]>
Commit | Line | Data |
---|---|---|
d6983cb4 | 1 | import base64 |
234416e4 | 2 | import collections |
3ec05685 | 3 | import hashlib |
cc16383f | 4 | import itertools |
3d3538e4 | 5 | import json |
f8271158 | 6 | import math |
4094b6e3 | 7 | import netrc |
d6983cb4 | 8 | import os |
773f291d | 9 | import random |
d6983cb4 | 10 | import re |
d6983cb4 | 11 | import sys |
4094b6e3 | 12 | import time |
f8271158 | 13 | import xml.etree.ElementTree |
d6983cb4 | 14 | |
8c25f81b | 15 | from ..compat import ( |
6c22cee6 | 16 | compat_cookiejar_Cookie, |
f7ad7160 | 17 | compat_cookies_SimpleCookie, |
e9c0cdd3 | 18 | compat_etree_fromstring, |
0001fcb5 | 19 | compat_expanduser, |
e64b7569 | 20 | compat_getpass, |
d6983cb4 | 21 | compat_http_client, |
e9c0cdd3 | 22 | compat_os_name, |
61d3665d | 23 | compat_Pattern, |
e9c0cdd3 | 24 | compat_str, |
d6983cb4 | 25 | compat_urllib_error, |
98763ee3 | 26 | compat_urllib_parse_unquote, |
15707c7e | 27 | compat_urllib_parse_urlencode, |
41d06b04 | 28 | compat_urllib_request, |
f0b5d6af | 29 | compat_urlparse, |
8c25f81b | 30 | ) |
eb8a4433 | 31 | from ..downloader import FileDownloader |
f8271158 | 32 | from ..downloader.f4m import get_base_url, remove_encrypted_media |
8c25f81b | 33 | from ..utils import ( |
f8271158 | 34 | JSON_LD_RE, |
35 | NO_DEFAULT, | |
36 | ExtractorError, | |
37 | GeoRestrictedError, | |
38 | GeoUtils, | |
39 | RegexNotFoundError, | |
40 | UnsupportedError, | |
05900629 | 41 | age_restricted, |
02dc0a36 | 42 | base_url, |
08f2a92c | 43 | bug_reports_message, |
d6983cb4 | 44 | clean_html, |
70f0f5a8 | 45 | determine_ext, |
46b18f23 | 46 | determine_protocol, |
d493f15c | 47 | dict_get, |
42676437 | 48 | encode_data_uri, |
9b9c5355 | 49 | error_to_compat_str, |
46b18f23 | 50 | extract_attributes, |
90137ca4 | 51 | filter_dict, |
97f4aecf | 52 | fix_xml_ampersands, |
b14f3a4c | 53 | float_or_none, |
b868936c | 54 | format_field, |
31bb8d3f | 55 | int_or_none, |
34921b43 | 56 | join_nonempty, |
a4a554a7 | 57 | js_to_json, |
46b18f23 | 58 | mimetype2ext, |
3158150c | 59 | network_exceptions, |
46b18f23 | 60 | orderedSet, |
d493f15c | 61 | parse_bitrate, |
46b18f23 JH |
62 | parse_codecs, |
63 | parse_duration, | |
4ca2a3cf | 64 | parse_iso8601, |
46b18f23 | 65 | parse_m3u8_attributes, |
d493f15c | 66 | parse_resolution, |
46b18f23 | 67 | sanitize_filename, |
b868936c | 68 | sanitized_Request, |
d493f15c | 69 | str_or_none, |
ce5b9040 | 70 | str_to_int, |
f856816b | 71 | strip_or_none, |
5d3a0e79 | 72 | traverse_obj, |
ffa89477 | 73 | try_get, |
f38de77f | 74 | unescapeHTML, |
647eab45 | 75 | unified_strdate, |
6b3a3098 | 76 | unified_timestamp, |
46b18f23 JH |
77 | update_Request, |
78 | update_url_query, | |
a107193e | 79 | url_basename, |
bebef109 | 80 | url_or_none, |
b868936c | 81 | urljoin, |
6606817a | 82 | variadic, |
a6571f10 | 83 | xpath_element, |
8d6765cf S |
84 | xpath_text, |
85 | xpath_with_ns, | |
d6983cb4 | 86 | ) |
c342041f | 87 | |
d6983cb4 | 88 | |
86e5f3ed | 89 | class InfoExtractor: |
d6983cb4 PH |
90 | """Information Extractor class. |
91 | ||
92 | Information extractors are the classes that, given a URL, extract | |
93 | information about the video (or videos) the URL refers to. This | |
94 | information includes the real video URL, the video title, author and | |
95 | others. The information is stored in a dictionary which is then | |
5d380852 | 96 | passed to the YoutubeDL. The YoutubeDL processes this |
d6983cb4 PH |
97 | information possibly downloading the video to the file system, among |
98 | other possible outcomes. | |
99 | ||
cf0649f8 | 100 | The type field determines the type of the result. |
fed5d032 PH |
101 | By far the most common value (and the default if _type is missing) is |
102 | "video", which indicates a single video. | |
103 | ||
104 | For a video, the dictionaries must include the following fields: | |
d6983cb4 PH |
105 | |
106 | id: Video identifier. | |
d6983cb4 | 107 | title: Video title, unescaped. |
d67b0b15 | 108 | |
f49d89ee | 109 | Additionally, it must contain either a formats entry or a url one: |
d67b0b15 | 110 | |
f49d89ee PH |
111 | formats: A list of dictionaries for each format available, ordered |
112 | from worst to best quality. | |
113 | ||
114 | Potential fields: | |
c790e93a S |
115 | * url The mandatory URL representing the media: |
116 | for plain file media - HTTP URL of this file, | |
117 | for RTMP - RTMP URL, | |
118 | for HLS - URL of the M3U8 media playlist, | |
119 | for HDS - URL of the F4M manifest, | |
79d2077e S |
120 | for DASH |
121 | - HTTP URL to plain file media (in case of | |
122 | unfragmented media) | |
123 | - URL of the MPD manifest or base URL | |
124 | representing the media if MPD manifest | |
8ed7a233 | 125 | is parsed from a string (in case of |
79d2077e | 126 | fragmented media) |
c790e93a | 127 | for MSS - URL of the ISM manifest. |
86f4d14f S |
128 | * manifest_url |
129 | The URL of the manifest file in case of | |
c790e93a S |
130 | fragmented media: |
131 | for HLS - URL of the M3U8 master playlist, | |
132 | for HDS - URL of the F4M manifest, | |
133 | for DASH - URL of the MPD manifest, | |
134 | for MSS - URL of the ISM manifest. | |
a44ca5a4 | 135 | * manifest_stream_number (For internal use only) |
136 | The index of the stream in the manifest file | |
10952eb2 | 137 | * ext Will be calculated from URL if missing |
d67b0b15 PH |
138 | * format A human-readable description of the format |
139 | ("mp4 container with h264/opus"). | |
140 | Calculated from the format_id, width, height. | |
141 | and format_note fields if missing. | |
142 | * format_id A short description of the format | |
5d4f3985 PH |
143 | ("mp4_h264_opus" or "19"). |
144 | Technically optional, but strongly recommended. | |
d67b0b15 PH |
145 | * format_note Additional info about the format |
146 | ("3D" or "DASH video") | |
147 | * width Width of the video, if known | |
148 | * height Height of the video, if known | |
f49d89ee | 149 | * resolution Textual description of width and height |
176f1866 | 150 | * dynamic_range The dynamic range of the video. One of: |
151 | "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV" | |
7217e148 | 152 | * tbr Average bitrate of audio and video in KBit/s |
d67b0b15 PH |
153 | * abr Average audio bitrate in KBit/s |
154 | * acodec Name of the audio codec in use | |
dd27fd17 | 155 | * asr Audio sampling rate in Hertz |
d67b0b15 | 156 | * vbr Average video bitrate in KBit/s |
fbb21cf5 | 157 | * fps Frame rate |
d67b0b15 | 158 | * vcodec Name of the video codec in use |
1394ce65 | 159 | * container Name of the container format |
d67b0b15 | 160 | * filesize The number of bytes, if known in advance |
9732d77e | 161 | * filesize_approx An estimate for the number of bytes |
d67b0b15 | 162 | * player_url SWF Player URL (used for rtmpdump). |
c7deaa4c | 163 | * protocol The protocol that will be used for the actual |
adbc4ec4 THD |
164 | download, lower-case. One of "http", "https" or |
165 | one of the protocols defined in downloader.PROTOCOL_MAP | |
c58c2d63 S |
166 | * fragment_base_url |
167 | Base URL for fragments. Each fragment's path | |
168 | value (if present) will be relative to | |
169 | this URL. | |
170 | * fragments A list of fragments of a fragmented media. | |
171 | Each fragment entry must contain either an url | |
172 | or a path. If an url is present it should be | |
173 | considered by a client. Otherwise both path and | |
174 | fragment_base_url must be present. Here is | |
175 | the list of all potential fields: | |
176 | * "url" - fragment's URL | |
177 | * "path" - fragment's path relative to | |
178 | fragment_base_url | |
a0d5077c S |
179 | * "duration" (optional, int or float) |
180 | * "filesize" (optional, int) | |
adbc4ec4 THD |
181 | * is_from_start Is a live format that can be downloaded |
182 | from the start. Boolean | |
f49d89ee | 183 | * preference Order number of this format. If this field is |
08d13955 | 184 | present and not None, the formats get sorted |
38d63d84 | 185 | by this field, regardless of all other values. |
f49d89ee PH |
186 | -1 for default (order by other properties), |
187 | -2 or smaller for less than default. | |
e65566a9 PH |
188 | < -1000 to hide the format (if there is |
189 | another one which is strictly better) | |
32f90364 PH |
190 | * language Language code, e.g. "de" or "en-US". |
191 | * language_preference Is this in the language mentioned in | |
192 | the URL? | |
aff2f4f4 PH |
193 | 10 if it's what the URL is about, |
194 | -1 for default (don't know), | |
195 | -10 otherwise, other values reserved for now. | |
5d73273f PH |
196 | * quality Order number of the video quality of this |
197 | format, irrespective of the file format. | |
198 | -1 for default (order by other properties), | |
199 | -2 or smaller for less than default. | |
c64ed2a3 PH |
200 | * source_preference Order number for this video source |
201 | (quality takes higher priority) | |
202 | -1 for default (order by other properties), | |
203 | -2 or smaller for less than default. | |
d769be6c PH |
204 | * http_headers A dictionary of additional HTTP headers |
205 | to add to the request. | |
6271f1ca | 206 | * stretched_ratio If given and not 1, indicates that the |
3dee7826 PH |
207 | video's pixels are not square. |
208 | width : height ratio as float. | |
209 | * no_resume The server does not support resuming the | |
210 | (HTTP or RTMP) download. Boolean. | |
88acdbc2 | 211 | * has_drm The format has DRM and cannot be downloaded. Boolean |
00c97e3e | 212 | * downloader_options A dictionary of downloader options as |
a44ca5a4 | 213 | described in FileDownloader (For internal use only) |
3b1fe47d | 214 | RTMP formats can also have the additional fields: page_url, |
215 | app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn, | |
216 | rtmp_protocol, rtmp_real_time | |
3dee7826 | 217 | |
c0ba0f48 | 218 | url: Final video URL. |
d6983cb4 | 219 | ext: Video filename extension. |
d67b0b15 PH |
220 | format: The video format, defaults to ext (used for --get-format) |
221 | player_url: SWF Player URL (used for rtmpdump). | |
2f5865cc | 222 | |
d6983cb4 PH |
223 | The following fields are optional: |
224 | ||
08d30158 | 225 | direct: True if a direct video file was given (must only be set by GenericIE) |
f5e43bc6 | 226 | alt_title: A secondary title of the video. |
0afef30b PH |
227 | display_id An alternative identifier for the video, not necessarily |
228 | unique, but available before title. Typically, id is | |
229 | something like "4234987", title "Dancing naked mole rats", | |
230 | and display_id "dancing-naked-mole-rats" | |
d5519808 | 231 | thumbnails: A list of dictionaries, with the following entries: |
cfb56d1a | 232 | * "id" (optional, string) - Thumbnail format ID |
d5519808 | 233 | * "url" |
cfb56d1a | 234 | * "preference" (optional, int) - quality of the image |
d5519808 PH |
235 | * "width" (optional, int) |
236 | * "height" (optional, int) | |
5e1c39ac | 237 | * "resolution" (optional, string "{width}x{height}", |
d5519808 | 238 | deprecated) |
2de624fd | 239 | * "filesize" (optional, int) |
297e9952 | 240 | * "http_headers" (dict) - HTTP headers for the request |
d6983cb4 | 241 | thumbnail: Full URL to a video thumbnail image. |
