org/html/rfc8216#section-3.1> | |
2021 | sub_info['ext'] = 'vtt' | |
2022 | sub_info['protocol'] = 'm3u8_native' | |
37a3bb66 | 2023 | lang = media.get('LANGUAGE') or 'und' |
a0c3b2d5 | 2024 | subtitles.setdefault(lang, []).append(sub_info) |
cb252080 S |
2025 | if media_type not in ('VIDEO', 'AUDIO'): |
2026 | return | |
2027 | media_url = media.get('URI') | |
2028 | if media_url: | |
310c2ed2 | 2029 | manifest_url = format_url(media_url) |
60755938 | 2030 | formats.extend({ |
34921b43 | 2031 | 'format_id': join_nonempty(m3u8_id, group_id, name, idx), |
60755938 | 2032 | 'format_note': name, |
2033 | 'format_index': idx, | |
2034 | 'url': manifest_url, | |
2035 | 'manifest_url': m3u8_url, | |
2036 | 'language': media.get('LANGUAGE'), | |
2037 | 'ext': ext, | |
2038 | 'protocol': entry_protocol, | |
2039 | 'preference': preference, | |
2040 | 'quality': quality, | |
2041 | 'vcodec': 'none' if media_type == 'AUDIO' else None, | |
2042 | } for idx in _extract_m3u8_playlist_indices(manifest_url)) | |
cb252080 S |
2043 | |
2044 | def build_stream_name(): | |
2045 | # Despite specification does not mention NAME attribute for | |
3019cb0c S |
2046 | # EXT-X-STREAM-INF tag it still sometimes may be present (see [1] |
2047 | # or vidio test in TestInfoExtractor.test_parse_m3u8_formats) | |
ddd258f9 | 2048 | # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015 |
cb252080 S |
2049 | stream_name = last_stream_inf.get('NAME') |
2050 | if stream_name: | |
2051 | return stream_name | |
2052 | # If there is no NAME in EXT-X-STREAM-INF it will be obtained | |
2053 | # from corresponding rendition group | |
2054 | stream_group_id = last_stream_inf.get('VIDEO') | |
2055 | if not stream_group_id: | |
2056 | return | |
2057 | stream_group = groups.get(stream_group_id) | |
2058 | if not stream_group: | |
2059 | return stream_group_id | |
2060 | rendition = stream_group[0] | |
2061 | return rendition.get('NAME') or stream_group_id | |
2062 | ||
379306ef | 2063 | # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the |
2bfc1d9d RA |
2064 | # chance to detect video only formats when EXT-X-STREAM-INF tags |
2065 | # precede EXT-X-MEDIA tags in HLS manifest such as [3]. | |
2066 | for line in m3u8_doc.splitlines(): | |
2067 | if line.startswith('#EXT-X-MEDIA:'): | |
2068 | extract_media(line) | |
2069 | ||
704df56d PH |
2070 | for line in m3u8_doc.splitlines(): |
2071 | if line.startswith('#EXT-X-STREAM-INF:'): | |
cb252080 | 2072 | last_stream_inf = parse_m3u8_attributes(line) |
704df56d PH |
2073 | elif line.startswith('#') or not line.strip(): |
2074 | continue | |
2075 | else: | |
9c99bef7 | 2076 | tbr = float_or_none( |
3089bc74 S |
2077 | last_stream_inf.get('AVERAGE-BANDWIDTH') |
2078 | or last_stream_inf.get('BANDWIDTH'), scale=1000) | |
30d0b549 | 2079 | manifest_url = format_url(line.strip()) |
5ef62fc4 | 2080 | |
60755938 | 2081 | for idx in _extract_m3u8_playlist_indices(manifest_url): |
2082 | format_id = [m3u8_id, None, idx] | |
310c2ed2 | 2083 | # Bandwidth of live streams may differ over time thus making |
2084 | # format_id unpredictable. So it's better to keep provided | |
2085 | # format_id intact. | |
2086 | if not live: | |
60755938 | 2087 | stream_name = build_stream_name() |
34921b43 | 2088 | format_id[1] = stream_name or '%d' % (tbr or len(formats)) |
310c2ed2 | 2089 | f = { |
34921b43 | 2090 | 'format_id': join_nonempty(*format_id), |
60755938 | 2091 | 'format_index': idx, |
310c2ed2 | 2092 | 'url': manifest_url, |
2093 | 'manifest_url': m3u8_url, | |
2094 | 'tbr': tbr, | |
2095 | 'ext': ext, | |
2096 | 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')), | |
2097 | 'protocol': entry_protocol, | |
2098 | 'preference': preference, | |
2099 | 'quality': quality, | |
2100 | } | |
2101 | resolution = last_stream_inf.get('RESOLUTION') | |
2102 | if resolution: | |
2103 | mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution) | |
2104 | if mobj: | |
2105 | f['width'] = int(mobj.group('width')) | |
2106 | f['height'] = int(mobj.group('height')) | |
2107 | # Unified Streaming Platform | |
2108 | mobj = re.search( | |
2109 | r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url']) | |
2110 | if mobj: | |
2111 | abr, vbr = mobj.groups() | |
2112 | abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000) | |
2113 | f.update({ | |
2114 | 'vbr': vbr, | |
2115 | 'abr': abr, | |
2116 | }) | |
2117 | codecs = parse_codecs(last_stream_inf.get('CODECS')) | |
2118 | f.update(codecs) | |
2119 | audio_group_id = last_stream_inf.get('AUDIO') | |
2120 | # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which | |
2121 | # references a rendition group MUST have a CODECS attribute. | |
62b58c09 | 2122 | # However, this is not always respected. E.g. [2] |
310c2ed2 | 2123 | # contains EXT-X-STREAM-INF tag which references AUDIO |
2124 | # rendition group but does not have CODECS and despite | |
2125 | # referencing an audio group it represents a complete | |
2126 | # (with audio and video) format. So, for such cases we will | |
2127 | # ignore references to rendition groups and treat them | |
2128 | # as complete formats. | |
2129 | if audio_group_id and codecs and f.get('vcodec') != 'none': | |
2130 | audio_group = groups.get(audio_group_id) | |
2131 | if audio_group and audio_group[0].get('URI'): | |
2132 | # TODO: update acodec for audio only formats with | |
2133 | # the same GROUP-ID | |
2134 | f['acodec'] = 'none' | |
fc21af50 | 2135 | if not f.get('ext'): |
2136 | f['ext'] = 'm4a' if f.get('vcodec') == 'none' else 'mp4' | |
310c2ed2 | 2137 | formats.append(f) |
2138 | ||
2139 | # for DailyMotion | |
2140 | progressive_uri = last_stream_inf.get('PROGRESSIVE-URI') | |
2141 | if progressive_uri: | |
2142 | http_f = f.copy() | |
2143 | del http_f['manifest_url'] | |
2144 | http_f.update({ | |
2145 | 'format_id': f['format_id'].replace('hls-', 'http-'), | |
2146 | 'protocol': 'http', | |
2147 | 'url': progressive_uri, | |
2148 | }) | |
2149 | formats.append(http_f) | |
5ef62fc4 | 2150 | |
cb252080 | 2151 | last_stream_inf = {} |
a0c3b2d5 | 2152 | return formats, subtitles |
704df56d | 2153 | |
3cf4b91d C |
2154 | def _extract_m3u8_vod_duration( |
2155 | self, m3u8_vod_url, video_id, note=None, errnote=None, data=None, headers={}, query={}): | |
2156 | ||
2157 | m3u8_vod = self._download_webpage( | |
2158 | m3u8_vod_url, video_id, | |
2159 | note='Downloading m3u8 VOD manifest' if note is None else note, | |
2160 | errnote='Failed to download VOD manifest' if errnote is None else errnote, | |
2161 | fatal=False, data=data, headers=headers, query=query) | |
2162 | ||
2163 | return self._parse_m3u8_vod_duration(m3u8_vod or '', video_id) | |
2164 | ||
2165 | def _parse_m3u8_vod_duration(self, m3u8_vod, video_id): | |
2166 | if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod: | |
2167 | return None | |
2168 | ||
2169 | return int(sum( | |
2170 | float(line[len('#EXTINF:'):].split(',')[0]) | |
2171 | for line in m3u8_vod.splitlines() if line.startswith('#EXTINF:'))) or None | |
2172 | ||
a107193e S |
2173 | @staticmethod |
2174 | def _xpath_ns(path, namespace=None): | |
2175 | if not namespace: | |
2176 | return path | |
2177 | out = [] | |
2178 | for c in path.split('/'): | |
2179 | if not c or c == '.': | |
2180 | out.append(c) | |
2181 | else: | |
2182 | out.append('{%s}%s' % (namespace, c)) | |
2183 | return '/'.join(out) | |
2184 | ||
da1c94ee | 2185 | def _extract_smil_formats_and_subtitles(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None): |
a076c1f9 E |
2186 | res = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source) |
2187 | if res is False: | |
995029a1 | 2188 | assert not fatal |
774a46c5 | 2189 | return [], {} |
e89a2aab | 2190 | |
a076c1f9 E |
2191 | smil, urlh = res |
2192 | smil_url = urlh.geturl() | |
2193 | ||
17712eeb | 2194 | namespace = self._parse_smil_namespace(smil) |
a107193e | 2195 | |
da1c94ee | 2196 | fmts = self._parse_smil_formats( |
a107193e | 2197 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) |
da1c94ee F |
2198 | subs = self._parse_smil_subtitles( |
2199 | smil, namespace=namespace) | |
2200 | ||
2201 | return fmts, subs | |
2202 | ||
2203 | def _extract_smil_formats(self, *args, **kwargs): | |
2204 | fmts, subs = self._extract_smil_formats_and_subtitles(*args, **kwargs) | |
2205 | if subs: | |
b5ae35ee | 2206 | self._report_ignoring_subs('SMIL') |
da1c94ee | 2207 | return fmts |
a107193e S |
2208 | |
2209 | def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None): | |
a076c1f9 E |
2210 | res = self._download_smil(smil_url, video_id, fatal=fatal) |
2211 | if res is False: | |
a107193e | 2212 | return {} |
a076c1f9 E |
2213 | |
2214 | smil, urlh = res | |
2215 | smil_url = urlh.geturl() | |
2216 | ||
a107193e S |
2217 | return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params) |
2218 | ||
09f572fb | 2219 | def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None): |
a076c1f9 | 2220 | return self._download_xml_handle( |
a107193e | 2221 | smil_url, video_id, 'Downloading SMIL file', |
09f572fb | 2222 | 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source) |
a107193e S |
2223 | |
2224 | def _parse_smil(self, smil, smil_url, video_id, f4m_params=None): | |
17712eeb | 2225 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
2226 | |
2227 | formats = self._parse_smil_formats( | |
2228 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
2229 | subtitles = self._parse_smil_subtitles(smil, namespace=namespace) | |
2230 | ||
2231 | video_id = os.path.splitext(url_basename(smil_url))[0] | |
2232 | title = None | |
2233 | description = None | |
647eab45 | 2234 | upload_date = None |
a107193e S |
2235 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): |
2236 | name = meta.attrib.get('name') | |
2237 | content = meta.attrib.get('content') | |
2238 | if not name or not content: | |
2239 | continue | |
2240 | if not title and name == 'title': | |
2241 | title = content | |
2242 | elif not description and name in ('description', 'abstract'): | |
2243 | description = content | |
647eab45 S |