f5e43bc6 | 242 | description: Full video description. |
d6983cb4 | 243 | uploader: Full name of the video uploader. |
2bc0c46f | 244 | license: License name the video is licensed under. |
8a92e51c | 245 | creator: The creator of the video. |
10db0d2f | 246 | timestamp: UNIX timestamp of the moment the video was uploaded |
ae6a1b95 | 247 | upload_date: Video upload date in UTC (YYYYMMDD). |
f0d785d3 | 248 | If not explicitly set, calculated from timestamp |
249 | release_timestamp: UNIX timestamp of the moment the video was released. | |
250 | If it is not clear whether to use timestamp or this, use the former | |
ae6a1b95 | 251 | release_date: The date (YYYYMMDD) when the video was released in UTC. |
f0d785d3 | 252 | If not explicitly set, calculated from release_timestamp |
253 | modified_timestamp: UNIX timestamp of the moment the video was last modified. | |
ae6a1b95 | 254 | modified_date: The date (YYYYMMDD) when the video was last modified in UTC. |
f0d785d3 | 255 | If not explicitly set, calculated from modified_timestamp |
d6983cb4 | 256 | uploader_id: Nickname or id of the video uploader. |
7bcd2830 | 257 | uploader_url: Full URL to a personal webpage of the video uploader. |
6f1f59f3 | 258 | channel: Full name of the channel the video is uploaded on. |
0e7b8d3e | 259 | Note that channel fields may or may not repeat uploader |
6f1f59f3 S |
260 | fields. This depends on a particular extractor. |
261 | channel_id: Id of the channel. | |
262 | channel_url: Full URL to a channel webpage. | |
6c73052c | 263 | channel_follower_count: Number of followers of the channel. |
da9ec3b9 | 264 | location: Physical location where the video was filmed. |
a504ced0 | 265 | subtitles: The available subtitles as a dictionary in the format |
4606c34e YCH |
266 | {tag: subformats}. "tag" is usually a language code, and |
267 | "subformats" is a list sorted from lower to higher | |
268 | preference, each element is a dictionary with the "ext" | |
269 | entry and one of: | |
a504ced0 | 270 | * "data": The subtitles file contents |
10952eb2 | 271 | * "url": A URL pointing to the subtitles file |
2412044c | 272 | It can optionally also have: |
273 | * "name": Name or description of the subtitles | |
08d30158 | 274 | * "http_headers": A dictionary of additional HTTP headers |
297e9952 | 275 | to add to the request. |
4bba3716 | 276 | "ext" will be calculated from URL if missing |
e167860c | 277 | automatic_captions: Like 'subtitles'; contains automatically generated |
278 | captions instead of normal subtitles | |
62d231c0 | 279 | duration: Length of the video in seconds, as an integer or float. |
f3d29461 | 280 | view_count: How many users have watched the video on the platform. |
19e3dfc9 PH |
281 | like_count: Number of positive ratings of the video |
282 | dislike_count: Number of negative ratings of the video | |
02835c6b | 283 | repost_count: Number of reposts of the video |
2d30521a | 284 | average_rating: Average rating give by users, the scale used depends on the webpage |
19e3dfc9 | 285 | comment_count: Number of comments on the video |
dd622d7c PH |
286 | comments: A list of comments, each with one or more of the following |
287 | properties (all but one of text or html optional): | |
288 | * "author" - human-readable name of the comment author | |
289 | * "author_id" - user ID of the comment author | |
a1c5d2ca | 290 | * "author_thumbnail" - The thumbnail of the comment author |
dd622d7c PH |
291 | * "id" - Comment ID |
292 | * "html" - Comment as HTML | |
293 | * "text" - Plain text of the comment | |
294 | * "timestamp" - UNIX timestamp of comment | |
295 | * "parent" - ID of the comment this one is replying to. | |
296 | Set to "root" to indicate that this is a | |
297 | comment to the original video. | |
a1c5d2ca M |
298 | * "like_count" - Number of positive ratings of the comment |
299 | * "dislike_count" - Number of negative ratings of the comment | |
300 | * "is_favorited" - Whether the comment is marked as | |
301 | favorite by the video uploader | |
302 | * "author_is_uploader" - Whether the comment is made by | |
303 | the video uploader | |
8dbe9899 | 304 | age_limit: Age restriction for the video, as an integer (years) |
7a5c1cfe | 305 | webpage_url: The URL to the video webpage, if given to yt-dlp it |
9103bbc5 JMF |
306 | should allow to get the same result again. (It will be set |
307 | by YoutubeDL if it's missing) | |
ad3bc6ac PH |
308 | categories: A list of categories that the video falls in, for example |
309 | ["Sports", "Berlin"] | |
864f24bd | 310 | tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"] |
d0fb4bd1 | 311 | cast: A list of the video cast |
7267bd53 PH |
312 | is_live: True, False, or None (=unknown). Whether this video is a |
313 | live stream that goes on instead of a fixed-length video. | |
f76ede8e | 314 | was_live: True, False, or None (=unknown). Whether this video was |
315 | originally a live stream. | |
3dbb2a9d | 316 | live_status: 'is_live', 'is_upcoming', 'was_live', 'not_live' or None (=unknown) |
ae30b840 | 317 | If absent, automatically set from is_live, was_live |
7c80519c | 318 | start_time: Time in seconds where the reproduction should start, as |
10952eb2 | 319 | specified in the URL. |
297a564b | 320 | end_time: Time in seconds where the reproduction should end, as |
10952eb2 | 321 | specified in the URL. |
55949fed | 322 | chapters: A list of dictionaries, with the following entries: |
323 | * "start_time" - The start time of the chapter in seconds | |
324 | * "end_time" - The end time of the chapter in seconds | |
325 | * "title" (optional, string) | |
6cfda058 | 326 | playable_in_embed: Whether this video is allowed to play in embedded |
327 | players on other sites. Can be True (=always allowed), | |
328 | False (=never allowed), None (=unknown), or a string | |
c224251a M |
329 | specifying the criteria for embedability (Eg: 'whitelist') |
330 | availability: Under what condition the video is available. One of | |
331 | 'private', 'premium_only', 'subscriber_only', 'needs_auth', | |
332 | 'unlisted' or 'public'. Use 'InfoExtractor._availability' | |
333 | to set it | |
277d6ff5 | 334 | __post_extractor: A function to be called just before the metadata is |
335 | written to either disk, logger or console. The function | |
336 | must return a dict which will be added to the info_dict. | |
337 | This is usefull for additional information that is | |
338 | time-consuming to extract. Note that the fields thus | |
339 | extracted will not be available to output template and | |
340 | match_filter. So, only "comments" and "comment_count" are | |
341 | currently allowed to be extracted via this method. | |
d6983cb4 | 342 | |
7109903e S |
343 | The following fields should only be used when the video belongs to some logical |
344 | chapter or section: | |
345 | ||
346 | chapter: Name or title of the chapter the video belongs to. | |
27bfd4e5 S |
347 | chapter_number: Number of the chapter the video belongs to, as an integer. |
348 | chapter_id: Id of the chapter the video belongs to, as a unicode string. | |
7109903e S |
349 | |
350 | The following fields should only be used when the video is an episode of some | |
8d76bdf1 | 351 | series, programme or podcast: |
7109903e S |
352 | |
353 | series: Title of the series or programme the video episode belongs to. | |
9ac24e23 | 354 | series_id: Id of the series or programme the video episode belongs to, as a unicode string. |
7109903e | 355 | season: Title of the season the video episode belongs to. |
27bfd4e5 S |
356 | season_number: Number of the season the video episode belongs to, as an integer. |
357 | season_id: Id of the season the video episode belongs to, as a unicode string. | |
7109903e S |
358 | episode: Title of the video episode. Unlike mandatory video title field, |
359 | this field should denote the exact title of the video episode | |
360 | without any kind of decoration. | |
27bfd4e5 S |
361 | episode_number: Number of the video episode within a season, as an integer. |
362 | episode_id: Id of the video episode, as a unicode string. | |
7109903e | 363 | |
7a93ab5f S |
364 | The following fields should only be used when the media is a track or a part of |
365 | a music album: | |
366 | ||
367 | track: Title of the track. | |
368 | track_number: Number of the track within an album or a disc, as an integer. | |
369 | track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii), | |
370 | as a unicode string. | |
371 | artist: Artist(s) of the track. | |
372 | genre: Genre(s) of the track. | |
373 | album: Title of the album the track belongs to. | |
374 | album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc). | |
375 | album_artist: List of all artists appeared on the album (e.g. | |
376 | "Ash Borer / Fell Voices" or "Various Artists", useful for splits | |
377 | and compilations). | |
378 | disc_number: Number of the disc or other physical medium the track belongs to, | |
379 | as an integer. | |
380 | release_year: Year (YYYY) when the album was released. | |
8bcd4048 | 381 | composer: Composer of the piece |
7a93ab5f | 382 | |
deefc05b | 383 | Unless mentioned otherwise, the fields should be Unicode strings. |
d6983cb4 | 384 | |
d838b1bd PH |
385 | Unless mentioned otherwise, None is equivalent to absence of information. |
386 | ||
fed5d032 PH |
387 | |
388 | _type "playlist" indicates multiple videos. | |
b82f815f PH |
389 | There must be a key "entries", which is a list, an iterable, or a PagedList |
390 | object, each element of which is a valid dictionary by this specification. | |
fed5d032 | 391 | |
b60419c5 | 392 | Additionally, playlists can have "id", "title", and any other relevent |
393 | attributes with the same semantics as videos (see above). | |
fed5d032 | 394 | |
f0d785d3 | 395 | It can also have the following optional fields: |
396 | ||
397 | playlist_count: The total number of videos in a playlist. If not given, | |
398 | YoutubeDL tries to calculate it from "entries" | |
399 | ||
fed5d032 PH |
400 | |
401 | _type "multi_video" indicates that there are multiple videos that | |
402 | form a single show, for examples multiple acts of an opera or TV episode. | |
403 | It must have an entries key like a playlist and contain all the keys | |
404 | required for a video at the same time. | |
405 | ||
406 | ||
407 | _type "url" indicates that the video must be extracted from another | |
408 | location, possibly by a different extractor. Its only required key is: | |
409 | "url" - the next URL to extract. | |
f58766ce PH |
410 | The key "ie_key" can be set to the class name (minus the trailing "IE", |
411 | e.g. "Youtube") if the extractor class is known in advance. | |
412 | Additionally, the dictionary may have any properties of the resolved entity | |
413 | known in advance, for example "title" if the title of the referred video is | |
fed5d032 PH |
414 | known ahead of time. |
415 | ||
416 | ||
417 | _type "url_transparent" entities have the same specification as "url", but | |
418 | indicate that the given additional information is more precise than the one | |
419 | associated with the resolved URL. | |
420 | This is useful when a site employs a video service that hosts the video and | |
421 | its technical metadata, but that video service does not embed a useful | |
422 | title, description etc. | |
423 | ||
424 | ||
08d30158 | 425 | Subclasses of this should define a _VALID_URL regexp and, re-define the |
426 | _real_extract() and (optionally) _real_initialize() methods. | |
d6983cb4 PH |
427 | Probably, they should also be added to the list of extractors. |
428 | ||
e6f21b3d | 429 | Subclasses may also override suitable() if necessary, but ensure the function |
430 | signature is preserved and that this function imports everything it needs | |
52efa4b3 | 431 | (except other extractors), so that lazy_extractors works correctly. |
432 | ||
433 | To support username + password (or netrc) login, the extractor must define a | |
434 | _NETRC_MACHINE and re-define _perform_login(username, password) and | |
435 | (optionally) _initialize_pre_login() methods. The _perform_login method will | |
436 | be called between _initialize_pre_login and _real_initialize if credentials | |
437 | are passed by the user. In cases where it is necessary to have the login | |
438 | process as part of the extraction rather than initialization, _perform_login | |
439 | can be left undefined. | |
e6f21b3d | 440 | |
4248dad9 | 441 | _GEO_BYPASS attribute may be set to False in order to disable |
773f291d S |
442 | geo restriction bypass mechanisms for a particular extractor. |
443 | Though it won't disable explicit geo restriction bypass based on | |
504f20dd | 444 | country code provided with geo_bypass_country. |
4248dad9 S |
445 | |
446 | _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted | |
447 | countries for this extractor. One of these countries will be used by | |
448 | geo restriction bypass mechanism right away in order to bypass | |
504f20dd | 449 | geo restriction, of course, if the mechanism is not disabled. |
773f291d | 450 | |
5f95927a S |
451 | _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted |
452 | IP blocks in CIDR notation for this extractor. One of these IP blocks | |
453 | will be used by geo restriction bypass mechanism similarly | |
504f20dd | 454 | to _GEO_COUNTRIES. |
3ccdde8c | 455 | |
e6f21b3d | 456 | The _WORKING attribute should be set to False for broken IEs |
d6983cb4 PH |
457 | in order to warn the users and skip the tests. |
458 | """ | |
459 | ||
460 | _ready = False | |
461 | _downloader = None | |
773f291d | 462 | _x_forwarded_for_ip = None |
4248dad9 S |
463 | _GEO_BYPASS = True |
464 | _GEO_COUNTRIES = None | |
5f95927a | 465 | _GEO_IP_BLOCKS = None |
d6983cb4 | 466 | _WORKING = True |
52efa4b3 | 467 | _NETRC_MACHINE = None |
231025c4 | 468 | IE_DESC = None |
d6983cb4 | 469 | |
9d5d4d64 | 470 | _LOGIN_HINTS = { |
52efa4b3 | 471 | 'any': 'Use --cookies, --cookies-from-browser, --username and --password, or --netrc to provide account credentials', |
9d5d4d64 | 472 | 'cookies': ( |
a0c716bb | 473 | 'Use --cookies-from-browser or --cookies for the authentication. ' |
474 | 'See https://github.com/ytdl-org/youtube-dl#how-do-i-pass-cookies-to-youtube-dl for how to manually pass cookies'), | |
0930b11f | 475 | 'password': 'Use --username and --password, or --netrc to provide account credentials', |
9d5d4d64 | 476 | } |
477 | ||
d6983cb4 | 478 | def __init__(self, downloader=None): |
49a57e70 | 479 | """Constructor. Receives an optional downloader (a YoutubeDL instance). |
480 | If a downloader is not passed during initialization, | |
481 | it must be set using "set_downloader()" before "extract()" is called""" | |
d6983cb4 | 482 | self._ready = False |
773f291d | 483 | self._x_forwarded_for_ip = None |
28f436ba | 484 | self._printed_messages = set() |
d6983cb4 PH |
485 | self.set_downloader(downloader) |
486 | ||
487 | @classmethod | |
5ad28e7f | 488 | def _match_valid_url(cls, url): |
79cb2577 PH |
489 | # This does not use has/getattr intentionally - we want to know whether |
490 | # we have cached the regexp for *this* class, whereas getattr would also | |
491 | # match the superclass | |
492 | if '_VALID_URL_RE' not in cls.__dict__: | |
2c4aaadd | 493 | if '_VALID_URL' not in cls.__dict__: |
494 | cls._VALID_URL = cls._make_valid_url() | |
79cb2577 | 495 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) |
5ad28e7f | 496 | return cls._VALID_URL_RE.match(url) |
497 | ||
498 | @classmethod | |
499 | def suitable(cls, url): | |
500 | """Receives a URL and returns True if suitable for this IE.""" | |
3fb4e21b | 501 | # This function must import everything it needs (except other extractors), |
502 | # so that lazy_extractors works correctly | |
5ad28e7f | 503 | return cls._match_valid_url(url) is not None |
d6983cb4 | 504 | |
ed9266db PH |
505 | @classmethod |
506 | def _match_id(cls, url): | |
5ad28e7f | 507 | return cls._match_valid_url(url).group('id') |
ed9266db | 508 | |
1151c407 | 509 | @classmethod |
510 | def get_temp_id(cls, url): | |
511 | try: | |
512 | return cls._match_id(url) | |
513 | except (IndexError, AttributeError): | |
514 | return None | |
515 | ||
d6983cb4 PH |
516 | @classmethod |
517 | def working(cls): | |
518 | """Getter method for _WORKING.""" | |
519 | return cls._WORKING | |
520 | ||
52efa4b3 | 521 | @classmethod |
522 | def supports_login(cls): | |
523 | return bool(cls._NETRC_MACHINE) | |
524 | ||
d6983cb4 PH |
525 | def initialize(self): |
526 | """Initializes an instance (authentication, etc).""" | |
28f436ba | 527 | self._printed_messages = set() |
5f95927a S |
528 | self._initialize_geo_bypass({ |
529 | 'countries': self._GEO_COUNTRIES, | |
530 | 'ip_blocks': self._GEO_IP_BLOCKS, | |
531 | }) | |
4248dad9 | 532 | if not self._ready: |
52efa4b3 | 533 | self._initialize_pre_login() |
534 | if self.supports_login(): | |
535 | username, password = self._get_login_info() | |
536 | if username: | |
537 | self._perform_login(username, password) | |
538 | elif self.get_param('username') and False not in (self.IE_DESC, self._NETRC_MACHINE): | |
539 | self.report_warning(f'Login with password is not supported for this website. {self._LOGIN_HINTS["cookies"]}') | |
4248dad9 S |
540 | self._real_initialize() |
541 | self._ready = True | |
542 | ||
5f95927a | 543 | def _initialize_geo_bypass(self, geo_bypass_context): |
e39b5d4a S |
544 | """ |
545 | Initialize geo restriction bypass mechanism. | |
546 | ||
547 | This method is used to initialize geo bypass mechanism based on faking | |
548 | X-Forwarded-For HTTP header. A random country from provided country list | |
dc0a869e | 549 | is selected and a random IP belonging to this country is generated. This |
e39b5d4a S |
550 | IP will be passed as X-Forwarded-For HTTP header in all subsequent |
551 | HTTP requests. | |
e39b5d4a S |
552 | |
553 | This method will be used for initial geo bypass mechanism initialization | |
5f95927a S |
554 | during the instance initialization with _GEO_COUNTRIES and |
555 | _GEO_IP_BLOCKS. | |
e39b5d4a | 556 | |
5f95927a | 557 | You may also manually call it from extractor's code if geo bypass |
e39b5d4a | 558 | information is not available beforehand (e.g. obtained during |
5f95927a S |
559 | extraction) or due to some other reason. In this case you should pass |
560 | this information in geo bypass context passed as first argument. It may | |
561 | contain following fields: | |
562 | ||
563 | countries: List of geo unrestricted countries (similar | |
564 | to _GEO_COUNTRIES) | |
565 | ip_blocks: List of geo unrestricted IP blocks in CIDR notation | |
566 | (similar to _GEO_IP_BLOCKS) | |
567 | ||
e39b5d4a | 568 | """ |
773f291d | 569 | if not self._x_forwarded_for_ip: |
5f95927a S |
570 | |
571 | # Geo bypass mechanism is explicitly disabled by user | |
a06916d9 | 572 | if not self.get_param('geo_bypass', True): |
5f95927a S |
573 | return |
574 | ||
575 | if not geo_bypass_context: | |
576 | geo_bypass_context = {} | |
577 | ||
578 | # Backward compatibility: previously _initialize_geo_bypass | |
579 | # expected a list of countries, some 3rd party code may still use | |
580 | # it this way | |
581 | if isinstance(geo_bypass_context, (list, tuple)): | |
582 | geo_bypass_context = { | |
583 | 'countries': geo_bypass_context, | |
584 | } | |
585 | ||
586 | # The whole point of geo bypass mechanism is to fake IP | |
587 | # as X-Forwarded-For HTTP header based on some IP block or | |
588 | # country code. | |
589 | ||
590 | # Path 1: bypassing based on IP block in CIDR notation | |
591 | ||
592 | # Explicit IP block specified by user, use it right away | |
593 | # regardless of whether extractor is geo bypassable or not | |
a06916d9 | 594 | ip_block = self.get_param('geo_bypass_ip_block', None) |
5f95927a S |
595 | |
596 | # Otherwise use random IP block from geo bypass context but only | |
597 | # if extractor is known as geo bypassable | |
598 | if not ip_block: | |
599 | ip_blocks = geo_bypass_context.get('ip_blocks') | |
600 | if self._GEO_BYPASS and ip_blocks: | |
601 | ip_block = random.choice(ip_blocks) | |
602 | ||
603 | if ip_block: | |
604 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(ip_block) | |
0760b0a7 | 605 | self._downloader.write_debug( |
606 | '[debug] Using fake IP %s as X-Forwarded-For' % self._x_forwarded_for_ip) | |
5f95927a S |
607 | return |
608 | ||
609 | # Path 2: bypassing based on country code | |
610 | ||
611 | # Explicit country code specified by user, use it right away | |
612 | # regardless of whether extractor is geo bypassable or not | |
a06916d9 | 613 | country = self.get_param('geo_bypass_country', None) |
5f95927a S |
614 | |
615 | # Otherwise use random country code from geo bypass context but | |
616 | # only if extractor is known as geo bypassable | |
617 | if not country: | |
618 | countries = geo_bypass_context.get('countries') | |
619 | if self._GEO_BYPASS and countries: | |
620 | country = random.choice(countries) | |
621 | ||
622 | if country: | |
623 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country) | |
0760b0a7 | 624 | self._downloader.write_debug( |
86e5f3ed | 625 | f'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For') |
d6983cb4 PH |
626 | |
627 | def extract(self, url): | |
628 | """Extracts URL information and returns it in list of dicts.""" | |
3a5bcd03 | 629 | try: |
773f291d S |
630 | for _ in range(2): |
631 | try: | |
632 | self.initialize() | |
a06916d9 | 633 | self.write_debug('Extracting URL: %s' % url) |
0016b84e | 634 | ie_result = self._real_extract(url) |
07cce701 | 635 | if ie_result is None: |
636 | return None | |
0016b84e S |
637 | if self._x_forwarded_for_ip: |
638 | ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip | |
53ed7066 | 639 | subtitles = ie_result.get('subtitles') |
640 | if (subtitles and 'live_chat' in subtitles | |
a06916d9 | 641 | and 'no-live-chat' in self.get_param('compat_opts', [])): |
53ed7066 | 642 | del subtitles['live_chat'] |
0016b84e | 643 | return ie_result |
773f291d | 644 | except GeoRestrictedError as e: |
4248dad9 S |
645 | if self.__maybe_fake_ip_and_retry(e.countries): |
646 | continue | |
773f291d | 647 | raise |
0db3bae8 | 648 | except UnsupportedError: |
649 | raise | |
1151c407 | 650 | except ExtractorError as e: |
0db3bae8 | 651 | kwargs = { |
652 | 'video_id': e.video_id or self.get_temp_id(url), | |
653 | 'ie': self.IE_NAME, | |