2244 | elif not upload_date and name == 'date': |
2245 | upload_date = unified_strdate(content) | |
a107193e | 2246 | |
1e5bcdec S |
2247 | thumbnails = [{ |
2248 | 'id': image.get('type'), | |
2249 | 'url': image.get('src'), | |
2250 | 'width': int_or_none(image.get('width')), | |
2251 | 'height': int_or_none(image.get('height')), | |
2252 | } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')] | |
2253 | ||
a107193e S |
2254 | return { |
2255 | 'id': video_id, | |
2256 | 'title': title or video_id, | |
2257 | 'description': description, | |
647eab45 | 2258 | 'upload_date': upload_date, |
1e5bcdec | 2259 | 'thumbnails': thumbnails, |
a107193e S |
2260 | 'formats': formats, |
2261 | 'subtitles': subtitles, | |
2262 | } | |
2263 | ||
17712eeb S |
2264 | def _parse_smil_namespace(self, smil): |
2265 | return self._search_regex( | |
2266 | r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None) | |
2267 | ||
f877c6ae | 2268 | def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None): |
a107193e S |
2269 | base = smil_url |
2270 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): | |
2271 | b = meta.get('base') or meta.get('httpBase') | |
2272 | if b: | |
2273 | base = b | |
2274 | break | |
e89a2aab S |
2275 | |
2276 | formats = [] | |
2277 | rtmp_count = 0 | |
a107193e | 2278 | http_count = 0 |
7f32e5dc | 2279 | m3u8_count = 0 |
9359f3d4 | 2280 | imgs_count = 0 |
a107193e | 2281 | |
9359f3d4 | 2282 | srcs = set() |
ad96b4c8 YCH |
2283 | media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace)) |
2284 | for medium in media: | |
2285 | src = medium.get('src') | |
81e1c4e2 | 2286 | if not src or src in srcs: |
a107193e | 2287 | continue |
9359f3d4 | 2288 | srcs.add(src) |
a107193e | 2289 | |
ad96b4c8 YCH |
2290 | bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000) |
2291 | filesize = int_or_none(medium.get('size') or medium.get('fileSize')) | |
2292 | width = int_or_none(medium.get('width')) | |
2293 | height = int_or_none(medium.get('height')) | |
2294 | proto = medium.get('proto') | |
2295 | ext = medium.get('ext') | |
a107193e | 2296 | src_ext = determine_ext(src) |
ad96b4c8 | 2297 | streamer = medium.get('streamer') or base |
a107193e S |
2298 | |
2299 | if proto == 'rtmp' or streamer.startswith('rtmp'): | |
2300 | rtmp_count += 1 | |
2301 | formats.append({ | |
2302 | 'url': streamer, | |
2303 | 'play_path': src, | |
2304 | 'ext': 'flv', | |
2305 | 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), | |
2306 | 'tbr': bitrate, | |
2307 | 'filesize': filesize, | |
2308 | 'width': width, | |
2309 | 'height': height, | |
2310 | }) | |
f877c6ae YCH |
2311 | if transform_rtmp_url: |
2312 | streamer, src = transform_rtmp_url(streamer, src) | |
2313 | formats[-1].update({ | |
2314 | 'url': streamer, | |
2315 | 'play_path': src, | |
2316 | }) | |
a107193e S |
2317 | continue |
2318 | ||
14f25df2 | 2319 | src_url = src if src.startswith('http') else urllib.parse.urljoin(base, src) |
c349456e | 2320 | src_url = src_url.strip() |
a107193e S |
2321 | |
2322 | if proto == 'm3u8' or src_ext == 'm3u8': | |
7f32e5dc | 2323 | m3u8_formats = self._extract_m3u8_formats( |
2324 | src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False) | |
2325 | if len(m3u8_formats) == 1: | |
2326 | m3u8_count += 1 | |
2327 | m3u8_formats[0].update({ | |
2328 | 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate), | |
2329 | 'tbr': bitrate, | |
2330 | 'width': width, | |
2331 | 'height': height, | |
2332 | }) | |
2333 | formats.extend(m3u8_formats) | |
bd21ead2 | 2334 | elif src_ext == 'f4m': |
a107193e S |
2335 | f4m_url = src_url |
2336 | if not f4m_params: | |
2337 | f4m_params = { | |
2338 | 'hdcore': '3.2.0', | |
2339 | 'plugin': 'flowplayer-3.2.0.1', | |
2340 | } | |
2341 | f4m_url += '&' if '?' in f4m_url else '?' | |
14f25df2 | 2342 | f4m_url += urllib.parse.urlencode(f4m_params) |
7e5edcfd | 2343 | formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False)) |
bd21ead2 RA |
2344 | elif src_ext == 'mpd': |
2345 | formats.extend(self._extract_mpd_formats( | |
2346 | src_url, video_id, mpd_id='dash', fatal=False)) | |
2347 | elif re.search(r'\.ism/[Mm]anifest', src_url): | |
2348 | formats.extend(self._extract_ism_formats( | |
2349 | src_url, video_id, ism_id='mss', fatal=False)) | |
2350 | elif src_url.startswith('http') and self._is_valid_url(src, video_id): | |
a107193e S |
2351 | http_count += 1 |
2352 | formats.append({ | |
2353 | 'url': src_url, | |
2354 | 'ext': ext or src_ext or 'flv', | |
2355 | 'format_id': 'http-%d' % (bitrate or http_count), | |
2356 | 'tbr': bitrate, | |
2357 | 'filesize': filesize, | |
2358 | 'width': width, | |
2359 | 'height': height, | |
2360 | }) | |
63757032 | 2361 | |
9359f3d4 F |
2362 | for medium in smil.findall(self._xpath_ns('.//imagestream', namespace)): |
2363 | src = medium.get('src') | |
2364 | if not src or src in srcs: | |
2365 | continue | |
2366 | srcs.add(src) | |
2367 | ||
2368 | imgs_count += 1 | |
2369 | formats.append({ | |
2370 | 'format_id': 'imagestream-%d' % (imgs_count), | |
2371 | 'url': src, | |
2372 | 'ext': mimetype2ext(medium.get('type')), | |
2373 | 'acodec': 'none', | |
2374 | 'vcodec': 'none', | |
2375 | 'width': int_or_none(medium.get('width')), | |
2376 | 'height': int_or_none(medium.get('height')), | |
2377 | 'format_note': 'SMIL storyboards', | |
2378 | }) | |
2379 | ||
e89a2aab S |
2380 | return formats |
2381 | ||
ce00af87 | 2382 | def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'): |
d413095f | 2383 | urls = [] |
a107193e S |
2384 | subtitles = {} |
2385 | for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))): | |
2386 | src = textstream.get('src') | |
d413095f | 2387 | if not src or src in urls: |
a107193e | 2388 | continue |
d413095f | 2389 | urls.append(src) |
df634be2 | 2390 | ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src) |
03bc7237 | 2391 | lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang |
a107193e S |
2392 | subtitles.setdefault(lang, []).append({ |
2393 | 'url': src, | |
2394 | 'ext': ext, | |
2395 | }) | |
2396 | return subtitles | |
63757032 | 2397 | |
47a5cb77 | 2398 | def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True): |
a076c1f9 | 2399 | res = self._download_xml_handle( |
47a5cb77 | 2400 | xspf_url, playlist_id, 'Downloading xpsf playlist', |
942acef5 | 2401 | 'Unable to download xspf manifest', fatal=fatal) |
a076c1f9 | 2402 | if res is False: |
942acef5 | 2403 | return [] |
a076c1f9 E |
2404 | |
2405 | xspf, urlh = res | |
2406 | xspf_url = urlh.geturl() | |
2407 | ||
47a5cb77 S |
2408 | return self._parse_xspf( |
2409 | xspf, playlist_id, xspf_url=xspf_url, | |
2410 | xspf_base_url=base_url(xspf_url)) | |
8d6765cf | 2411 | |
47a5cb77 | 2412 | def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None): |
8d6765cf S |
2413 | NS_MAP = { |
2414 | 'xspf': 'http://xspf.org/ns/0/', | |
2415 | 's1': 'http://static.streamone.nl/player/ns/0', | |
2416 | } | |
2417 | ||
2418 | entries = [] | |
47a5cb77 | 2419 | for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)): |
8d6765cf | 2420 | title = xpath_text( |
98044462 | 2421 | track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id) |
8d6765cf S |
2422 | description = xpath_text( |
2423 | track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description') | |
2424 | thumbnail = xpath_text( | |
2425 | track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail') | |
2426 | duration = float_or_none( | |
2427 | xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000) | |
2428 | ||
47a5cb77 S |
2429 | formats = [] |
2430 | for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)): | |
2431 | format_url = urljoin(xspf_base_url, location.text) | |
2432 | if not format_url: | |
2433 | continue | |
2434 | formats.append({ | |
2435 | 'url': format_url, | |
2436 | 'manifest_url': xspf_url, | |
2437 | 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)), | |
2438 | 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))), | |
2439 | 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))), | |
2440 | }) | |
8d6765cf S |
2441 | |
2442 | entries.append({ | |
2443 | 'id': playlist_id, | |
2444 | 'title': title, | |
2445 | 'description': description, | |
2446 | 'thumbnail': thumbnail, | |
2447 | 'duration': duration, | |
2448 | 'formats': formats, | |
2449 | }) | |
2450 | return entries | |
2451 | ||
171e59ed F |
2452 | def _extract_mpd_formats(self, *args, **kwargs): |
2453 | fmts, subs = self._extract_mpd_formats_and_subtitles(*args, **kwargs) | |
2454 | if subs: | |
b5ae35ee | 2455 | self._report_ignoring_subs('DASH') |
171e59ed F |
2456 | return fmts |
2457 | ||
2458 | def _extract_mpd_formats_and_subtitles( | |
2459 | self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, | |
2460 | fatal=True, data=None, headers={}, query={}): | |
47a5cb77 | 2461 | res = self._download_xml_handle( |
1bac3455 | 2462 | mpd_url, video_id, |
37a3bb66 | 2463 | note='Downloading MPD manifest' if note is None else note, |
2464 | errnote='Failed to download MPD manifest' if errnote is None else errnote, | |
7360c06f | 2465 | fatal=fatal, data=data, headers=headers, query=query) |
1bac3455 | 2466 | if res is False: |
171e59ed | 2467 | return [], {} |
47a5cb77 | 2468 | mpd_doc, urlh = res |
c25720ef | 2469 | if mpd_doc is None: |
171e59ed | 2470 | return [], {} |
779da8e3 E |
2471 | |
2472 | # We could have been redirected to a new url when we retrieved our mpd file. | |
2473 | mpd_url = urlh.geturl() | |
2474 | mpd_base_url = base_url(mpd_url) | |
1bac3455 | 2475 | |
171e59ed | 2476 | return self._parse_mpd_formats_and_subtitles( |
545cc85d | 2477 | mpd_doc, mpd_id, mpd_base_url, mpd_url) |
2d2fa82d | 2478 | |
171e59ed F |
2479 | def _parse_mpd_formats(self, *args, **kwargs): |
2480 | fmts, subs = self._parse_mpd_formats_and_subtitles(*args, **kwargs) | |
2481 | if subs: | |
b5ae35ee | 2482 | self._report_ignoring_subs('DASH') |
171e59ed F |
2483 | return fmts |
2484 | ||
2485 | def _parse_mpd_formats_and_subtitles( | |
2486 | self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None): | |
f0948348 S |
2487 | """ |