b69fd25c | 654 | 'tb': e.traceback or sys.exc_info()[2], |
0db3bae8 | 655 | 'expected': e.expected, |
656 | 'cause': e.cause | |
657 | } | |
658 | if hasattr(e, 'countries'): | |
659 | kwargs['countries'] = e.countries | |
7265a219 | 660 | raise type(e)(e.orig_msg, **kwargs) |
3a5bcd03 | 661 | except compat_http_client.IncompleteRead as e: |
1151c407 | 662 | raise ExtractorError('A network error has occurred.', cause=e, expected=True, video_id=self.get_temp_id(url)) |
9650885b | 663 | except (KeyError, StopIteration) as e: |
1151c407 | 664 | raise ExtractorError('An extractor error has occurred.', cause=e, video_id=self.get_temp_id(url)) |
d6983cb4 | 665 | |
4248dad9 | 666 | def __maybe_fake_ip_and_retry(self, countries): |
a06916d9 | 667 | if (not self.get_param('geo_bypass_country', None) |
3089bc74 | 668 | and self._GEO_BYPASS |
a06916d9 | 669 | and self.get_param('geo_bypass', True) |
3089bc74 S |
670 | and not self._x_forwarded_for_ip |
671 | and countries): | |
eea0716c S |
672 | country_code = random.choice(countries) |
673 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) | |
4248dad9 S |
674 | if self._x_forwarded_for_ip: |
675 | self.report_warning( | |
eea0716c S |
676 | 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.' |
677 | % (self._x_forwarded_for_ip, country_code.upper())) | |
4248dad9 S |
678 | return True |
679 | return False | |
680 | ||
d6983cb4 | 681 | def set_downloader(self, downloader): |
08d30158 | 682 | """Sets a YoutubeDL instance as the downloader for this IE.""" |
d6983cb4 PH |
683 | self._downloader = downloader |
684 | ||
52efa4b3 | 685 | def _initialize_pre_login(self): |
686 | """ Intialization before login. Redefine in subclasses.""" | |
687 | pass | |
688 | ||
689 | def _perform_login(self, username, password): | |
690 | """ Login with username and password. Redefine in subclasses.""" | |
691 | pass | |
692 | ||
d6983cb4 PH |
693 | def _real_initialize(self): |
694 | """Real initialization process. Redefine in subclasses.""" | |
695 | pass | |
696 | ||
697 | def _real_extract(self, url): | |
698 | """Real extraction process. Redefine in subclasses.""" | |
08d30158 | 699 | raise NotImplementedError('This method must be implemented by subclasses') |
d6983cb4 | 700 | |
56c73665 JMF |
701 | @classmethod |
702 | def ie_key(cls): | |
703 | """A string for getting the InfoExtractor with get_info_extractor""" | |
3fb4e21b | 704 | return cls.__name__[:-2] |
56c73665 | 705 | |
d6983cb4 PH |
706 | @property |
707 | def IE_NAME(self): | |
dc519b54 | 708 | return compat_str(type(self).__name__[:-2]) |
d6983cb4 | 709 | |
d391b7e2 S |
710 | @staticmethod |
711 | def __can_accept_status_code(err, expected_status): | |
712 | assert isinstance(err, compat_urllib_error.HTTPError) | |
713 | if expected_status is None: | |
714 | return False | |
d391b7e2 S |
715 | elif callable(expected_status): |
716 | return expected_status(err.code) is True | |
717 | else: | |
6606817a | 718 | return err.code in variadic(expected_status) |
d391b7e2 S |
719 | |
720 | def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers={}, query={}, expected_status=None): | |
721 | """ | |
722 | Return the response handle. | |
723 | ||
724 | See _download_webpage docstring for arguments specification. | |
725 | """ | |
1cf376f5 | 726 | if not self._downloader._first_webpage_request: |
49a57e70 | 727 | sleep_interval = self.get_param('sleep_interval_requests') or 0 |
1cf376f5 | 728 | if sleep_interval > 0: |
5ef7d9bd | 729 | self.to_screen('Sleeping %s seconds ...' % sleep_interval) |
1cf376f5 | 730 | time.sleep(sleep_interval) |
731 | else: | |
732 | self._downloader._first_webpage_request = False | |
733 | ||
d6983cb4 PH |
734 | if note is None: |
735 | self.report_download_webpage(video_id) | |
736 | elif note is not False: | |
7cc3570e | 737 | if video_id is None: |
86e5f3ed | 738 | self.to_screen(str(note)) |
7cc3570e | 739 | else: |
86e5f3ed | 740 | self.to_screen(f'{video_id}: {note}') |
2132edaa S |
741 | |
742 | # Some sites check X-Forwarded-For HTTP header in order to figure out | |
743 | # the origin of the client behind proxy. This allows bypassing geo | |
744 | # restriction by faking this header's value to IP that belongs to some | |
745 | # geo unrestricted country. We will do so once we encounter any | |
746 | # geo restriction error. | |
747 | if self._x_forwarded_for_ip: | |
748 | if 'X-Forwarded-For' not in headers: | |
749 | headers['X-Forwarded-For'] = self._x_forwarded_for_ip | |
750 | ||
41d06b04 S |
751 | if isinstance(url_or_request, compat_urllib_request.Request): |
752 | url_or_request = update_Request( | |
753 | url_or_request, data=data, headers=headers, query=query) | |
754 | else: | |
cdfee168 | 755 | if query: |
756 | url_or_request = update_url_query(url_or_request, query) | |
2c0d9c62 | 757 | if data is not None or headers: |
41d06b04 | 758 | url_or_request = sanitized_Request(url_or_request, data, headers) |
d6983cb4 | 759 | try: |
dca08720 | 760 | return self._downloader.urlopen(url_or_request) |
3158150c | 761 | except network_exceptions as err: |
d391b7e2 S |
762 | if isinstance(err, compat_urllib_error.HTTPError): |
763 | if self.__can_accept_status_code(err, expected_status): | |
95e42d73 XDG |
764 | # Retain reference to error to prevent file object from |
765 | # being closed before it can be read. Works around the | |
766 | # effects of <https://bugs.python.org/issue15002> | |
767 | # introduced in Python 3.4.1. | |
768 | err.fp._error = err | |
d391b7e2 S |
769 | return err.fp |
770 | ||
aa94a6d3 PH |
771 | if errnote is False: |
772 | return False | |
d6983cb4 | 773 | if errnote is None: |
f1a9d64e | 774 | errnote = 'Unable to download webpage' |
7f8b2714 | 775 | |
86e5f3ed | 776 | errmsg = f'{errnote}: {error_to_compat_str(err)}' |
7cc3570e | 777 | if fatal: |
497d2fab | 778 | raise ExtractorError(errmsg, cause=err) |
7cc3570e | 779 | else: |
6a39ee13 | 780 | self.report_warning(errmsg) |
7cc3570e | 781 | return False |
d6983cb4 | 782 | |
d391b7e2 S |
783 | def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None): |
784 | """ | |
785 | Return a tuple (page content as string, URL handle). | |
786 | ||
787 | See _download_webpage docstring for arguments specification. | |
788 | """ | |
b9d3e163 PH |
789 | # Strip hashes from the URL (#1038) |
790 | if isinstance(url_or_request, (compat_str, str)): | |
791 | url_or_request = url_or_request.partition('#')[0] | |
792 | ||
d391b7e2 | 793 | urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query, expected_status=expected_status) |
7cc3570e PH |
794 | if urlh is False: |
795 | assert not fatal | |
796 | return False | |
c9a77969 | 797 | content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding) |
23be51d8 PH |
798 | return (content, urlh) |
799 | ||
c9a77969 YCH |
800 | @staticmethod |
801 | def _guess_encoding_from_content(content_type, webpage_bytes): | |
d6983cb4 PH |
802 | m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type) |
803 | if m: | |
804 | encoding = m.group(1) | |
805 | else: | |
0d75ae2c | 806 | m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]', |
f143d86a PH |
807 | webpage_bytes[:1024]) |
808 | if m: | |
809 | encoding = m.group(1).decode('ascii') | |
b60016e8 PH |
810 | elif webpage_bytes.startswith(b'\xff\xfe'): |
811 | encoding = 'utf-16' | |
f143d86a PH |
812 | else: |
813 | encoding = 'utf-8' | |
c9a77969 YCH |
814 | |
815 | return encoding | |
816 | ||
4457823d S |
817 | def __check_blocked(self, content): |
818 | first_block = content[:512] | |
3089bc74 S |
819 | if ('<title>Access to this site is blocked</title>' in content |
820 | and 'Websense' in first_block): | |
4457823d S |
821 | msg = 'Access to this webpage has been blocked by Websense filtering software in your network.' |
822 | blocked_iframe = self._html_search_regex( | |
823 | r'<iframe src="([^"]+)"', content, | |
824 | 'Websense information URL', default=None) | |
825 | if blocked_iframe: | |
826 | msg += ' Visit %s for more details' % blocked_iframe | |
827 | raise ExtractorError(msg, expected=True) | |
828 | if '<title>The URL you requested has been blocked</title>' in first_block: | |
829 | msg = ( | |
830 | 'Access to this webpage has been blocked by Indian censorship. ' | |
831 | 'Use a VPN or proxy server (with --proxy) to route around it.') | |
832 | block_msg = self._html_search_regex( | |
833 | r'</h1><p>(.*?)</p>', | |
834 | content, 'block message', default=None) | |
835 | if block_msg: | |
836 | msg += ' (Message: "%s")' % block_msg.replace('\n', ' ') | |
837 | raise ExtractorError(msg, expected=True) | |
3089bc74 S |
838 | if ('<title>TTK :: Доступ к ресурсу ограничен</title>' in content |
839 | and 'blocklist.rkn.gov.ru' in content): | |
4457823d S |
840 | raise ExtractorError( |
841 | 'Access to this webpage has been blocked by decision of the Russian government. ' | |
842 | 'Visit http://blocklist.rkn.gov.ru/ for a block reason.', | |
843 | expected=True) | |
844 | ||
c9a77969 YCH |
845 | def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None): |
846 | content_type = urlh.headers.get('Content-Type', '') | |
847 | webpage_bytes = urlh.read() | |
848 | if prefix is not None: | |
849 | webpage_bytes = prefix + webpage_bytes | |
850 | if not encoding: | |
851 | encoding = self._guess_encoding_from_content(content_type, webpage_bytes) | |
a06916d9 | 852 | if self.get_param('dump_intermediate_pages', False): |
f610dbb0 | 853 | self.to_screen('Dumping request to ' + urlh.geturl()) |
d6983cb4 PH |
854 | dump = base64.b64encode(webpage_bytes).decode('ascii') |
855 | self._downloader.to_screen(dump) | |
a06916d9 | 856 | if self.get_param('write_pages', False): |
86e5f3ed | 857 | basen = f'{video_id}_{urlh.geturl()}' |
bd6f722d | 858 | trim_length = self.get_param('trim_file_name') or 240 |
859 | if len(basen) > trim_length: | |
f1a9d64e | 860 | h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest() |
bd6f722d | 861 | basen = basen[:trim_length - len(h)] + h |
c1bce22f | 862 | raw_filename = basen + '.dump' |
d41e6efc | 863 | filename = sanitize_filename(raw_filename, restricted=True) |
f1a9d64e | 864 | self.to_screen('Saving request to ' + filename) |
5f58165d S |
865 | # Working around MAX_PATH limitation on Windows (see |
866 | # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx) | |
e9c0cdd3 | 867 | if compat_os_name == 'nt': |
5f58165d S |
868 | absfilepath = os.path.abspath(filename) |
869 | if len(absfilepath) > 259: | |
870 | filename = '\\\\?\\' + absfilepath | |
d41e6efc PH |
871 | with open(filename, 'wb') as outf: |
872 | outf.write(webpage_bytes) | |
873 | ||
ec0fafbb AA |
874 | try: |
875 | content = webpage_bytes.decode(encoding, 'replace') | |
876 | except LookupError: | |
877 | content = webpage_bytes.decode('utf-8', 'replace') | |
2410c43d | 878 | |
4457823d | 879 | self.__check_blocked(content) |
2410c43d | 880 | |
23be51d8 | 881 | return content |
d6983cb4 | 882 | |
d391b7e2 S |
883 | def _download_webpage( |
884 | self, url_or_request, video_id, note=None, errnote=None, | |
885 | fatal=True, tries=1, timeout=5, encoding=None, data=None, | |
886 | headers={}, query={}, expected_status=None): | |
887 | """ | |
888 | Return the data of the page as a string. | |
889 | ||
890 | Arguments: | |