2488 | Parse formats from MPD manifest. | |
2489 | References: | |
2490 | 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E), | |
2491 | http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip | |
2492 | 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP | |
2493 | """ | |
a06916d9 | 2494 | if not self.get_param('dynamic_mpd', True): |
78895bd3 | 2495 | if mpd_doc.get('type') == 'dynamic': |
171e59ed | 2496 | return [], {} |
2d2fa82d | 2497 | |
91cb6b50 | 2498 | namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None) |
f14be228 | 2499 | |
2500 | def _add_ns(path): | |
2501 | return self._xpath_ns(path, namespace) | |
2502 | ||
675d0016 | 2503 | def is_drm_protected(element): |
2504 | return element.find(_add_ns('ContentProtection')) is not None | |
2505 | ||
1bac3455 | 2506 | def extract_multisegment_info(element, ms_parent_info): |
2507 | ms_info = ms_parent_info.copy() | |
b4c1d6e8 S |
2508 | |
2509 | # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some | |
2510 | # common attributes and elements. We will only extract relevant | |
2511 | # for us. | |
2512 | def extract_common(source): | |
2513 | segment_timeline = source.find(_add_ns('SegmentTimeline')) | |
2514 | if segment_timeline is not None: | |
2515 | s_e = segment_timeline.findall(_add_ns('S')) | |
2516 | if s_e: | |
2517 | ms_info['total_number'] = 0 | |
2518 | ms_info['s'] = [] | |
2519 | for s in s_e: | |
2520 | r = int(s.get('r', 0)) | |
2521 | ms_info['total_number'] += 1 + r | |
2522 | ms_info['s'].append({ | |
2523 | 't': int(s.get('t', 0)), | |
2524 | # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60]) | |
2525 | 'd': int(s.attrib['d']), | |
2526 | 'r': r, | |
2527 | }) | |
2528 | start_number = source.get('startNumber') | |
2529 | if start_number: | |
2530 | ms_info['start_number'] = int(start_number) | |
2531 | timescale = source.get('timescale') | |
2532 | if timescale: | |
2533 | ms_info['timescale'] = int(timescale) | |
2534 | segment_duration = source.get('duration') | |
2535 | if segment_duration: | |
48504785 | 2536 | ms_info['segment_duration'] = float(segment_duration) |
b4c1d6e8 S |
2537 | |
2538 | def extract_Initialization(source): | |
2539 | initialization = source.find(_add_ns('Initialization')) | |
2540 | if initialization is not None: | |
2541 | ms_info['initialization_url'] = initialization.attrib['sourceURL'] | |
2542 | ||
f14be228 | 2543 | segment_list = element.find(_add_ns('SegmentList')) |
1bac3455 | 2544 | if segment_list is not None: |
b4c1d6e8 S |
2545 | extract_common(segment_list) |
2546 | extract_Initialization(segment_list) | |
f14be228 | 2547 | segment_urls_e = segment_list.findall(_add_ns('SegmentURL')) |
1bac3455 | 2548 | if segment_urls_e: |
2549 | ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e] | |
1bac3455 | 2550 | else: |
f14be228 | 2551 | segment_template = element.find(_add_ns('SegmentTemplate')) |
1bac3455 | 2552 | if segment_template is not None: |
b4c1d6e8 | 2553 | extract_common(segment_template) |
e228616c S |
2554 | media = segment_template.get('media') |
2555 | if media: | |
2556 | ms_info['media'] = media | |
1bac3455 | 2557 | initialization = segment_template.get('initialization') |
2558 | if initialization: | |
e228616c | 2559 | ms_info['initialization'] = initialization |
1bac3455 | 2560 | else: |
b4c1d6e8 | 2561 | extract_Initialization(segment_template) |
1bac3455 | 2562 | return ms_info |
b323e170 | 2563 | |
1bac3455 | 2564 | mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration')) |
6251555f | 2565 | formats, subtitles = [], {} |
234416e4 | 2566 | stream_numbers = collections.defaultdict(int) |
f14be228 | 2567 | for period in mpd_doc.findall(_add_ns('Period')): |
1bac3455 | 2568 | period_duration = parse_duration(period.get('duration')) or mpd_duration |
2569 | period_ms_info = extract_multisegment_info(period, { | |
2570 | 'start_number': 1, | |
2571 | 'timescale': 1, | |
2572 | }) | |
f14be228 | 2573 | for adaptation_set in period.findall(_add_ns('AdaptationSet')): |
1bac3455 | 2574 | adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info) |
f14be228 | 2575 | for representation in adaptation_set.findall(_add_ns('Representation')): |
1bac3455 | 2576 | representation_attrib = adaptation_set.attrib.copy() |
2577 | representation_attrib.update(representation.attrib) | |
f0948348 | 2578 | # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory |
a6c8b759 | 2579 | mime_type = representation_attrib['mimeType'] |
171e59ed F |
2580 | content_type = representation_attrib.get('contentType', mime_type.split('/')[0]) |
2581 | ||
21633673 | 2582 | codec_str = representation_attrib.get('codecs', '') |
2583 | # Some kind of binary subtitle found in some youtube livestreams | |
2584 | if mime_type == 'application/x-rawcc': | |
2585 | codecs = {'scodec': codec_str} | |
2586 | else: | |
2587 | codecs = parse_codecs(codec_str) | |
be2fc5b2 | 2588 | if content_type not in ('video', 'audio', 'text'): |
2589 | if mime_type == 'image/jpeg': | |
a8731fcc | 2590 | content_type = mime_type |
21633673 | 2591 | elif codecs.get('vcodec', 'none') != 'none': |
4afa3ec4 | 2592 | content_type = 'video' |
21633673 | 2593 | elif codecs.get('acodec', 'none') != 'none': |
4afa3ec4 | 2594 | content_type = 'audio' |
3fe75fdc | 2595 | elif codecs.get('scodec', 'none') != 'none': |
be2fc5b2 | 2596 | content_type = 'text' |
6993f78d | 2597 | elif mimetype2ext(mime_type) in ('tt', 'dfxp', 'ttml', 'xml', 'json'): |
2598 | content_type = 'text' | |
cdb19aa4 | 2599 | else: |
be2fc5b2 | 2600 | self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) |
2601 | continue | |
2602 | ||
2603 | base_url = '' | |
2604 | for element in (representation, adaptation_set, period, mpd_doc): | |
2605 | base_url_e = element.find(_add_ns('BaseURL')) | |
47046464 | 2606 | if try_call(lambda: base_url_e.text) is not None: |
be2fc5b2 | 2607 | base_url = base_url_e.text + base_url |
2608 | if re.match(r'^https?://', base_url): | |
2609 | break | |
f9cc0161 | 2610 | if mpd_base_url and base_url.startswith('/'): |
14f25df2 | 2611 | base_url = urllib.parse.urljoin(mpd_base_url, base_url) |
f9cc0161 D |
2612 | elif mpd_base_url and not re.match(r'^https?://', base_url): |
2613 | if not mpd_base_url.endswith('/'): | |
be2fc5b2 | 2614 | mpd_base_url += '/' |
2615 | base_url = mpd_base_url + base_url | |
2616 | representation_id = representation_attrib.get('id') | |
2617 | lang = representation_attrib.get('lang') | |
2618 | url_el = representation.find(_add_ns('BaseURL')) | |
2619 | filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None) | |
2620 | bandwidth = int_or_none(representation_attrib.get('bandwidth')) | |
2621 | if representation_id is not None: | |
2622 | format_id = representation_id | |
2623 | else: | |
2624 | format_id = content_type | |
2625 | if mpd_id: | |
2626 | format_id = mpd_id + '-' + format_id | |
2627 | if content_type in ('video', 'audio'): | |
2628 | f = { | |
2629 | 'format_id': format_id, | |
2630 | 'manifest_url': mpd_url, | |
2631 | 'ext': mimetype2ext(mime_type), | |
2632 | 'width': int_or_none(representation_attrib.get('width')), | |
2633 | 'height': int_or_none(representation_attrib.get('height')), | |
2634 | 'tbr': float_or_none(bandwidth, 1000), | |
2635 | 'asr': int_or_none(representation_attrib.get('audioSamplingRate')), | |
2636 | 'fps': int_or_none(representation_attrib.get('frameRate')), | |
2637 | 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None, | |
2638 | 'format_note': 'DASH %s' % content_type, | |
2639 | 'filesize': filesize, | |
2640 | 'container': mimetype2ext(mime_type) + '_dash', | |
4afa3ec4 | 2641 | **codecs |
be2fc5b2 | 2642 | } |
be2fc5b2 | 2643 | elif content_type == 'text': |
2644 | f = { | |
2645 | 'ext': mimetype2ext(mime_type), | |
2646 | 'manifest_url': mpd_url, | |
2647 | 'filesize': filesize, | |
2648 | } | |
2649 | elif content_type == 'image/jpeg': | |
2650 | # See test case in VikiIE | |
2651 | # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1 | |
2652 | f = { | |
2653 | 'format_id': format_id, | |
2654 | 'ext': 'mhtml', | |
2655 | 'manifest_url': mpd_url, | |
2656 | 'format_note': 'DASH storyboards (jpeg)', | |
2657 | 'acodec': 'none', | |
2658 | 'vcodec': 'none', | |
2659 | } | |
88acdbc2 | 2660 | if is_drm_protected(adaptation_set) or is_drm_protected(representation): |
2661 | f['has_drm'] = True | |
be2fc5b2 | 2662 | representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info) |
2663 | ||
2664 | def prepare_template(template_name, identifiers): | |
2665 | tmpl = representation_ms_info[template_name] | |
0cb0fdbb | 2666 | if representation_id is not None: |
2667 | tmpl = tmpl.replace('$RepresentationID$', representation_id) | |
be2fc5b2 | 2668 | # First of, % characters outside $...$ templates |
2669 | # must be escaped by doubling for proper processing | |
2670 | # by % operator string formatting used further (see | |
2671 | # https://github.com/ytdl-org/youtube-dl/issues/16867). | |
2672 | t = '' | |
2673 | in_template = False | |
2674 | for c in tmpl: | |
2675 | t += c | |
2676 | if c == '$': | |
2677 | in_template = not in_template | |
2678 | elif c == '%' and not in_template: | |
eca1f0d1 | 2679 | t += c |
be2fc5b2 | 2680 | # Next, $...$ templates are translated to their |
2681 | # %(...) counterparts to be used with % operator | |
be2fc5b2 | 2682 | t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t) |
2683 | t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t) | |
2684 | t.replace('$$', '$') | |
2685 | return t | |
2686 | ||
2687 | # @initialization is a regular template like @media one | |
2688 | # so it should be handled just the same way (see | |
2689 | # https://github.com/ytdl-org/youtube-dl/issues/11605) | |
2690 | if 'initialization' in representation_ms_info: | |
2691 | initialization_template = prepare_template( | |
2692 | 'initialization', | |
2693 | # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and | |
2694 | # $Time$ shall not be included for @initialization thus | |
2695 | # only $Bandwidth$ remains | |
2696 | ('Bandwidth', )) | |
2697 | representation_ms_info['initialization_url'] = initialization_template % { | |
2698 | 'Bandwidth': bandwidth, | |
2699 | } | |
2700 | ||
2701 | def location_key(location): | |
2702 | return 'url' if re.match(r'^https?://', location) else 'path' | |
2703 | ||
2704 | if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info: | |
2705 | ||
2706 | media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time')) | |
2707 | media_location_key = location_key(media_template) | |
2708 | ||
2709 | # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$ | |
2710 | # can't be used at the same time | |
2711 | if '%(Number' in media_template and 's' not in representation_ms_info: | |
2712 | segment_duration = None | |
2713 | if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info: | |
2714 | segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale']) | |
ffa89477 | 2715 | representation_ms_info['total_number'] = int(math.ceil( |
2716 | float_or_none(period_duration, segment_duration, default=0))) | |
be2fc5b2 | 2717 | representation_ms_info['fragments'] = [{ |
2718 | media_location_key: media_template % { | |
2719 | 'Number': segment_number, | |
2720 | 'Bandwidth': bandwidth, | |
2721 | }, | |
2722 | 'duration': segment_duration, | |
2723 | } for segment_number in range( | |
2724 | representation_ms_info['start_number'], | |
2725 | representation_ms_info['total_number'] + representation_ms_info['start_number'])] | |
2726 | else: | |
2727 | # $Number*$ or $Time$ in media template with S list available | |
2728 | # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg | |
2729 | # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411 | |
2730 | representation_ms_info['fragments'] = [] | |
2731 | segment_time = 0 | |
2732 | segment_d = None | |
2733 | segment_number = representation_ms_info['start_number'] | |
2734 | ||
2735 | def add_segment_url(): | |
2736 | segment_url = media_template % { | |
2737 | 'Time': segment_time, | |
2738 | 'Bandwidth': bandwidth, | |
2739 | 'Number': segment_number, | |
2740 | } | |
2741 | representation_ms_info['fragments'].append({ | |
2742 | media_location_key: segment_url, | |
2743 | 'duration': float_or_none(segment_d, representation_ms_info['timescale']), | |
2744 | }) | |
2745 | ||
2746 | for num, s in enumerate(representation_ms_info['s']): | |
2747 | segment_time = s.get('t') or segment_time | |
2748 | segment_d = s['d'] | |
2749 | add_segment_url() | |
2750 | segment_number += 1 | |
2751 | for r in range(s.get('r', 0)): | |
2752 | segment_time += segment_d | |
f0948348 | 2753 | add_segment_url() |
b4c1d6e8 | 2754 | segment_number += 1 |
be2fc5b2 | 2755 | segment_time += segment_d |
2756 | elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info: | |
62b58c09 L |
2757 | # No media template, |
2758 | # e.g. https://www.youtube.com/watch?v=iXZV5uAYMJI | |
be2fc5b2 | 2759 | # or any YouTube dashsegments video |
2760 | fragments = [] | |
2761 | segment_index = 0 | |
2762 | timescale = representation_ms_info['timescale'] | |
2763 | for s in representation_ms_info['s']: | |
2764 | duration = float_or_none(s['d'], timescale) | |
2765 | for r in range(s.get('r', 0) + 1): | |
2766 | segment_uri = representation_ms_info['segment_urls'][segment_index] | |
2767 | fragments.append({ | |
2768 | location_key(segment_uri): segment_uri, | |
2769 | 'duration': duration, | |
2770 | }) | |
2771 | segment_index += 1 | |
2772 | representation_ms_info['fragments'] = fragments | |
2773 | elif 'segment_urls' in representation_ms_info: | |
2774 | # Segment URLs with no SegmentTimeline | |
62b58c09 | 2775 | # E.g. https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091 |
be2fc5b2 | 2776 | # https://github.com/ytdl-org/youtube-dl/pull/14844 |
2777 | fragments = [] | |
2778 | segment_duration = float_or_none( | |
2779 | representation_ms_info['segment_duration'], | |
2780 | representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None | |
2781 | for segment_url in representation_ms_info['segment_urls']: | |
2782 | fragment = { | |
2783 | location_key(segment_url): segment_url, | |
2784 | } | |
2785 | if segment_duration: | |
2786 | fragment['duration'] = segment_duration | |
2787 | fragments.append(fragment) | |
2788 | representation_ms_info['fragments'] = fragments | |
2789 | # If there is a fragments key available then we correctly recognized fragmented media. | |
2790 | # Otherwise we will assume unfragmented media with direct access. Technically, such | |
2791 | # assumption is not necessarily correct since we may simply have no support for | |
2792 | # some forms of fragmented media renditions yet, but for now we'll use this fallback. | |
2793 | if 'fragments' in representation_ms_info: | |
2794 | f.update({ | |
2795 | # NB: mpd_url may be empty when MPD manifest is parsed from a string | |
2796 | 'url': mpd_url or base_url, | |
2797 | 'fragment_base_url': base_url, | |
2798 | 'fragments': [], | |
2799 | 'protocol': 'http_dash_segments' if mime_type != 'image/jpeg' else 'mhtml', | |
2800 | }) | |
2801 | if 'initialization_url' in representation_ms_info: | |
2802 | initialization_url = representation_ms_info['initialization_url'] | |
2803 | if not f.get('url'): | |
2804 | f['url'] = initialization_url | |
2805 | f['fragments'].append({location_key(initialization_url): initialization_url}) | |
2806 | f['fragments'].extend(representation_ms_info['fragments']) | |
ffa89477 | 2807 | if not period_duration: |
2808 | period_duration = try_get( | |
2809 | representation_ms_info, | |
2810 | lambda r: sum(frag['duration'] for frag in r['fragments']), float) | |
17b598d3 | 2811 | else: |
be2fc5b2 | 2812 | # Assuming direct URL to unfragmented media. |
2813 | f['url'] = base_url | |
234416e4 | 2814 | if content_type in ('video', 'audio', 'image/jpeg'): |
2815 | f['manifest_stream_number'] = stream_numbers[f['url']] | |
2816 | stream_numbers[f['url']] += 1 | |
be2fc5b2 | 2817 | formats.append(f) |
2818 | elif content_type == 'text': | |
2819 | subtitles.setdefault(lang or 'und', []).append(f) | |
2820 | ||
171e59ed | 2821 | return formats, subtitles |
17b598d3 | 2822 | |
fd76a142 F |
2823 | def _extract_ism_formats(self, *args, **kwargs): |
2824 | fmts, subs = self._extract_ism_formats_and_subtitles(*args, **kwargs) | |
2825 | if subs: | |
b5ae35ee | 2826 | self._report_ignoring_subs('ISM') |
fd76a142 F |
2827 | return fmts |
2828 | ||
2829 | def _extract_ism_formats_and_subtitles(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}): | |
47a5cb77 | 2830 | res = self._download_xml_handle( |
b2758123 | 2831 | ism_url, video_id, |
37a3bb66 | 2832 | note='Downloading ISM manifest' if note is None else note, |
2833 | errnote='Failed to download ISM manifest' if errnote is None else errnote, | |
7360c06f | 2834 | fatal=fatal, data=data, headers=headers, query=query) |
b2758123 | 2835 | if res is False: |
fd76a142 | 2836 | return [], {} |
47a5cb77 | 2837 | ism_doc, urlh = res |
13b08034 | 2838 | if ism_doc is None: |
fd76a142 | 2839 | return [], {} |
b2758123 | 2840 | |
fd76a142 | 2841 | return self._parse_ism_formats_and_subtitles(ism_doc, urlh.geturl(), ism_id) |
b2758123 | 2842 | |
fd76a142 | 2843 | def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None): |
76d5a363 S |
2844 | """ |
2845 | Parse formats from ISM manifest. | |
2846 | References: | |
2847 | 1. [MS-SSTR]: Smooth Streaming Protocol, | |
2848 | https://msdn.microsoft.com/en-us/library/ff469518.aspx | |
2849 | """ | |
06869367 | 2850 | if ism_doc.get('IsLive') == 'TRUE': |
fd76a142 | 2851 | return [], {} |
b2758123 | 2852 | |
b2758123 RA |
2853 | duration = int(ism_doc.attrib['Duration']) |
2854 | timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000 | |
2855 | ||
2856 | formats = [] | |
fd76a142 | 2857 | subtitles = {} |
b2758123 RA |
2858 | for stream in ism_doc.findall('StreamIndex'): |
2859 | stream_type = stream.get('Type') | |
fd76a142 | 2860 | if stream_type not in ('video', 'audio', 'text'): |
b2758123 RA |
2861 | continue |
2862 | url_pattern = stream.attrib['Url'] | |
2863 | stream_timescale = int_or_none(stream.get('TimeScale')) or timescale | |
2864 | stream_name = stream.get('Name') | |
fd76a142 | 2865 | stream_language = stream.get('Language', 'und') |
b2758123 | 2866 | for track in stream.findall('QualityLevel'): |
81b6102d | 2867 | KNOWN_TAGS = {'255': 'AACL', '65534': 'EC-3'} |
2868 | fourcc = track.get('FourCC') or KNOWN_TAGS.get(track.get('AudioTag')) | |
b2758123 | 2869 | # TODO: add support for WVC1 and WMAP |
81b6102d | 2870 | if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML', 'EC-3'): |
b2758123 RA |
2871 | self.report_warning('%s is not a supported codec' % fourcc) |
2872 | continue | |
2873 | tbr = int(track.attrib['Bitrate']) // 1000 | |
76d5a363 S |
2874 | # [1] does not mention Width and Height attributes. However, |
2875 | # they're often present while MaxWidth and MaxHeight are | |
2876 | # missing, so should be used as fallbacks | |
2877 | width = int_or_none(track.get('MaxWidth') or track.get('Width')) | |
2878 | height = int_or_none(track.get('MaxHeight') or track.get('Height')) | |
b2758123 RA |
2879 | sampling_rate = int_or_none(track.get('SamplingRate')) |
2880 | ||
2881 | track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern) | |
14f25df2 | 2882 | track_url_pattern = urllib.parse.urljoin(ism_url, track_url_pattern) |
b2758123 RA |
2883 | |
2884 | fragments = [] | |
2885 | fragment_ctx = { | |
2886 | 'time': 0, | |
2887 | } | |
2888 | stream_fragments = stream.findall('c') | |
2889 | for stream_fragment_index, stream_fragment in enumerate(stream_fragments): | |
2890 | fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time'] | |
2891 | fragment_repeat = int_or_none(stream_fragment.get('r')) or 1 | |
2892 | fragment_ctx['duration'] = int_or_none(stream_fragment.get('d')) | |
2893 | if not fragment_ctx['duration']: | |
2894 | try: | |
2895 | next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t']) | |
2896 | except IndexError: | |
2897 | next_fragment_time = duration | |
1616f9b4 | 2898 | fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat |
b2758123 RA |
2899 | for _ in range(fragment_repeat): |