891 | url_or_request -- plain text URL as a string or | |
892 | a compat_urllib_request.Requestobject | |
893 | video_id -- Video/playlist/item identifier (string) | |
894 | ||
895 | Keyword arguments: | |
896 | note -- note printed before downloading (string) | |
897 | errnote -- note printed in case of an error (string) | |
898 | fatal -- flag denoting whether error should be considered fatal, | |
899 | i.e. whether it should cause ExtractionError to be raised, | |
900 | otherwise a warning will be reported and extraction continued | |
901 | tries -- number of tries | |
902 | timeout -- sleep interval between tries | |
903 | encoding -- encoding for a page content decoding, guessed automatically | |
904 | when not explicitly specified | |
905 | data -- POST data (bytes) | |
906 | headers -- HTTP headers (dict) | |
907 | query -- URL query (dict) | |
908 | expected_status -- allows to accept failed HTTP requests (non 2xx | |
909 | status code) by explicitly specifying a set of accepted status | |
910 | codes. Can be any of the following entities: | |
911 | - an integer type specifying an exact failed status code to | |
912 | accept | |
913 | - a list or a tuple of integer types specifying a list of | |
914 | failed status codes to accept | |
915 | - a callable accepting an actual failed status code and | |
916 | returning True if it should be accepted | |
917 | Note that this argument does not affect success status codes (2xx) | |
918 | which are always accepted. | |
919 | """ | |
920 | ||
995ad69c TF |
921 | success = False |
922 | try_count = 0 | |
923 | while success is False: | |
924 | try: | |
d391b7e2 S |
925 | res = self._download_webpage_handle( |
926 | url_or_request, video_id, note, errnote, fatal, | |
927 | encoding=encoding, data=data, headers=headers, query=query, | |
928 | expected_status=expected_status) | |
995ad69c TF |
929 | success = True |
930 | except compat_http_client.IncompleteRead as e: | |
931 | try_count += 1 | |
932 | if try_count >= tries: | |
933 | raise e | |
934 | self._sleep(timeout, video_id) | |
7cc3570e PH |
935 | if res is False: |
936 | return res | |
937 | else: | |
938 | content, _ = res | |
939 | return content | |
d6983cb4 | 940 | |
e0d198c1 S |
941 | def _download_xml_handle( |
942 | self, url_or_request, video_id, note='Downloading XML', | |
943 | errnote='Unable to download XML', transform_source=None, | |
d391b7e2 S |
944 | fatal=True, encoding=None, data=None, headers={}, query={}, |
945 | expected_status=None): | |
946 | """ | |
f9934b96 | 947 | Return a tuple (xml as an xml.etree.ElementTree.Element, URL handle). |
d391b7e2 S |
948 | |
949 | See _download_webpage docstring for arguments specification. | |
950 | """ | |
e0d198c1 S |
951 | res = self._download_webpage_handle( |
952 | url_or_request, video_id, note, errnote, fatal=fatal, | |
d391b7e2 S |
953 | encoding=encoding, data=data, headers=headers, query=query, |
954 | expected_status=expected_status) | |
e0d198c1 S |
955 | if res is False: |
956 | return res | |
957 | xml_string, urlh = res | |
958 | return self._parse_xml( | |
959 | xml_string, video_id, transform_source=transform_source, | |
960 | fatal=fatal), urlh | |
961 | ||
d391b7e2 S |
962 | def _download_xml( |
963 | self, url_or_request, video_id, | |
964 | note='Downloading XML', errnote='Unable to download XML', | |
965 | transform_source=None, fatal=True, encoding=None, | |
966 | data=None, headers={}, query={}, expected_status=None): | |
967 | """ | |
f9934b96 | 968 | Return the xml as an xml.etree.ElementTree.Element. |
d391b7e2 S |
969 | |
970 | See _download_webpage docstring for arguments specification. | |
971 | """ | |
e0d198c1 S |
972 | res = self._download_xml_handle( |
973 | url_or_request, video_id, note=note, errnote=errnote, | |
974 | transform_source=transform_source, fatal=fatal, encoding=encoding, | |
d391b7e2 S |
975 | data=data, headers=headers, query=query, |
976 | expected_status=expected_status) | |
e0d198c1 | 977 | return res if res is False else res[0] |
e01c3d2e S |
978 | |
979 | def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True): | |
e2b38da9 PH |
980 | if transform_source: |
981 | xml_string = transform_source(xml_string) | |
e01c3d2e S |
982 | try: |
983 | return compat_etree_fromstring(xml_string.encode('utf-8')) | |
f9934b96 | 984 | except xml.etree.ElementTree.ParseError as ve: |
e01c3d2e S |
985 | errmsg = '%s: Failed to parse XML ' % video_id |
986 | if fatal: | |
987 | raise ExtractorError(errmsg, cause=ve) | |
988 | else: | |
989 | self.report_warning(errmsg + str(ve)) | |
267ed0c5 | 990 | |
0fe7783e S |
991 | def _download_json_handle( |
992 | self, url_or_request, video_id, note='Downloading JSON metadata', | |
993 | errnote='Unable to download JSON metadata', transform_source=None, | |
d391b7e2 S |
994 | fatal=True, encoding=None, data=None, headers={}, query={}, |
995 | expected_status=None): | |
996 | """ | |
997 | Return a tuple (JSON object, URL handle). | |
998 | ||
999 | See _download_webpage docstring for arguments specification. | |
1000 | """ | |
0fe7783e | 1001 | res = self._download_webpage_handle( |
c9a77969 | 1002 | url_or_request, video_id, note, errnote, fatal=fatal, |
d391b7e2 S |
1003 | encoding=encoding, data=data, headers=headers, query=query, |
1004 | expected_status=expected_status) | |
0fe7783e S |
1005 | if res is False: |
1006 | return res | |
1007 | json_string, urlh = res | |
ebb64199 | 1008 | return self._parse_json( |
0fe7783e S |
1009 | json_string, video_id, transform_source=transform_source, |
1010 | fatal=fatal), urlh | |
1011 | ||
1012 | def _download_json( | |
1013 | self, url_or_request, video_id, note='Downloading JSON metadata', | |
1014 | errnote='Unable to download JSON metadata', transform_source=None, | |
d391b7e2 S |
1015 | fatal=True, encoding=None, data=None, headers={}, query={}, |
1016 | expected_status=None): | |
1017 | """ | |
1018 | Return the JSON object as a dict. | |
1019 | ||
1020 | See _download_webpage docstring for arguments specification. | |
1021 | """ | |
0fe7783e S |
1022 | res = self._download_json_handle( |
1023 | url_or_request, video_id, note=note, errnote=errnote, | |
1024 | transform_source=transform_source, fatal=fatal, encoding=encoding, | |
d391b7e2 S |
1025 | data=data, headers=headers, query=query, |
1026 | expected_status=expected_status) | |
0fe7783e | 1027 | return res if res is False else res[0] |
ebb64199 TF |
1028 | |
1029 | def _parse_json(self, json_string, video_id, transform_source=None, fatal=True): | |
81c2f20b PH |
1030 | if transform_source: |
1031 | json_string = transform_source(json_string) | |
3d3538e4 | 1032 | try: |
700ccbe3 | 1033 | return json.loads(json_string, strict=False) |
3d3538e4 | 1034 | except ValueError as ve: |
e7b6d122 PH |
1035 | errmsg = '%s: Failed to parse JSON ' % video_id |
1036 | if fatal: | |
1037 | raise ExtractorError(errmsg, cause=ve) | |
1038 | else: | |
1039 | self.report_warning(errmsg + str(ve)) | |
3d3538e4 | 1040 | |
adddc50c | 1041 | def _parse_socket_response_as_json(self, data, video_id, transform_source=None, fatal=True): |
1042 | return self._parse_json( | |
1043 | data[data.find('{'):data.rfind('}') + 1], | |
1044 | video_id, transform_source, fatal) | |
1045 | ||
1046 | def _download_socket_json_handle( | |
1047 | self, url_or_request, video_id, note='Polling socket', | |
1048 | errnote='Unable to poll socket', transform_source=None, | |
1049 | fatal=True, encoding=None, data=None, headers={}, query={}, | |
1050 | expected_status=None): | |
1051 | """ | |
1052 | Return a tuple (JSON object, URL handle). | |
1053 | ||
1054 | See _download_webpage docstring for arguments specification. | |
1055 | """ | |
1056 | res = self._download_webpage_handle( | |
1057 | url_or_request, video_id, note, errnote, fatal=fatal, | |
1058 | encoding=encoding, data=data, headers=headers, query=query, | |
1059 | expected_status=expected_status) | |
1060 | if res is False: | |
1061 | return res | |
1062 | webpage, urlh = res | |
1063 | return self._parse_socket_response_as_json( | |
1064 | webpage, video_id, transform_source=transform_source, | |
1065 | fatal=fatal), urlh | |
1066 | ||
1067 | def _download_socket_json( | |
1068 | self, url_or_request, video_id, note='Polling socket', | |
1069 | errnote='Unable to poll socket', transform_source=None, | |
1070 | fatal=True, encoding=None, data=None, headers={}, query={}, | |
1071 | expected_status=None): | |
1072 | """ | |
1073 | Return the JSON object as a dict. | |
1074 | ||
1075 | See _download_webpage docstring for arguments specification. | |
1076 | """ | |
1077 | res = self._download_socket_json_handle( | |
1078 | url_or_request, video_id, note=note, errnote=errnote, | |
1079 | transform_source=transform_source, fatal=fatal, encoding=encoding, | |
1080 | data=data, headers=headers, query=query, | |
1081 | expected_status=expected_status) | |
1082 | return res if res is False else res[0] | |
1083 | ||
28f436ba | 1084 | def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs): |
b868936c | 1085 | idstr = format_field(video_id, template='%s: ') |
28f436ba | 1086 | msg = f'[{self.IE_NAME}] {idstr}{msg}' |
1087 | if only_once: | |
1088 | if f'WARNING: {msg}' in self._printed_messages: | |
1089 | return | |
1090 | self._printed_messages.add(f'WARNING: {msg}') | |
1091 | self._downloader.report_warning(msg, *args, **kwargs) | |
f45f96f8 | 1092 | |
a06916d9 | 1093 | def to_screen(self, msg, *args, **kwargs): |
d6983cb4 | 1094 | """Print msg to screen, prefixing it with '[ie_name]'""" |
86e5f3ed | 1095 | self._downloader.to_screen(f'[{self.IE_NAME}] {msg}', *args, **kwargs) |
a06916d9 | 1096 | |
1097 | def write_debug(self, msg, *args, **kwargs): | |
86e5f3ed | 1098 | self._downloader.write_debug(f'[{self.IE_NAME}] {msg}', *args, **kwargs) |
a06916d9 | 1099 | |
1100 | def get_param(self, name, default=None, *args, **kwargs): | |
1101 | if self._downloader: | |
1102 | return self._downloader.params.get(name, default, *args, **kwargs) | |
1103 | return default | |
d6983cb4 | 1104 | |
88acdbc2 | 1105 | def report_drm(self, video_id, partial=False): |
1106 | self.raise_no_formats('This video is DRM protected', expected=True, video_id=video_id) | |
1107 | ||
d6983cb4 PH |
1108 | def report_extraction(self, id_or_name): |
1109 | """Report information extraction.""" | |
f1a9d64e | 1110 | self.to_screen('%s: Extracting information' % id_or_name) |
d6983cb4 PH |
1111 | |
1112 | def report_download_webpage(self, video_id): | |
1113 | """Report webpage download.""" | |
f1a9d64e | 1114 | self.to_screen('%s: Downloading webpage' % video_id) |
d6983cb4 PH |
1115 | |
1116 | def report_age_confirmation(self): | |
1117 | """Report attempt to confirm age.""" | |
f1a9d64e | 1118 | self.to_screen('Confirming age') |
d6983cb4 | 1119 | |
fc79158d JMF |
1120 | def report_login(self): |
1121 | """Report attempt to log in.""" | |
f1a9d64e | 1122 | self.to_screen('Logging in') |
fc79158d | 1123 | |
b7da73eb | 1124 | def raise_login_required( |
9d5d4d64 | 1125 | self, msg='This video is only available for registered users', |
52efa4b3 | 1126 | metadata_available=False, method=NO_DEFAULT): |
f2ebc5c7 | 1127 | if metadata_available and ( |
1128 | self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')): | |