2900 | fragments.append({ | |
14f25df2 | 2901 | 'url': re.sub(r'{start[ _]time}', str(fragment_ctx['time']), track_url_pattern), |
b2758123 RA |
2902 | 'duration': fragment_ctx['duration'] / stream_timescale, |
2903 | }) | |
2904 | fragment_ctx['time'] += fragment_ctx['duration'] | |
2905 | ||
fd76a142 F |
2906 | if stream_type == 'text': |
2907 | subtitles.setdefault(stream_language, []).append({ | |
2908 | 'ext': 'ismt', | |
2909 | 'protocol': 'ism', | |
2910 | 'url': ism_url, | |
2911 | 'manifest_url': ism_url, | |
2912 | 'fragments': fragments, | |
2913 | '_download_params': { | |
2914 | 'stream_type': stream_type, | |
2915 | 'duration': duration, | |
2916 | 'timescale': stream_timescale, | |
2917 | 'fourcc': fourcc, | |
2918 | 'language': stream_language, | |
2919 | 'codec_private_data': track.get('CodecPrivateData'), | |
2920 | } | |
2921 | }) | |
2922 | elif stream_type in ('video', 'audio'): | |
2923 | formats.append({ | |
34921b43 | 2924 | 'format_id': join_nonempty(ism_id, stream_name, tbr), |
fd76a142 F |
2925 | 'url': ism_url, |
2926 | 'manifest_url': ism_url, | |
2927 | 'ext': 'ismv' if stream_type == 'video' else 'isma', | |
2928 | 'width': width, | |
2929 | 'height': height, | |
2930 | 'tbr': tbr, | |
2931 | 'asr': sampling_rate, | |
2932 | 'vcodec': 'none' if stream_type == 'audio' else fourcc, | |
2933 | 'acodec': 'none' if stream_type == 'video' else fourcc, | |
2934 | 'protocol': 'ism', | |
2935 | 'fragments': fragments, | |
88acdbc2 | 2936 | 'has_drm': ism_doc.find('Protection') is not None, |
fd76a142 F |
2937 | '_download_params': { |
2938 | 'stream_type': stream_type, | |
2939 | 'duration': duration, | |
2940 | 'timescale': stream_timescale, | |
2941 | 'width': width or 0, | |
2942 | 'height': height or 0, | |
2943 | 'fourcc': fourcc, | |
2944 | 'language': stream_language, | |
2945 | 'codec_private_data': track.get('CodecPrivateData'), | |
2946 | 'sampling_rate': sampling_rate, | |
2947 | 'channels': int_or_none(track.get('Channels', 2)), | |
2948 | 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)), | |
2949 | 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)), | |
2950 | }, | |
2951 | }) | |
2952 | return formats, subtitles | |
b2758123 | 2953 | |
079a7cfc | 2954 | def _parse_html5_media_entries(self, base_url, webpage, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8_native', mpd_id=None, preference=None, quality=None): |
6780154e S |
2955 | def absolute_url(item_url): |
2956 | return urljoin(base_url, item_url) | |
59bbe491 | 2957 | |
2958 | def parse_content_type(content_type): | |
2959 | if not content_type: | |
2960 | return {} | |
2961 | ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type) | |
2962 | if ctr: | |
2963 | mimetype, codecs = ctr.groups() | |
2964 | f = parse_codecs(codecs) | |
2965 | f['ext'] = mimetype2ext(mimetype) | |
2966 | return f | |
2967 | return {} | |
2968 | ||
222a2308 L |
2969 | def _media_formats(src, cur_media_type, type_info=None): |
2970 | type_info = type_info or {} | |
520251c0 | 2971 | full_url = absolute_url(src) |
82889d4a | 2972 | ext = type_info.get('ext') or determine_ext(full_url) |
87a449c1 | 2973 | if ext == 'm3u8': |
520251c0 YCH |
2974 | is_plain_url = False |
2975 | formats = self._extract_m3u8_formats( | |
ad120ae1 | 2976 | full_url, video_id, ext='mp4', |
eeb0a956 | 2977 | entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id, |
f983b875 | 2978 | preference=preference, quality=quality, fatal=False) |
87a449c1 S |
2979 | elif ext == 'mpd': |
2980 | is_plain_url = False | |
2981 | formats = self._extract_mpd_formats( | |
b359e977 | 2982 | full_url, video_id, mpd_id=mpd_id, fatal=False) |
520251c0 YCH |
2983 | else: |
2984 | is_plain_url = True | |
2985 | formats = [{ | |
2986 | 'url': full_url, | |
2987 | 'vcodec': 'none' if cur_media_type == 'audio' else None, | |
222a2308 | 2988 | 'ext': ext, |
520251c0 YCH |
2989 | }] |
2990 | return is_plain_url, formats | |
2991 | ||
59bbe491 | 2992 | entries = [] |
4328ddf8 | 2993 | # amp-video and amp-audio are very similar to their HTML5 counterparts |
962ffcf8 | 2994 | # so we will include them right here (see |
4328ddf8 | 2995 | # https://www.ampproject.org/docs/reference/components/amp-video) |
29f7c58a | 2996 | # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/ |
2997 | _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)' | |
2998 | media_tags = [(media_tag, media_tag_name, media_type, '') | |
2999 | for media_tag, media_tag_name, media_type | |
3000 | in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)] | |
2aec7256 S |
3001 | media_tags.extend(re.findall( |
3002 | # We only allow video|audio followed by a whitespace or '>'. | |
3003 | # Allowing more characters may end up in significant slow down (see | |
62b58c09 L |
3004 | # https://github.com/ytdl-org/youtube-dl/issues/11979, |
3005 | # e.g. http://www.porntrex.com/maps/videositemap.xml). | |
29f7c58a | 3006 | r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage)) |
3007 | for media_tag, _, media_type, media_content in media_tags: | |
59bbe491 | 3008 | media_info = { |
3009 | 'formats': [], | |
3010 | 'subtitles': {}, | |
3011 | } | |
3012 | media_attributes = extract_attributes(media_tag) | |
bfbecd11 | 3013 | src = strip_or_none(dict_get(media_attributes, ('src', 'data-video-src', 'data-src', 'data-source'))) |
59bbe491 | 3014 | if src: |
222a2308 L |
3015 | f = parse_content_type(media_attributes.get('type')) |
3016 | _, formats = _media_formats(src, media_type, f) | |
520251c0 | 3017 | media_info['formats'].extend(formats) |
6780154e | 3018 | media_info['thumbnail'] = absolute_url(media_attributes.get('poster')) |
59bbe491 | 3019 | if media_content: |
3020 | for source_tag in re.findall(r'<source[^>]+>', media_content): | |
d493f15c S |
3021 | s_attr = extract_attributes(source_tag) |
3022 | # data-video-src and data-src are non standard but seen | |
3023 | # several times in the wild | |
bfbecd11 | 3024 | src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src', 'data-source'))) |
59bbe491 | 3025 | if not src: |
3026 | continue | |
d493f15c | 3027 | f = parse_content_type(s_attr.get('type')) |
868f79db | 3028 | is_plain_url, formats = _media_formats(src, media_type, f) |
520251c0 | 3029 | if is_plain_url: |
d493f15c S |
3030 | # width, height, res, label and title attributes are |
3031 | # all not standard but seen several times in the wild | |
3032 | labels = [ | |
3033 | s_attr.get(lbl) | |
3034 | for lbl in ('label', 'title') | |
3035 | if str_or_none(s_attr.get(lbl)) | |
3036 | ] | |
3037 | width = int_or_none(s_attr.get('width')) | |
3089bc74 S |
3038 | height = (int_or_none(s_attr.get('height')) |
3039 | or int_or_none(s_attr.get('res'))) | |
d493f15c S |
3040 | if not width or not height: |
3041 | for lbl in labels: | |
3042 | resolution = parse_resolution(lbl) | |
3043 | if not resolution: | |
3044 | continue | |
3045 | width = width or resolution.get('width') | |
3046 | height = height or resolution.get('height') | |
3047 | for lbl in labels: | |
3048 | tbr = parse_bitrate(lbl) | |
3049 | if tbr: | |
3050 | break | |
3051 | else: | |
3052 | tbr = None | |
1ed45499 | 3053 | f.update({ |
d493f15c S |
3054 | 'width': width, |
3055 | 'height': height, | |
3056 | 'tbr': tbr, | |
3057 | 'format_id': s_attr.get('label') or s_attr.get('title'), | |
1ed45499 | 3058 | }) |
520251c0 YCH |
3059 | f.update(formats[0]) |
3060 | media_info['formats'].append(f) | |
3061 | else: | |
3062 | media_info['formats'].extend(formats) | |
59bbe491 | 3063 | for track_tag in re.findall(r'<track[^>]+>', media_content): |
3064 | track_attributes = extract_attributes(track_tag) | |
3065 | kind = track_attributes.get('kind') | |
5968d7d2 | 3066 | if not kind or kind in ('subtitles', 'captions'): |
f856816b | 3067 | src = strip_or_none(track_attributes.get('src')) |
59bbe491 | 3068 | if not src: |
3069 | continue | |
3070 | lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label') | |
3071 | media_info['subtitles'].setdefault(lang, []).append({ | |
3072 | 'url': absolute_url(src), | |
3073 | }) | |
5e8e2fa5 S |
3074 | for f in media_info['formats']: |
3075 | f.setdefault('http_headers', {})['Referer'] = base_url | |
5968d7d2 | 3076 | if media_info['formats'] or media_info['subtitles']: |
59bbe491 | 3077 | entries.append(media_info) |
3078 | return entries | |
3079 | ||
f6a1d69a F |
3080 | def _extract_akamai_formats(self, *args, **kwargs): |
3081 | fmts, subs = self._extract_akamai_formats_and_subtitles(*args, **kwargs) | |
3082 | if subs: | |
b5ae35ee | 3083 | self._report_ignoring_subs('akamai') |
f6a1d69a F |
3084 | return fmts |
3085 | ||
3086 | def _extract_akamai_formats_and_subtitles(self, manifest_url, video_id, hosts={}): | |
29f7c58a | 3087 | signed = 'hdnea=' in manifest_url |
3088 | if not signed: | |
3089 | # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html | |
3090 | manifest_url = re.sub( | |
3091 | r'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?', | |
3092 | '', manifest_url).strip('?') | |
3093 | ||
c7c43a93 | 3094 | formats = [] |
f6a1d69a | 3095 | subtitles = {} |
70c5802b | 3096 | |
e71a4509 | 3097 | hdcore_sign = 'hdcore=3.7.0' |
ff6f9a67 | 3098 | f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') |
c4251b9a RA |
3099 | hds_host = hosts.get('hds') |
3100 | if hds_host: | |
3101 | f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url) | |
e71a4509 RA |
3102 | if 'hdcore=' not in f4m_url: |
3103 | f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign | |
3104 | f4m_formats = self._extract_f4m_formats( | |
3105 | f4m_url, video_id, f4m_id='hds', fatal=False) | |
3106 | for entry in f4m_formats: | |
3107 | entry.update({'extra_param_to_segment_url': hdcore_sign}) | |
3108 | formats.extend(f4m_formats) | |
70c5802b | 3109 | |
c4251b9a RA |
3110 | m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8') |
3111 | hls_host = hosts.get('hls') | |
3112 | if hls_host: | |
3113 | m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url) | |
f6a1d69a | 3114 | m3u8_formats, m3u8_subtitles = self._extract_m3u8_formats_and_subtitles( |
c7c43a93 | 3115 | m3u8_url, video_id, 'mp4', 'm3u8_native', |
29f7c58a | 3116 | m3u8_id='hls', fatal=False) |
3117 | formats.extend(m3u8_formats) | |
f6a1d69a | 3118 | subtitles = self._merge_subtitles(subtitles, m3u8_subtitles) |
70c5802b | 3119 | |
3120 | http_host = hosts.get('http') | |
29f7c58a | 3121 | if http_host and m3u8_formats and not signed: |
3122 | REPL_REGEX = r'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+' | |
70c5802b | 3123 | qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',') |
3124 | qualities_length = len(qualities) | |
29f7c58a | 3125 | if len(m3u8_formats) in (qualities_length, qualities_length + 1): |
70c5802b | 3126 | i = 0 |
29f7c58a | 3127 | for f in m3u8_formats: |
3128 | if f['vcodec'] != 'none': | |
70c5802b | 3129 | for protocol in ('http', 'https'): |
3130 | http_f = f.copy() | |
3131 | del http_f['manifest_url'] | |
3132 | http_url = re.sub( | |
86e5f3ed | 3133 | REPL_REGEX, protocol + fr'://{http_host}/\g<1>{qualities[i]}\3', f['url']) |
70c5802b | 3134 | http_f.update({ |
3135 | 'format_id': http_f['format_id'].replace('hls-', protocol + '-'), | |
3136 | 'url': http_url, | |
3137 | 'protocol': protocol, | |
3138 | }) | |
29f7c58a | 3139 | formats.append(http_f) |
70c5802b | 3140 | i += 1 |
70c5802b | 3141 | |
f6a1d69a | 3142 | return formats, subtitles |
c7c43a93 | 3143 | |
6ad02195 | 3144 | def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]): |
14f25df2 | 3145 | query = urllib.parse.urlparse(url).query |
6ad02195 | 3146 | url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url) |
240f2622 S |
3147 | mobj = re.search( |
3148 | r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url) | |
3149 | url_base = mobj.group('url') | |
3150 | http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base) | |
6ad02195 | 3151 | formats = [] |
044eeb14 S |
3152 | |
3153 | def manifest_url(manifest): | |
86e5f3ed | 3154 | m_url = f'{http_base_url}/{manifest}' |
044eeb14 S |
3155 | if query: |
3156 | m_url += '?%s' % query | |
3157 | return m_url | |
3158 | ||
6ad02195 RA |
3159 | if 'm3u8' not in skip_protocols: |
3160 | formats.extend(self._extract_m3u8_formats( | |
044eeb14 | 3161 | manifest_url('playlist.m3u8'), video_id, 'mp4', |
6ad02195 RA |
3162 | m3u8_entry_protocol, m3u8_id='hls', fatal=False)) |
3163 | if 'f4m' not in skip_protocols: | |
3164 | formats.extend(self._extract_f4m_formats( | |
044eeb14 | 3165 | manifest_url('manifest.f4m'), |
6ad02195 | 3166 | video_id, f4m_id='hds', fatal=False)) |
0384932e RA |
3167 | if 'dash' not in skip_protocols: |
3168 | formats.extend(self._extract_mpd_formats( | |
044eeb14 | 3169 | manifest_url('manifest.mpd'), |
0384932e | 3170 | video_id, mpd_id='dash', fatal=False)) |
6ad02195 | 3171 | if re.search(r'(?:/smil:|\.smil)', url_base): |
6ad02195 RA |
3172 | if 'smil' not in skip_protocols: |
3173 | rtmp_formats = self._extract_smil_formats( | |
044eeb14 | 3174 | manifest_url('jwplayer.smil'), |
6ad02195 RA |
3175 | video_id, fatal=False) |
3176 | for rtmp_format in rtmp_formats: | |
3177 | rtsp_format = rtmp_format.copy() | |
3178 | rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path']) | |
3179 | del rtsp_format['play_path'] | |
3180 | del rtsp_format['ext'] | |
3181 | rtsp_format.update({ | |
3182 | 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'), | |
3183 | 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'), | |
3184 | 'protocol': 'rtsp', | |
3185 | }) | |
3186 | formats.extend([rtmp_format, rtsp_format]) | |
3187 | else: | |
3188 | for protocol in ('rtmp', 'rtsp'): | |
3189 | if protocol not in skip_protocols: | |
3190 | formats.append({ | |
86e5f3ed | 3191 | 'url': f'{protocol}:{url_base}', |
6ad02195 RA |
3192 | 'format_id': protocol, |
3193 | 'protocol': protocol, | |
3194 | }) | |
3195 | return formats | |
3196 | ||
c73e330e | 3197 | def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json): |
a4a554a7 | 3198 | mobj = re.search( |
ac9c69ac | 3199 | r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)', |
a4a554a7 YCH |
3200 | webpage) |
3201 | if mobj: | |
c73e330e RU |
3202 | try: |
3203 | jwplayer_data = self._parse_json(mobj.group('options'), | |
3204 | video_id=video_id, | |
3205 | transform_source=transform_source) | |
3206 | except ExtractorError: | |
3207 | pass | |
3208 | else: | |
3209 | if isinstance(jwplayer_data, dict): | |
3210 | return jwplayer_data | |
a4a554a7 YCH |
3211 | |
3212 | def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs): | |
c73e330e RU |
3213 | jwplayer_data = self._find_jwplayer_data( |
3214 | webpage, video_id, transform_source=js_to_json) | |
a4a554a7 YCH |
3215 | return self._parse_jwplayer_data( |
3216 | jwplayer_data, video_id, *args, **kwargs) | |
3217 | ||
3218 | def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True, | |
3219 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
3220 | # JWPlayer backward compatibility: flattened playlists | |
3221 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96 | |
3222 | if 'playlist' not in jwplayer_data: | |
3223 | jwplayer_data = {'playlist': [jwplayer_data]} | |
3224 | ||
3225 | entries = [] | |
3226 | ||
3227 | # JWPlayer backward compatibility: single playlist item | |
3228 | # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10 | |
3229 | if not isinstance(jwplayer_data['playlist'], list): | |
3230 | jwplayer_data['playlist'] = [jwplayer_data['playlist']] | |
3231 | ||
3232 | for video_data in jwplayer_data['playlist']: | |
3233 | # JWPlayer backward compatibility: flattened sources | |
3234 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35 | |
3235 | if 'sources' not in video_data: | |
3236 | video_data['sources'] = [video_data] | |
3237 | ||
3238 | this_video_id = video_id or video_data['mediaid'] | |
3239 | ||
1a2192cb S |
3240 | formats = self._parse_jwplayer_formats( |
3241 | video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id, | |
3242 | mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url) | |
a4a554a7 YCH |
3243 | |
3244 | subtitles = {} | |
3245 | tracks = video_data.get('tracks') | |
3246 | if tracks and isinstance(tracks, list): | |
3247 | for track in tracks: | |
96a2daa1 S |
3248 | if not isinstance(track, dict): |
3249 | continue | |
f4b74272 | 3250 | track_kind = track.get('kind') |
14f25df2 | 3251 | if not track_kind or not isinstance(track_kind, str): |
f4b74272 S |
3252 | continue |
3253 | if track_kind.lower() not in ('captions', 'subtitles'): | |
a4a554a7 YCH |
3254 | continue |
3255 | track_url = urljoin(base_url, track.get('file')) | |
3256 | if not track_url: | |
3257 | continue | |
3258 | subtitles.setdefault(track.get('label') or 'en', []).append({ | |
3259 | 'url': self._proto_relative_url(track_url) | |
3260 | }) | |
3261 | ||
50d808f5 | 3262 | entry = { |
a4a554a7 | 3263 | 'id': this_video_id, |
50d808f5 | 3264 | 'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')), |
f81dd65b | 3265 | 'description': clean_html(video_data.get('description')), |
6945b9e7 | 3266 | 'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))), |
a4a554a7 YCH |
3267 | 'timestamp': int_or_none(video_data.get('pubdate')), |
3268 | 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')), | |
3269 | 'subtitles': subtitles, | |
50d808f5 RA |
3270 | } |
3271 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32 | |
3272 | if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']): | |
3273 | entry.update({ | |
3274 | '_type': 'url_transparent', | |
3275 | 'url': formats[0]['url'], | |
3276 | }) | |
3277 | else: | |
50d808f5 RA |
3278 | entry['formats'] = formats |
3279 | entries.append(entry) | |
a4a554a7 YCH |
3280 | if len(entries) == 1: |
3281 | return entries[0] | |
3282 | else: | |
3283 | return self.playlist_result(entries) | |
3284 | ||
ed0cf9b3 S |
3285 | def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None, |
3286 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
bf1b87cd | 3287 | urls = [] |
ed0cf9b3 | 3288 | formats = [] |
1a2192cb | 3289 | for source in jwplayer_sources_data: |
0a268c6e S |
3290 | if not isinstance(source, dict): |
3291 | continue | |
6945b9e7 RA |
3292 | source_url = urljoin( |
3293 | base_url, self._proto_relative_url(source.get('file'))) | |
3294 | if not source_url or source_url in urls: | |
bf1b87cd RA |
3295 | continue |
3296 | urls.append(source_url) | |
ed0cf9b3 S |
3297 | source_type = source.get('type') or '' |
3298 | ext = mimetype2ext(source_type) or determine_ext(source_url) | |
3299 | if source_type == 'hls' or ext == 'm3u8': | |
3300 | formats.extend(self._extract_m3u8_formats( | |
0236cd0d S |
3301 | source_url, video_id, 'mp4', entry_protocol='m3u8_native', |
3302 | m3u8_id=m3u8_id, fatal=False)) | |
0d9c48de | 3303 | elif source_type == 'dash' or ext == 'mpd': |
ed0cf9b3 S |
3304 | formats.extend(self._extract_mpd_formats( |
3305 | source_url, video_id, mpd_id=mpd_id, fatal=False)) | |
b51dc9db S |
3306 | elif ext == 'smil': |
3307 | formats.extend(self._extract_smil_formats( | |
3308 | source_url, video_id, fatal=False)) | |
ed0cf9b3 | 3309 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67 |
0236cd0d S |
3310 | elif source_type.startswith('audio') or ext in ( |
3311 | 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'): | |
ed0cf9b3 S |
3312 | formats.append({ |
3313 | 'url': source_url, | |
3314 | 'vcodec': 'none', | |
3315 | 'ext': ext, | |
3316 | }) | |
3317 | else: | |
3318 | height = int_or_none(source.get('height')) | |
3319 | if height is None: | |
3320 | # Often no height is provided but there is a label in | |
0236cd0d | 3321 | # format like "1080p", "720p SD", or 1080. |
ed0cf9b3 | 3322 | height = int_or_none(self._search_regex( |
14f25df2 | 3323 | r'^(\d{3,4})[pP]?(?:\b|$)', str(source.get('label') or ''), |
ed0cf9b3 S |
3324 | 'height', default=None)) |
3325 | a_format = { | |
3326 | 'url': source_url, | |
3327 | 'width': int_or_none(source.get('width')), | |
3328 | 'height': height, | |
d3a3d7f0 | 3329 | 'tbr': int_or_none(source.get('bitrate'), scale=1000), |
3330 | 'filesize': int_or_none(source.get('filesize')), | |
ed0cf9b3 S |
3331 | 'ext': ext, |