b7da73eb | 1129 | self.report_warning(msg) |
7265a219 | 1130 | return |
52efa4b3 | 1131 | if method is NO_DEFAULT: |
1132 | method = 'any' if self.supports_login() else 'cookies' | |
46890374 | 1133 | if method is not None: |
52efa4b3 | 1134 | assert method in self._LOGIN_HINTS, 'Invalid login method' |
86e5f3ed | 1135 | msg = f'{msg}. {self._LOGIN_HINTS[method]}' |
46890374 | 1136 | raise ExtractorError(msg, expected=True) |
43e7d3c9 | 1137 | |
b7da73eb | 1138 | def raise_geo_restricted( |
1139 | self, msg='This video is not available from your location due to geo restriction', | |
1140 | countries=None, metadata_available=False): | |
f2ebc5c7 | 1141 | if metadata_available and ( |
1142 | self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')): | |
b7da73eb | 1143 | self.report_warning(msg) |
1144 | else: | |
1145 | raise GeoRestrictedError(msg, countries=countries) | |
1146 | ||
1147 | def raise_no_formats(self, msg, expected=False, video_id=None): | |
f2ebc5c7 | 1148 | if expected and ( |
1149 | self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')): | |
b7da73eb | 1150 | self.report_warning(msg, video_id) |
68f5867c L |
1151 | elif isinstance(msg, ExtractorError): |
1152 | raise msg | |
b7da73eb | 1153 | else: |
1154 | raise ExtractorError(msg, expected=expected, video_id=video_id) | |
c430802e | 1155 | |
5f6a1245 | 1156 | # Methods for following #608 |
c0d0b01f | 1157 | @staticmethod |
311b6615 | 1158 | def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs): |
10952eb2 | 1159 | """Returns a URL that points to a page that should be processed""" |
311b6615 | 1160 | if ie is not None: |
1161 | kwargs['ie_key'] = ie if isinstance(ie, str) else ie.ie_key() | |
7012b23c | 1162 | if video_id is not None: |
311b6615 | 1163 | kwargs['id'] = video_id |
830d53bf | 1164 | if video_title is not None: |
311b6615 | 1165 | kwargs['title'] = video_title |
1166 | return { | |
1167 | **kwargs, | |
1168 | '_type': 'url_transparent' if url_transparent else 'url', | |
1169 | 'url': url, | |
1170 | } | |
1171 | ||
27231526 ZM |
1172 | def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None, video_kwargs=None, **kwargs): |
1173 | urls = (self.url_result(self._proto_relative_url(m), ie, **(video_kwargs or {})) | |
311b6615 | 1174 | for m in orderedSet(map(getter, matches) if getter else matches)) |
1175 | return self.playlist_result(urls, playlist_id, playlist_title, **kwargs) | |
46b18f23 | 1176 | |
c0d0b01f | 1177 | @staticmethod |
311b6615 | 1178 | def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs): |
d6983cb4 | 1179 | """Returns a playlist""" |
d6983cb4 | 1180 | if playlist_id: |
311b6615 | 1181 | kwargs['id'] = playlist_id |
d6983cb4 | 1182 | if playlist_title: |
311b6615 | 1183 | kwargs['title'] = playlist_title |
ecc97af3 | 1184 | if playlist_description is not None: |
311b6615 | 1185 | kwargs['description'] = playlist_description |
1186 | return { | |
1187 | **kwargs, | |
1188 | '_type': 'multi_video' if multi_video else 'playlist', | |
1189 | 'entries': entries, | |
1190 | } | |
d6983cb4 | 1191 | |
c342041f | 1192 | def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
1193 | """ |
1194 | Perform a regex search on the given string, using a single or a list of | |
1195 | patterns returning the first matching group. | |
1196 | In case of failure return a default value or raise a WARNING or a | |
55b3e45b | 1197 | RegexNotFoundError, depending on fatal, specifying the field name. |
d6983cb4 | 1198 | """ |
61d3665d | 1199 | if string is None: |
1200 | mobj = None | |
1201 | elif isinstance(pattern, (str, compat_Pattern)): | |
d6983cb4 PH |
1202 | mobj = re.search(pattern, string, flags) |
1203 | else: | |
1204 | for p in pattern: | |
1205 | mobj = re.search(p, string, flags) | |
c3415d1b PH |
1206 | if mobj: |
1207 | break | |
d6983cb4 | 1208 | |
ec11a9f4 | 1209 | _name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS) |
d6983cb4 PH |
1210 | |
1211 | if mobj: | |
711ede6e PH |
1212 | if group is None: |
1213 | # return the first matching group | |
1214 | return next(g for g in mobj.groups() if g is not None) | |
198f7ea8 | 1215 | elif isinstance(group, (list, tuple)): |
1216 | return tuple(mobj.group(g) for g in group) | |
711ede6e PH |
1217 | else: |
1218 | return mobj.group(group) | |
c342041f | 1219 | elif default is not NO_DEFAULT: |
d6983cb4 PH |
1220 | return default |
1221 | elif fatal: | |
f1a9d64e | 1222 | raise RegexNotFoundError('Unable to extract %s' % _name) |
d6983cb4 | 1223 | else: |
6a39ee13 | 1224 | self.report_warning('unable to extract %s' % _name + bug_reports_message()) |
d6983cb4 PH |
1225 | return None |
1226 | ||
c342041f | 1227 | def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
1228 | """ |
1229 | Like _search_regex, but strips HTML tags and unescapes entities. | |
1230 | """ | |
711ede6e | 1231 | res = self._search_regex(pattern, string, name, default, fatal, flags, group) |
d6983cb4 PH |
1232 | if res: |
1233 | return clean_html(res).strip() | |
1234 | else: | |
1235 | return res | |
1236 | ||
2118fdd1 RA |
1237 | def _get_netrc_login_info(self, netrc_machine=None): |
1238 | username = None | |
1239 | password = None | |
1240 | netrc_machine = netrc_machine or self._NETRC_MACHINE | |
1241 | ||
a06916d9 | 1242 | if self.get_param('usenetrc', False): |
2118fdd1 | 1243 | try: |
0001fcb5 | 1244 | netrc_file = compat_expanduser(self.get_param('netrc_location') or '~') |
1245 | if os.path.isdir(netrc_file): | |
1246 | netrc_file = os.path.join(netrc_file, '.netrc') | |
1247 | info = netrc.netrc(file=netrc_file).authenticators(netrc_machine) | |
2118fdd1 RA |
1248 | if info is not None: |
1249 | username = info[0] | |
1250 | password = info[2] | |
1251 | else: | |
dcce092e S |
1252 | raise netrc.NetrcParseError( |
1253 | 'No authenticators for %s' % netrc_machine) | |
86e5f3ed | 1254 | except (OSError, netrc.NetrcParseError) as err: |
6a39ee13 | 1255 | self.report_warning( |
dcce092e | 1256 | 'parsing .netrc: %s' % error_to_compat_str(err)) |
2118fdd1 | 1257 | |
dcce092e | 1258 | return username, password |
2118fdd1 | 1259 | |
1b6712ab | 1260 | def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None): |
fc79158d | 1261 | """ |
cf0649f8 | 1262 | Get the login info as (username, password) |
32443dd3 S |
1263 | First look for the manually specified credentials using username_option |
1264 | and password_option as keys in params dictionary. If no such credentials | |
1265 | available look in the netrc file using the netrc_machine or _NETRC_MACHINE | |
1266 | value. | |
fc79158d JMF |
1267 | If there's no info available, return (None, None) |
1268 | """ | |
fc79158d JMF |
1269 | |
1270 | # Attempt to use provided username and password or .netrc data | |
a06916d9 | 1271 | username = self.get_param(username_option) |
1272 | if username is not None: | |
1273 | password = self.get_param(password_option) | |
2118fdd1 | 1274 | else: |
1b6712ab | 1275 | username, password = self._get_netrc_login_info(netrc_machine) |
5f6a1245 | 1276 | |
2133565c | 1277 | return username, password |
fc79158d | 1278 | |
e64b7569 | 1279 | def _get_tfa_info(self, note='two-factor verification code'): |
83317f69 | 1280 | """ |
1281 | Get the two-factor authentication info | |
1282 | TODO - asking the user will be required for sms/phone verify | |
1283 | currently just uses the command line option | |
1284 | If there's no info available, return None | |
1285 | """ | |
83317f69 | 1286 | |
a06916d9 | 1287 | tfa = self.get_param('twofactor') |
1288 | if tfa is not None: | |
1289 | return tfa | |
83317f69 | 1290 | |
e64b7569 | 1291 | return compat_getpass('Type %s and press [Return]: ' % note) |
83317f69 | 1292 | |
46720279 JMF |
1293 | # Helper functions for extracting OpenGraph info |
1294 | @staticmethod | |
ab2d5247 | 1295 | def _og_regexes(prop): |
448ef1f3 | 1296 | content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))' |
fbfde1c3 F |
1297 | property_re = (r'(?:name|property)=(?:\'og%(sep)s%(prop)s\'|"og%(sep)s%(prop)s"|\s*og%(sep)s%(prop)s\b)' |
1298 | % {'prop': re.escape(prop), 'sep': '(?::|[:-])'}) | |
78fb87b2 | 1299 | template = r'<meta[^>]+?%s[^>]+?%s' |
ab2d5247 | 1300 | return [ |
78fb87b2 JMF |
1301 | template % (property_re, content_re), |
1302 | template % (content_re, property_re), | |
ab2d5247 | 1303 | ] |
46720279 | 1304 | |
864f24bd S |
1305 | @staticmethod |
1306 | def _meta_regex(prop): | |
1307 | return r'''(?isx)<meta | |
8b9848ac | 1308 | (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1) |
864f24bd S |
1309 | [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop) |
1310 | ||
3c4e6d83 | 1311 | def _og_search_property(self, prop, html, name=None, **kargs): |
6606817a | 1312 | prop = variadic(prop) |
46720279 | 1313 | if name is None: |
b070564e S |
1314 | name = 'OpenGraph %s' % prop[0] |
1315 | og_regexes = [] | |
1316 | for p in prop: | |
1317 | og_regexes.extend(self._og_regexes(p)) | |
1318 | escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs) | |
eb0a8398 PH |
1319 | if escaped is None: |
1320 | return None | |
1321 | return unescapeHTML(escaped) | |
46720279 JMF |
1322 | |
1323 | def _og_search_thumbnail(self, html, **kargs): | |
10952eb2 | 1324 | return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs) |
46720279 JMF |
1325 | |
1326 | def _og_search_description(self, html, **kargs): | |
1327 | return self._og_search_property('description', html, fatal=False, **kargs) | |
1328 | ||
04f3fd2c | 1329 | def _og_search_title(self, html, *, fatal=False, **kargs): |
1330 | return self._og_search_property('title', html, fatal=fatal, **kargs) | |
46720279 | 1331 | |
8ffa13e0 | 1332 | def _og_search_video_url(self, html, name='video url', secure=True, **kargs): |
a3681973 PH |
1333 | regexes = self._og_regexes('video') + self._og_regexes('video:url') |
1334 | if secure: | |
1335 | regexes = self._og_regexes('video:secure_url') + regexes | |
8ffa13e0 | 1336 | return self._html_search_regex(regexes, html, name, **kargs) |
46720279 | 1337 | |
78338f71 JMF |
1338 | def _og_search_url(self, html, **kargs): |
1339 | return self._og_search_property('url', html, **kargs) | |
1340 | ||
04f3fd2c | 1341 | def _html_extract_title(self, html, name='title', *, fatal=False, **kwargs): |
1342 | return self._html_search_regex(r'(?s)<title>([^<]+)</title>', html, name, fatal=fatal, **kwargs) | |
77cc7c6e | 1343 | |
40c696e5 | 1344 | def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs): |
6606817a | 1345 | name = variadic(name) |
59040888 | 1346 | if display_name is None: |
88d9f6c0 | 1347 | display_name = name[0] |
59040888 | 1348 | return self._html_search_regex( |
88d9f6c0 | 1349 | [self._meta_regex(n) for n in name], |
711ede6e | 1350 | html, display_name, fatal=fatal, group='content', **kwargs) |
59040888 PH |
1351 | |
1352 | def _dc_search_uploader(self, html): | |
1353 | return self._html_search_meta('dc.creator', html, 'uploader') | |
1354 | ||
8dbe9899 PH |
1355 | def _rta_search(self, html): |
1356 | # See http://www.rtalabel.org/index.php?content=howtofaq#single | |
1357 | if re.search(r'(?ix)<meta\s+name="rating"\s+' | |
1358 | r' content="RTA-5042-1996-1400-1577-RTA"', | |
1359 | html): | |
1360 | return 18 | |
1361 | return 0 | |
1362 | ||
59040888 PH |
1363 | def _media_rating_search(self, html): |
1364 | # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/ | |
1365 | rating = self._html_search_meta('rating', html) | |
1366 | ||
1367 | if not rating: | |
1368 | return None | |
1369 | ||
1370 | RATING_TABLE = { | |
1371 | 'safe for kids': 0, | |
1372 | 'general': 8, | |
1373 | '14 years': 14, | |
1374 | 'mature': 17, | |
1375 | 'restricted': 19, | |
1376 | } | |
d800609c | 1377 | return RATING_TABLE.get(rating.lower()) |
59040888 | 1378 | |
69319969 | 1379 | def _family_friendly_search(self, html): |
6ca7732d | 1380 | # See http://schema.org/VideoObject |
ac8491fc S |
1381 | family_friendly = self._html_search_meta( |
1382 | 'isFamilyFriendly', html, default=None) | |
69319969 NJ |
1383 | |
1384 | if not family_friendly: | |
1385 | return None | |
1386 | ||
1387 | RATING_TABLE = { | |
1388 | '1': 0, | |
1389 | 'true': 0, | |
1390 | '0': 18, | |
1391 | 'false': 18, | |
1392 | } | |
d800609c | 1393 | return RATING_TABLE.get(family_friendly.lower()) |
69319969 | 1394 | |
0c708f11 JMF |
1395 | def _twitter_search_player(self, html): |
1396 | return self._html_search_meta('twitter:player', html, | |
9e1a5b84 | 1397 | 'twitter card player') |
0c708f11 | 1398 | |
95b31e26 | 1399 | def _search_json_ld(self, html, video_id, expected_type=None, **kwargs): |
4433bb02 | 1400 | json_ld_list = list(re.finditer(JSON_LD_RE, html)) |
321b5e08 | 1401 | default = kwargs.get('default', NO_DEFAULT) |
321b5e08 S |
1402 | # JSON-LD may be malformed and thus `fatal` should be respected. |
1403 | # At the same time `default` may be passed that assumes `fatal=False` | |
1404 | # for _search_regex. Let's simulate the same behavior here as well. | |
dbf5416a | 1405 | fatal = kwargs.get('fatal', True) if default is NO_DEFAULT else False |
4433bb02 S |
1406 | json_ld = [] |
1407 | for mobj in json_ld_list: | |
1408 | json_ld_item = self._parse_json( | |
1409 | mobj.group('json_ld'), video_id, fatal=fatal) | |
1410 | if not json_ld_item: | |
1411 | continue | |
1412 | if isinstance(json_ld_item, dict): | |
1413 | json_ld.append(json_ld_item) | |
1414 | elif isinstance(json_ld_item, (list, tuple)): | |
1415 | json_ld.extend(json_ld_item) | |
1416 | if json_ld: | |
1417 | json_ld = self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type) | |
1418 | if json_ld: | |
1419 | return json_ld | |
1420 | if default is not NO_DEFAULT: | |
1421 | return default | |
1422 | elif fatal: | |
1423 | raise RegexNotFoundError('Unable to extract JSON-LD') | |
1424 | else: | |
6a39ee13 | 1425 | self.report_warning('unable to extract JSON-LD %s' % bug_reports_message()) |
4433bb02 | 1426 | return {} |
4ca2a3cf | 1427 | |
95b31e26 | 1428 | def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): |
4ca2a3cf S |
1429 | if isinstance(json_ld, compat_str): |
1430 | json_ld = self._parse_json(json_ld, video_id, fatal=fatal) | |
1431 | if not json_ld: | |
1432 | return {} | |
1433 | info = {} | |
46933a15 S |
1434 | if not isinstance(json_ld, (list, tuple, dict)): |
1435 | return info | |
1436 | if isinstance(json_ld, dict): | |
1437 | json_ld = [json_ld] | |
bae14048 | 1438 | |
e7e4a6e0 S |
1439 | INTERACTION_TYPE_MAP = { |
1440 | 'CommentAction': 'comment', | |
1441 | 'AgreeAction': 'like', | |
1442 | 'DisagreeAction': 'dislike', | |
1443 | 'LikeAction': 'like', | |
1444 | 'DislikeAction': 'dislike', | |
1445 | 'ListenAction': 'view', | |
1446 | 'WatchAction': 'view', | |
1447 | 'ViewAction': 'view', | |
1448 | } | |
1449 | ||
29f7c58a | 1450 | def extract_interaction_type(e): |
1451 | interaction_type = e.get('interactionType') | |
1452 | if isinstance(interaction_type, dict): | |
1453 | interaction_type = interaction_type.get('@type') | |
1454 | return str_or_none(interaction_type) | |
1455 | ||
e7e4a6e0 S |
1456 | def extract_interaction_statistic(e): |
1457 | interaction_statistic = e.get('interactionStatistic') | |
29f7c58a | 1458 | if isinstance(interaction_statistic, dict): |
1459 | interaction_statistic = [interaction_statistic] | |
e7e4a6e0 S |
1460 | if not isinstance(interaction_statistic, list): |
1461 | return | |
1462 | for is_e in interaction_statistic: | |
1463 | if not isinstance(is_e, dict): | |
1464 | continue | |
1465 | if is_e.get('@type') != 'InteractionCounter': | |
1466 | continue | |
29f7c58a | 1467 | interaction_type = extract_interaction_type(is_e) |
1468 | if not interaction_type: | |
e7e4a6e0 | 1469 | continue |
ce5b9040 S |
1470 | # For interaction count some sites provide string instead of |
1471 | # an integer (as per spec) with non digit characters (e.g. ",") | |
1472 | # so extracting count with more relaxed str_to_int | |
1473 | interaction_count = str_to_int(is_e.get('userInteractionCount')) | |
e7e4a6e0 S |
1474 | if interaction_count is None: |
1475 | continue | |
1476 | count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1]) | |
1477 | if not count_kind: | |
1478 | continue | |
1479 | count_key = '%s_count' % count_kind | |
1480 | if info.get(count_key) is not None: | |
1481 | continue | |
1482 | info[count_key] = interaction_count | |
1483 | ||
f5225737 | 1484 | def extract_chapter_information(e): |
1485 | chapters = [{ | |
1486 | 'title': part.get('name'), | |
1487 | 'start_time': part.get('startOffset'), | |
1488 | 'end_time': part.get('endOffset'), | |
85553414 | 1489 | } for part in variadic(e.get('hasPart') or []) if part.get('@type') == 'Clip'] |
f5225737 | 1490 | for idx, (last_c, current_c, next_c) in enumerate(zip( |
1491 | [{'end_time': 0}] + chapters, chapters, chapters[1:])): | |
1492 | current_c['end_time'] = current_c['end_time'] or next_c['start_time'] | |
1493 | current_c['start_time'] = current_c['start_time'] or last_c['end_time'] | |
1494 | if None in current_c.values(): | |
1495 | self.report_warning(f'Chapter {idx} contains broken data. Not extracting chapters') | |
1496 | return | |
1497 | if chapters: | |
1498 | chapters[-1]['end_time'] = chapters[-1]['end_time'] or info['duration'] | |
1499 | info['chapters'] = chapters | |
1500 | ||
bae14048 S |
1501 | def extract_video_object(e): |
1502 | assert e['@type'] == 'VideoObject' | |
f7ad7160 | 1503 | author = e.get('author') |
bae14048 | 1504 | info.update({ |
bebef109 | 1505 | 'url': url_or_none(e.get('contentUrl')), |
bae14048 S |
1506 | 'title': unescapeHTML(e.get('name')), |
1507 | 'description': unescapeHTML(e.get('description')), | |
7592749c | 1508 | 'thumbnails': [{'url': url_or_none(url)} |
1509 | for url in variadic(traverse_obj(e, 'thumbnailUrl', 'thumbnailURL'))], | |
bae14048 S |
1510 | 'duration': parse_duration(e.get('duration')), |
1511 | 'timestamp': unified_timestamp(e.get('uploadDate')), | |
f7ad7160 | 1512 | # author can be an instance of 'Organization' or 'Person' types. |
1513 | # both types can have 'name' property(inherited from 'Thing' type). [1] | |
1514 | # however some websites are using 'Text' type instead. | |
1515 | # 1. https://schema.org/VideoObject | |
1516 | 'uploader': author.get('name') if isinstance(author, dict) else author if isinstance(author, compat_str) else None, | |
bae14048 S |
1517 | 'filesize': float_or_none(e.get('contentSize')), |
1518 | 'tbr': int_or_none(e.get('bitrate')), | |
1519 | 'width': int_or_none(e.get('width')), | |
1520 | 'height': int_or_none(e.get('height')), | |
33a81c2c | 1521 | 'view_count': int_or_none(e.get('interactionCount')), |
bae14048 | 1522 | }) |
e7e4a6e0 | 1523 | extract_interaction_statistic(e) |
f5225737 | 1524 | extract_chapter_information(e) |
bae14048 | 1525 | |
d5c32548 ZM |
1526 | def traverse_json_ld(json_ld, at_top_level=True): |
1527 | for e in json_ld: | |
1528 | if at_top_level and '@context' not in e: | |
1529 | continue | |
1530 | if at_top_level and set(e.keys()) == {'@context', '@graph'}: | |
1531 | traverse_json_ld(variadic(e['@graph'], allowed_types=(dict,)), at_top_level=False) | |
1532 | break | |
46933a15 S |
1533 | item_type = e.get('@type') |
1534 | if expected_type is not None and expected_type != item_type: | |
4433bb02 | 1535 | continue |
8f122fa0 | 1536 | rating = traverse_obj(e, ('aggregateRating', 'ratingValue'), expected_type=float_or_none) |
1537 | if rating is not None: | |
1538 | info['average_rating'] = rating | |
c69701c6 | 1539 | if item_type in ('TVEpisode', 'Episode'): |
440863ad | 1540 | episode_name = unescapeHTML(e.get('name')) |
46933a15 | 1541 | info.update({ |
440863ad | 1542 | 'episode': episode_name, |
46933a15 S |
1543 | 'episode_number': int_or_none(e.get('episodeNumber')), |
1544 | 'description': unescapeHTML(e.get('description')), | |
1545 | }) | |
440863ad S |
1546 | if not info.get('title') and episode_name: |
1547 | info['title'] = episode_name | |
46933a15 | 1548 | part_of_season = e.get('partOfSeason') |
c69701c6 | 1549 | if isinstance(part_of_season, dict) and part_of_season.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'): |
458fd30f S |
1550 | info.update({ |
1551 | 'season': unescapeHTML(part_of_season.get('name')), | |
1552 | 'season_number': int_or_none(part_of_season.get('seasonNumber')), | |
1553 | }) | |
d16b3c66 | 1554 | part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries') |
c69701c6 | 1555 | if isinstance(part_of_series, dict) and part_of_series.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'): |
46933a15 | 1556 | info['series'] = unescapeHTML(part_of_series.get('name')) |
391256dc S |
1557 | elif item_type == 'Movie': |
1558 | info.update({ | |
1559 | 'title': unescapeHTML(e.get('name')), | |
1560 | 'description': unescapeHTML(e.get('description')), | |
1561 | 'duration': parse_duration(e.get('duration')), | |
1562 | 'timestamp': unified_timestamp(e.get('dateCreated')), | |
1563 | }) | |
3931b845 | 1564 | elif item_type in ('Article', 'NewsArticle'): |
46933a15 S |
1565 | info.update({ |
1566 | 'timestamp': parse_iso8601(e.get('datePublished')), | |
1567 | 'title': unescapeHTML(e.get('headline')), | |
d5c32548 | 1568 | 'description': unescapeHTML(e.get('articleBody') or e.get('description')), |
46933a15 | 1569 | }) |
2edb38e8 | 1570 | if traverse_obj(e, ('video', 0, '@type')) == 'VideoObject': |
1571 | extract_video_object(e['video'][0]) | |
46933a15 | 1572 | elif item_type == 'VideoObject': |
bae14048 | 1573 | extract_video_object(e) |
4433bb02 S |
1574 | if expected_type is None: |
1575 | continue | |
1576 | else: | |
1577 | break | |
c69701c6 S |
1578 | video = e.get('video') |
1579 | if isinstance(video, dict) and video.get('@type') == 'VideoObject': | |
1580 | extract_video_object(video) | |
4433bb02 S |
1581 | if expected_type is None: |
1582 | continue | |
1583 | else: | |
1584 | break | |
d5c32548 ZM |
1585 | traverse_json_ld(json_ld) |
1586 | ||
90137ca4 | 1587 | return filter_dict(info) |
4ca2a3cf | 1588 | |
135dfa2c | 1589 | def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw): |
f98709af LL |
1590 | return self._parse_json( |
1591 | self._search_regex( | |
1592 | r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>', | |
135dfa2c | 1593 | webpage, 'next.js data', fatal=fatal, **kw), |
1594 | video_id, transform_source=transform_source, fatal=fatal) | |
f98709af | 1595 | |
66f4c04e THD |
1596 | def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__'): |
1597 | ''' Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function. ''' | |