3332 | } | |
3333 | if source_url.startswith('rtmp'): | |
3334 | a_format['ext'] = 'flv' | |
ed0cf9b3 S |
3335 | # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as |
3336 | # of jwplayer.flash.swf | |
3337 | rtmp_url_parts = re.split( | |
3338 | r'((?:mp4|mp3|flv):)', source_url, 1) | |
3339 | if len(rtmp_url_parts) == 3: | |
3340 | rtmp_url, prefix, play_path = rtmp_url_parts | |
3341 | a_format.update({ | |
3342 | 'url': rtmp_url, | |
3343 | 'play_path': prefix + play_path, | |
3344 | }) | |
3345 | if rtmp_params: | |
3346 | a_format.update(rtmp_params) | |
3347 | formats.append(a_format) | |
3348 | return formats | |
3349 | ||
f4b1c7ad | 3350 | def _live_title(self, name): |
39ca3b5c | 3351 | self._downloader.deprecation_warning('yt_dlp.InfoExtractor._live_title is deprecated and does not work as expected') |
3352 | return name | |
f4b1c7ad | 3353 | |
b14f3a4c PH |
3354 | def _int(self, v, name, fatal=False, **kwargs): |
3355 | res = int_or_none(v, **kwargs) | |
b14f3a4c | 3356 | if res is None: |
86e5f3ed | 3357 | msg = f'Failed to extract {name}: Could not parse value {v!r}' |
b14f3a4c PH |
3358 | if fatal: |
3359 | raise ExtractorError(msg) | |
3360 | else: | |
6a39ee13 | 3361 | self.report_warning(msg) |
b14f3a4c PH |
3362 | return res |
3363 | ||
3364 | def _float(self, v, name, fatal=False, **kwargs): | |
3365 | res = float_or_none(v, **kwargs) | |
3366 | if res is None: | |
86e5f3ed | 3367 | msg = f'Failed to extract {name}: Could not parse value {v!r}' |
b14f3a4c PH |
3368 | if fatal: |
3369 | raise ExtractorError(msg) | |
3370 | else: | |
6a39ee13 | 3371 | self.report_warning(msg) |
b14f3a4c PH |
3372 | return res |
3373 | ||
40e41780 TF |
3374 | def _set_cookie(self, domain, name, value, expire_time=None, port=None, |
3375 | path='/', secure=False, discard=False, rest={}, **kwargs): | |
ac668111 | 3376 | cookie = http.cookiejar.Cookie( |
4ed2d7b7 | 3377 | 0, name, value, port, port is not None, domain, True, |
40e41780 TF |
3378 | domain.startswith('.'), path, True, secure, expire_time, |
3379 | discard, None, None, rest) | |
9809740b | 3380 | self.cookiejar.set_cookie(cookie) |
42939b61 | 3381 | |
799207e8 | 3382 | def _get_cookies(self, url): |
ac668111 | 3383 | """ Return a http.cookies.SimpleCookie with the cookies for the url """ |
8817a80d | 3384 | return LenientSimpleCookie(self._downloader._calc_cookies(url)) |
799207e8 | 3385 | |
e3c1266f | 3386 | def _apply_first_set_cookie_header(self, url_handle, cookie): |
ce2fe4c0 S |
3387 | """ |
3388 | Apply first Set-Cookie header instead of the last. Experimental. | |
3389 | ||
3390 | Some sites (e.g. [1-3]) may serve two cookies under the same name | |
3391 | in Set-Cookie header and expect the first (old) one to be set rather | |
3392 | than second (new). However, as of RFC6265 the newer one cookie | |
3393 | should be set into cookie store what actually happens. | |
3394 | We will workaround this issue by resetting the cookie to | |
3395 | the first one manually. | |
3396 | 1. https://new.vk.com/ | |
3397 | 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201 | |
3398 | 3. https://learning.oreilly.com/ | |
3399 | """ | |
e3c1266f S |
3400 | for header, cookies in url_handle.headers.items(): |
3401 | if header.lower() != 'set-cookie': | |
3402 | continue | |
cfb0511d | 3403 | cookies = cookies.encode('iso-8859-1').decode('utf-8') |
e3c1266f S |
3404 | cookie_value = re.search( |
3405 | r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies) | |
3406 | if cookie_value: | |
3407 | value, domain = cookie_value.groups() | |
3408 | self._set_cookie(domain, cookie, value) | |
3409 | break | |
3410 | ||
82d02080 | 3411 | @classmethod |
3412 | def get_testcases(cls, include_onlymatching=False): | |
6368e2e6 | 3413 | # Do not look in super classes |
3414 | t = vars(cls).get('_TEST') | |
05900629 | 3415 | if t: |
82d02080 | 3416 | assert not hasattr(cls, '_TESTS'), f'{cls.ie_key()}IE has _TEST and _TESTS' |
05900629 PH |
3417 | tests = [t] |
3418 | else: | |
6368e2e6 | 3419 | tests = vars(cls).get('_TESTS', []) |
05900629 PH |
3420 | for t in tests: |
3421 | if not include_onlymatching and t.get('only_matching', False): | |
3422 | continue | |
82d02080 | 3423 | t['name'] = cls.ie_key() |
05900629 PH |
3424 | yield t |
3425 | ||
f2e8dbcc | 3426 | @classmethod |
3427 | def get_webpage_testcases(cls): | |
6368e2e6 | 3428 | tests = vars(cls).get('_WEBPAGE_TESTS', []) |
f2e8dbcc | 3429 | for t in tests: |
3430 | t['name'] = cls.ie_key() | |
3431 | return tests | |
3432 | ||
6368e2e6 | 3433 | @classproperty(cache=True) |
24146491 | 3434 | def age_limit(cls): |
3435 | """Get age limit from the testcases""" | |
3436 | return max(traverse_obj( | |
f2e8dbcc | 3437 | (*cls.get_testcases(include_onlymatching=False), *cls.get_webpage_testcases()), |
24146491 | 3438 | (..., (('playlist', 0), None), 'info_dict', 'age_limit')) or [0]) |
3439 | ||
171a31db | 3440 | @classproperty(cache=True) |
3441 | def _RETURN_TYPE(cls): | |
3442 | """What the extractor returns: "video", "playlist", "any", or None (Unknown)""" | |
3443 | tests = tuple(cls.get_testcases(include_onlymatching=False)) | |
3444 | if not tests: | |
3445 | return None | |
3446 | elif not any(k.startswith('playlist') for test in tests for k in test): | |
3447 | return 'video' | |
3448 | elif all(any(k.startswith('playlist') for k in test) for test in tests): | |
3449 | return 'playlist' | |
3450 | return 'any' | |
3451 | ||
3452 | @classmethod | |
3453 | def is_single_video(cls, url): | |
3454 | """Returns whether the URL is of a single video, None if unknown""" | |
3455 | assert cls.suitable(url), 'The URL must be suitable for the extractor' | |
3456 | return {'video': True, 'playlist': False}.get(cls._RETURN_TYPE) | |
3457 | ||
82d02080 | 3458 | @classmethod |
3459 | def is_suitable(cls, age_limit): | |
24146491 | 3460 | """Test whether the extractor is generally suitable for the given age limit""" |
3461 | return not age_restricted(cls.age_limit, age_limit) | |
05900629 | 3462 | |
82d02080 | 3463 | @classmethod |
3464 | def description(cls, *, markdown=True, search_examples=None): | |
8dcce6a8 | 3465 | """Description of the extractor""" |
3466 | desc = '' | |
82d02080 | 3467 | if cls._NETRC_MACHINE: |
8dcce6a8 | 3468 | if markdown: |
82d02080 | 3469 | desc += f' [<abbr title="netrc machine"><em>{cls._NETRC_MACHINE}</em></abbr>]' |
8dcce6a8 | 3470 | else: |
82d02080 | 3471 | desc += f' [{cls._NETRC_MACHINE}]' |
3472 | if cls.IE_DESC is False: | |
8dcce6a8 | 3473 | desc += ' [HIDDEN]' |
82d02080 | 3474 | elif cls.IE_DESC: |
3475 | desc += f' {cls.IE_DESC}' | |
3476 | if cls.SEARCH_KEY: | |
3477 | desc += f'; "{cls.SEARCH_KEY}:" prefix' | |
8dcce6a8 | 3478 | if search_examples: |
3479 | _COUNTS = ('', '5', '10', 'all') | |
62b58c09 | 3480 | desc += f' (e.g. "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")' |
82d02080 | 3481 | if not cls.working(): |
8dcce6a8 | 3482 | desc += ' (**Currently broken**)' if markdown else ' (Currently broken)' |
3483 | ||
46d09f87 | 3484 | # Escape emojis. Ref: https://github.com/github/markup/issues/1153 |
3485 | name = (' - **%s**' % re.sub(r':(\w+:)', ':\u200B\\g<1>', cls.IE_NAME)) if markdown else cls.IE_NAME | |
8dcce6a8 | 3486 | return f'{name}:{desc}' if desc else name |
3487 | ||
a504ced0 | 3488 | def extract_subtitles(self, *args, **kwargs): |
a06916d9 | 3489 | if (self.get_param('writesubtitles', False) |
3490 | or self.get_param('listsubtitles')): | |
9868ea49 JMF |
3491 | return self._get_subtitles(*args, **kwargs) |
3492 | return {} | |
a504ced0 JMF |
3493 | |
3494 | def _get_subtitles(self, *args, **kwargs): | |
611c1dd9 | 3495 | raise NotImplementedError('This method must be implemented by subclasses') |
a504ced0 | 3496 | |
0cf643b2 M |
3497 | class CommentsDisabled(Exception): |
3498 | """Raise in _get_comments if comments are disabled for the video""" | |
3499 | ||
a2160aa4 | 3500 | def extract_comments(self, *args, **kwargs): |
3501 | if not self.get_param('getcomments'): | |
3502 | return None | |
3503 | generator = self._get_comments(*args, **kwargs) | |
3504 | ||
3505 | def extractor(): | |
3506 | comments = [] | |
d2b2fca5 | 3507 | interrupted = True |
a2160aa4 | 3508 | try: |
3509 | while True: | |
3510 | comments.append(next(generator)) | |
a2160aa4 | 3511 | except StopIteration: |
3512 | interrupted = False | |
d2b2fca5 | 3513 | except KeyboardInterrupt: |
3514 | self.to_screen('Interrupted by user') | |
0cf643b2 M |
3515 | except self.CommentsDisabled: |
3516 | return {'comments': None, 'comment_count': None} | |
d2b2fca5 | 3517 | except Exception as e: |
3518 | if self.get_param('ignoreerrors') is not True: | |
3519 | raise | |
3520 | self._downloader.report_error(e) | |
a2160aa4 | 3521 | comment_count = len(comments) |
3522 | self.to_screen(f'Extracted {comment_count} comments') | |
3523 | return { | |
3524 | 'comments': comments, | |
3525 | 'comment_count': None if interrupted else comment_count | |
3526 | } | |
3527 | return extractor | |
3528 | ||
3529 | def _get_comments(self, *args, **kwargs): | |
3530 | raise NotImplementedError('This method must be implemented by subclasses') | |
3531 | ||
912e0b7e YCH |
3532 | @staticmethod |
3533 | def _merge_subtitle_items(subtitle_list1, subtitle_list2): | |
a825ffbf | 3534 | """ Merge subtitle items for one language. Items with duplicated URLs/data |
912e0b7e | 3535 | will be dropped. """ |
86e5f3ed | 3536 | list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1} |
912e0b7e | 3537 | ret = list(subtitle_list1) |
a44ca5a4 | 3538 | ret.extend(item for item in subtitle_list2 if (item.get('url'), item.get('data')) not in list1_data) |
912e0b7e YCH |
3539 | return ret |
3540 | ||
3541 | @classmethod | |
46890374 | 3542 | def _merge_subtitles(cls, *dicts, target=None): |
19bb3920 | 3543 | """ Merge subtitle dictionaries, language by language. """ |
19bb3920 F |
3544 | if target is None: |
3545 | target = {} | |
3546 | for d in dicts: | |
3547 | for lang, subs in d.items(): | |
3548 | target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs) | |
3549 | return target | |
912e0b7e | 3550 | |
360e1ca5 | 3551 | def extract_automatic_captions(self, *args, **kwargs): |
a06916d9 | 3552 | if (self.get_param('writeautomaticsub', False) |
3553 | or self.get_param('listsubtitles')): | |
9868ea49 JMF |
3554 | return self._get_automatic_captions(*args, **kwargs) |
3555 | return {} | |
360e1ca5 JMF |
3556 | |
3557 | def _get_automatic_captions(self, *args, **kwargs): | |
611c1dd9 | 3558 | raise NotImplementedError('This method must be implemented by subclasses') |
360e1ca5 | 3559 | |
2762dbb1 | 3560 | @functools.cached_property |
24146491 | 3561 | def _cookies_passed(self): |
3562 | """Whether cookies have been passed to YoutubeDL""" | |
3563 | return self.get_param('cookiefile') is not None or self.get_param('cookiesfrombrowser') is not None | |
3564 | ||
d77ab8e2 | 3565 | def mark_watched(self, *args, **kwargs): |
1813a6cc | 3566 | if not self.get_param('mark_watched', False): |
3567 | return | |
24146491 | 3568 | if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed: |
d77ab8e2 S |
3569 | self._mark_watched(*args, **kwargs) |
3570 | ||
3571 | def _mark_watched(self, *args, **kwargs): | |
3572 | raise NotImplementedError('This method must be implemented by subclasses') | |
3573 | ||
38cce791 YCH |
3574 | def geo_verification_headers(self): |
3575 | headers = {} | |
a06916d9 | 3576 | geo_verification_proxy = self.get_param('geo_verification_proxy') |
38cce791 YCH |
3577 | if geo_verification_proxy: |
3578 | headers['Ytdl-request-proxy'] = geo_verification_proxy | |
3579 | return headers | |
3580 | ||
8f97a15d | 3581 | @staticmethod |
3582 | def _generic_id(url): | |
14f25df2 | 3583 | return urllib.parse.unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0]) |
98763ee3 | 3584 | |
62b8dac4 | 3585 | def _generic_title(self, url='', webpage='', *, default=None): |
3586 | return (self._og_search_title(webpage, default=None) | |
3587 | or self._html_extract_title(webpage, default=None) | |
3588 | or urllib.parse.unquote(os.path.splitext(url_basename(url))[0]) | |
3589 | or default) | |
98763ee3 | 3590 | |
c224251a | 3591 | @staticmethod |
b0089e89 | 3592 | def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None): |
c224251a M |
3593 | all_known = all(map( |
3594 | lambda x: x is not None, | |
3595 | (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted))) | |
3596 | return ( | |
3597 | 'private' if is_private | |
3598 | else 'premium_only' if needs_premium | |
3599 | else 'subscriber_only' if needs_subscription | |
3600 | else 'needs_auth' if needs_auth | |
3601 | else 'unlisted' if is_unlisted | |
3602 | else 'public' if all_known | |
3603 | else None) | |
3604 | ||
d43de682 | 3605 | def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False): |
4bb6b02f | 3606 | ''' |
3607 | @returns A list of values for the extractor argument given by "key" | |
3608 | or "default" if no such key is present | |
3609 | @param default The default value to return when the key is not present (default: []) | |
3610 | @param casesense When false, the values are converted to lower case | |
3611 | ''' | |
5225df50 | 3612 | ie_key = ie_key if isinstance(ie_key, str) else (ie_key or self).ie_key() |
3613 | val = traverse_obj(self._downloader.params, ('extractor_args', ie_key.lower(), key)) | |
4bb6b02f | 3614 | if val is None: |
3615 | return [] if default is NO_DEFAULT else default | |
3616 | return list(val) if casesense else [x.lower() for x in val] | |
5d3a0e79 | 3617 | |
f40ee5e9 | 3618 | def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist', video_label='video'): |
3619 | if not playlist_id or not video_id: | |
3620 | return not video_id | |
3621 | ||
3622 | no_playlist = (smuggled_data or {}).get('force_noplaylist') | |
3623 | if no_playlist is not None: | |
3624 | return not no_playlist | |
3625 | ||
3626 | video_id = '' if video_id is True else f' {video_id}' | |
3627 | playlist_id = '' if playlist_id is True else f' {playlist_id}' | |
3628 | if self.get_param('noplaylist'): | |
3629 | self.to_screen(f'Downloading just the {video_label}{video_id} because of --no-playlist') | |
3630 | return False | |
3631 | self.to_screen(f'Downloading {playlist_label}{playlist_id} - add --no-playlist to download just the {video_label}{video_id}') | |
3632 | return True | |
3633 | ||
be5c1ae8 | 3634 | def _error_or_warning(self, err, _count=None, _retries=0, *, fatal=True): |
8ca48a1a | 3635 | RetryManager.report_retry( |
3636 | err, _count or int(fatal), _retries, | |
3637 | info=self.to_screen, warn=self.report_warning, error=None if fatal else self.report_warning, | |
3638 | sleep_func=self.get_param('retry_sleep_functions', {}).get('extractor')) | |
be5c1ae8 | 3639 | |
3640 | def RetryManager(self, **kwargs): | |
3641 | return RetryManager(self.get_param('extractor_retries', 3), self._error_or_warning, **kwargs) | |
3642 | ||
ade1fa70 | 3643 | def _extract_generic_embeds(self, url, *args, info_dict={}, note='Extracting generic embeds', **kwargs): |
3644 | display_id = traverse_obj(info_dict, 'display_id', 'id') | |
3645 | self.to_screen(f'{format_field(display_id, None, "%s: ")}{note}') | |
3646 | return self._downloader.get_info_extractor('Generic')._extract_embeds( | |
3647 | smuggle_url(url, {'block_ies': [self.ie_key()]}), *args, **kwargs) | |
3648 | ||
8f97a15d | 3649 | @classmethod |
3650 | def extract_from_webpage(cls, ydl, url, webpage): | |
3651 | ie = (cls if isinstance(cls._extract_from_webpage, types.MethodType) | |
3652 | else ydl.get_info_extractor(cls.ie_key())) | |
f2e8dbcc | 3653 | for info in ie._extract_from_webpage(url, webpage) or []: |
3654 | # url = None since we do not want to set (webpage/original)_url | |
3655 | ydl.add_default_extra_info(info, ie, None) | |
3656 | yield info | |
8f97a15d | 3657 | |
3658 | @classmethod | |
3659 | def _extract_from_webpage(cls, url, webpage): | |
3660 | for embed_url in orderedSet( | |
3661 | cls._extract_embed_urls(url, webpage) or [], lazy=True): | |
d2c8aadf | 3662 | yield cls.url_result(embed_url, None if cls._VALID_URL is False else cls) |
8f97a15d | 3663 | |
3664 | @classmethod | |
3665 | def _extract_embed_urls(cls, url, webpage): | |
3666 | """@returns all the embed urls on the webpage""" | |
3667 | if '_EMBED_URL_RE' not in cls.__dict__: | |
3668 | assert isinstance(cls._EMBED_REGEX, (list, tuple)) | |
3669 | for idx, regex in enumerate(cls._EMBED_REGEX): | |
3670 | assert regex.count('(?P<url>') == 1, \ | |
3671 | f'{cls.__name__}._EMBED_REGEX[{idx}] must have exactly 1 url group\n\t{regex}' | |
3672 | cls._EMBED_URL_RE = tuple(map(re.compile, cls._EMBED_REGEX)) | |
3673 | ||
3674 | for regex in cls._EMBED_URL_RE: | |
3675 | for mobj in regex.finditer(webpage): | |
3676 | embed_url = urllib.parse.urljoin(url, unescapeHTML(mobj.group('url'))) | |
3677 | if cls._VALID_URL is False or cls.suitable(embed_url): | |
3678 | yield embed_url | |
3679 | ||
3680 | class StopExtraction(Exception): | |
3681 | pass | |
3682 | ||
bfd973ec | 3683 | @classmethod |
3684 | def _extract_url(cls, webpage): # TODO: Remove | |
3685 | """Only for compatibility with some older extractors""" | |
3686 | return next(iter(cls._extract_embed_urls(None, webpage) or []), None) | |
3687 | ||
2314b4d8 | 3688 | @classmethod |
3689 | def __init_subclass__(cls, *, plugin_name=None, **kwargs): | |
3690 | if plugin_name: | |
3691 | mro = inspect.getmro(cls) | |
3692 | super_class = cls.__wrapped__ = mro[mro.index(cls) + 1] | |
3693 | cls.IE_NAME, cls.ie_key = f'{super_class.IE_NAME}+{plugin_name}', super_class.ie_key | |
3694 | while getattr(super_class, '__wrapped__', None): | |
3695 | super_class = super_class.__wrapped__ | |
3696 | setattr(sys.modules[super_class.__module__], super_class.__name__, cls) | |
3697 | ||
3698 | return super().__init_subclass__(**kwargs) | |
3699 | ||
8dbe9899 | 3700 | |
d6983cb4 PH |
3701 | class SearchInfoExtractor(InfoExtractor): |
3702 | """ | |
3703 | Base class for paged search queries extractors. | |
10952eb2 | 3704 | They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query} |
96565c7e | 3705 | Instances should define _SEARCH_KEY and optionally _MAX_RESULTS |
d6983cb4 PH |
3706 | """ |
3707 | ||
96565c7e | 3708 | _MAX_RESULTS = float('inf') |
171a31db | 3709 | _RETURN_TYPE = 'playlist' |
96565c7e | 3710 | |
8f97a15d | 3711 | @classproperty |
3712 | def _VALID_URL(cls): | |
d6983cb4 PH |
3713 | return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY |
3714 | ||
d6983cb4 | 3715 | def _real_extract(self, query): |
2c4aaadd | 3716 | prefix, query = self._match_valid_url(query).group('prefix', 'query') |
d6983cb4 PH |
3717 | if prefix == '': |
3718 | return self._get_n_results(query, 1) | |
3719 | elif prefix == 'all': | |
3720 | return self._get_n_results(query, self._MAX_RESULTS) | |
3721 | else: | |
3722 | n = int(prefix) | |
3723 | if n <= 0: | |
86e5f3ed | 3724 | raise ExtractorError(f'invalid download number {n} for query "{query}"') |
d6983cb4 | 3725 | elif n > self._MAX_RESULTS: |
6a39ee13 | 3726 | self.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n)) |
d6983cb4 PH |
3727 | n = self._MAX_RESULTS |
3728 | return self._get_n_results(query, n) | |
3729 | ||
3730 | def _get_n_results(self, query, n): | |
cc16383f | 3731 | """Get a specified number of results for a query. |
3732 | Either this function or _search_results must be overridden by subclasses """ | |
3733 | return self.playlist_result( | |
3734 | itertools.islice(self._search_results(query), 0, None if n == float('inf') else n), | |
3735 | query, query) | |
3736 | ||
3737 | def _search_results(self, query): | |
3738 | """Returns an iterator of search results""" | |
611c1dd9 | 3739 | raise NotImplementedError('This method must be implemented by subclasses') |
0f818663 | 3740 | |
82d02080 | 3741 | @classproperty |
3742 | def SEARCH_KEY(cls): | |
3743 | return cls._SEARCH_KEY | |
fe7866d0 | 3744 | |
3745 | ||
3746 | class UnsupportedURLIE(InfoExtractor): | |
3747 | _VALID_URL = '.*' | |
3748 | _ENABLED = False | |
3749 | IE_DESC = False | |
3750 | ||
3751 | def _real_extract(self, url): | |
3752 | raise UnsupportedError(url) |