1598 | # not all website do this, but it can be changed | |
1599 | # https://stackoverflow.com/questions/67463109/how-to-change-or-hide-nuxt-and-nuxt-keyword-in-page-source | |
1600 | rectx = re.escape(context_name) | |
1601 | js, arg_keys, arg_vals = self._search_regex( | |
1602 | (r'<script>window\.%s=\(function\((?P<arg_keys>.*?)\)\{return\s(?P<js>\{.*?\})\}\((?P<arg_vals>.+?)\)\);?</script>' % rectx, | |
1603 | r'%s\(.*?\(function\((?P<arg_keys>.*?)\)\{return\s(?P<js>\{.*?\})\}\((?P<arg_vals>.*?)\)' % rectx), | |
1604 | webpage, context_name, group=['js', 'arg_keys', 'arg_vals']) | |
1605 | ||
1606 | args = dict(zip(arg_keys.split(','), arg_vals.split(','))) | |
1607 | ||
1608 | for key, val in args.items(): | |
1609 | if val in ('undefined', 'void 0'): | |
1610 | args[key] = 'null' | |
1611 | ||
1612 | return self._parse_json(js_to_json(js, args), video_id)['data'][0] | |
1613 | ||
27713812 | 1614 | @staticmethod |
f8da79f8 | 1615 | def _hidden_inputs(html): |
586f1cc5 | 1616 | html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html) |
201ea3ee | 1617 | hidden_inputs = {} |
c8498368 S |
1618 | for input in re.findall(r'(?i)(<input[^>]+>)', html): |
1619 | attrs = extract_attributes(input) | |
1620 | if not input: | |
201ea3ee | 1621 | continue |
c8498368 | 1622 | if attrs.get('type') not in ('hidden', 'submit'): |
201ea3ee | 1623 | continue |
c8498368 S |
1624 | name = attrs.get('name') or attrs.get('id') |
1625 | value = attrs.get('value') | |
1626 | if name and value is not None: | |
1627 | hidden_inputs[name] = value | |
201ea3ee | 1628 | return hidden_inputs |
27713812 | 1629 | |
cf61d96d S |
1630 | def _form_hidden_inputs(self, form_id, html): |
1631 | form = self._search_regex( | |
73eb13df | 1632 | r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id, |
cf61d96d S |
1633 | html, '%s form' % form_id, group='form') |
1634 | return self._hidden_inputs(form) | |
1635 | ||
eb8a4433 | 1636 | class FormatSort: |
b050d210 | 1637 | regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$' |
eb8a4433 | 1638 | |
8326b00a | 1639 | default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality', |
176f1866 | 1640 | 'res', 'fps', 'hdr:12', 'codec:vp9.2', 'size', 'br', 'asr', |
f304da8a | 1641 | 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases |
198e3a04 | 1642 | ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr', |
53ed7066 | 1643 | 'height', 'width', 'proto', 'vext', 'abr', 'aext', |
f304da8a | 1644 | 'fps', 'fs_approx', 'source', 'id') |
eb8a4433 | 1645 | |
1646 | settings = { | |
1647 | 'vcodec': {'type': 'ordered', 'regex': True, | |
155d2b48 | 1648 | 'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']}, |
eb8a4433 | 1649 | 'acodec': {'type': 'ordered', 'regex': True, |
a10aa588 | 1650 | 'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']}, |
176f1866 | 1651 | 'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range', |
1652 | 'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]}, | |
f137c99e | 1653 | 'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol', |
f304da8a | 1654 | 'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']}, |
eb8a4433 | 1655 | 'vext': {'type': 'ordered', 'field': 'video_ext', |
91ebc640 | 1656 | 'order': ('mp4', 'webm', 'flv', '', 'none'), |
eb8a4433 | 1657 | 'order_free': ('webm', 'mp4', 'flv', '', 'none')}, |
1658 | 'aext': {'type': 'ordered', 'field': 'audio_ext', | |
1659 | 'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'), | |
1660 | 'order_free': ('opus', 'ogg', 'webm', 'm4a', 'mp3', 'aac', '', 'none')}, | |
1661 | 'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000}, | |
f5510afe | 1662 | 'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple', |
8326b00a | 1663 | 'field': ('vcodec', 'acodec'), |
1664 | 'function': lambda it: int(any(v != 'none' for v in it))}, | |
f983b875 | 1665 | 'ie_pref': {'priority': True, 'type': 'extractor'}, |
63be1aab | 1666 | 'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)}, |
1667 | 'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)}, | |
10beccc9 | 1668 | 'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1}, |
1669 | 'quality': {'convert': 'float', 'default': -1}, | |
eb8a4433 | 1670 | 'filesize': {'convert': 'bytes'}, |
f137c99e | 1671 | 'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'}, |
1672 | 'id': {'convert': 'string', 'field': 'format_id'}, | |
eb8a4433 | 1673 | 'height': {'convert': 'float_none'}, |
1674 | 'width': {'convert': 'float_none'}, | |
1675 | 'fps': {'convert': 'float_none'}, | |
1676 | 'tbr': {'convert': 'float_none'}, | |
1677 | 'vbr': {'convert': 'float_none'}, | |
1678 | 'abr': {'convert': 'float_none'}, | |
1679 | 'asr': {'convert': 'float_none'}, | |
10beccc9 | 1680 | 'source': {'convert': 'float', 'field': 'source_preference', 'default': -1}, |
63be1aab | 1681 | |
eb8a4433 | 1682 | 'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')}, |
63be1aab | 1683 | 'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True}, |
1684 | 'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')}, | |
1685 | 'ext': {'type': 'combined', 'field': ('vext', 'aext')}, | |
f5510afe | 1686 | 'res': {'type': 'multiple', 'field': ('height', 'width'), |
dbf5416a | 1687 | 'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))}, |
63be1aab | 1688 | |
19188702 | 1689 | # For compatibility with youtube-dl |
1690 | 'format_id': {'type': 'alias', 'field': 'id'}, | |
1691 | 'preference': {'type': 'alias', 'field': 'ie_pref'}, | |
1692 | 'language_preference': {'type': 'alias', 'field': 'lang'}, | |
63be1aab | 1693 | 'source_preference': {'type': 'alias', 'field': 'source'}, |
08d30158 | 1694 | 'protocol': {'type': 'alias', 'field': 'proto'}, |
63be1aab | 1695 | 'filesize_approx': {'type': 'alias', 'field': 'fs_approx'}, |
08d30158 | 1696 | |
1697 | # Deprecated | |
1698 | 'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True}, | |
1699 | 'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True}, | |
1700 | 'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True}, | |
1701 | 'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True}, | |
1702 | 'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True}, | |
1703 | 'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True}, | |
1704 | 'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True}, | |
1705 | 'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True}, | |
1706 | 'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True}, | |
1707 | 'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True}, | |
1708 | 'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True}, | |
1709 | 'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True}, | |
1710 | 'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True}, | |
1711 | 'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True}, | |
1712 | 'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}, | |
1713 | 'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True}, | |
1714 | 'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}, | |
1715 | 'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True}, | |
1716 | 'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}, | |
1717 | 'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True}, | |
63be1aab | 1718 | } |
eb8a4433 | 1719 | |
f304da8a | 1720 | def __init__(self, ie, field_preference): |
1721 | self._order = [] | |
1722 | self.ydl = ie._downloader | |
1723 | self.evaluate_params(self.ydl.params, field_preference) | |
1724 | if ie.get_param('verbose'): | |
1725 | self.print_verbose_info(self.ydl.write_debug) | |
eb8a4433 | 1726 | |
1727 | def _get_field_setting(self, field, key): | |
1728 | if field not in self.settings: | |
ee8dd27a | 1729 | if key in ('forced', 'priority'): |
1730 | return False | |
1731 | self.ydl.deprecation_warning( | |
1732 | f'Using arbitrary fields ({field}) for format sorting is deprecated ' | |
1733 | 'and may be removed in a future version') | |
eb8a4433 | 1734 | self.settings[field] = {} |
1735 | propObj = self.settings[field] | |
1736 | if key not in propObj: | |
1737 | type = propObj.get('type') | |
1738 | if key == 'field': | |
1739 | default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field | |
1740 | elif key == 'convert': | |
1741 | default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore' | |
4bcc7bd1 | 1742 | else: |
f5510afe | 1743 | default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None) |
eb8a4433 | 1744 | propObj[key] = default |
1745 | return propObj[key] | |
1746 | ||
1747 | def _resolve_field_value(self, field, value, convertNone=False): | |
1748 | if value is None: | |
1749 | if not convertNone: | |
1750 | return None | |
4bcc7bd1 | 1751 | else: |
eb8a4433 | 1752 | value = value.lower() |
1753 | conversion = self._get_field_setting(field, 'convert') | |
1754 | if conversion == 'ignore': | |
1755 | return None | |
1756 | if conversion == 'string': | |
1757 | return value | |
1758 | elif conversion == 'float_none': | |
1759 | return float_or_none(value) | |
1760 | elif conversion == 'bytes': | |
1761 | return FileDownloader.parse_bytes(value) | |
1762 | elif conversion == 'order': | |
da9be05e | 1763 | order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order') |
eb8a4433 | 1764 | use_regex = self._get_field_setting(field, 'regex') |
1765 | list_length = len(order_list) | |
1766 | empty_pos = order_list.index('') if '' in order_list else list_length + 1 | |
1767 | if use_regex and value is not None: | |
da9be05e | 1768 | for i, regex in enumerate(order_list): |
eb8a4433 | 1769 | if regex and re.match(regex, value): |
1770 | return list_length - i | |
1771 | return list_length - empty_pos # not in list | |
1772 | else: # not regex or value = None | |
1773 | return list_length - (order_list.index(value) if value in order_list else empty_pos) | |
1774 | else: | |
1775 | if value.isnumeric(): | |
1776 | return float(value) | |
4bcc7bd1 | 1777 | else: |
eb8a4433 | 1778 | self.settings[field]['convert'] = 'string' |
1779 | return value | |
1780 | ||
1781 | def evaluate_params(self, params, sort_extractor): | |
1782 | self._use_free_order = params.get('prefer_free_formats', False) | |
1783 | self._sort_user = params.get('format_sort', []) | |
1784 | self._sort_extractor = sort_extractor | |
1785 | ||
1786 | def add_item(field, reverse, closest, limit_text): | |
1787 | field = field.lower() | |
1788 | if field in self._order: | |
1789 | return | |
1790 | self._order.append(field) | |
1791 | limit = self._resolve_field_value(field, limit_text) | |
1792 | data = { | |
1793 | 'reverse': reverse, | |
1794 | 'closest': False if limit is None else closest, | |
1795 | 'limit_text': limit_text, | |
1796 | 'limit': limit} | |
1797 | if field in self.settings: | |
1798 | self.settings[field].update(data) | |
1799 | else: | |
1800 | self.settings[field] = data | |
1801 | ||
1802 | sort_list = ( | |
1803 | tuple(field for field in self.default if self._get_field_setting(field, 'forced')) | |
1804 | + (tuple() if params.get('format_sort_force', False) | |
1805 | else tuple(field for field in self.default if self._get_field_setting(field, 'priority'))) | |
1806 | + tuple(self._sort_user) + tuple(sort_extractor) + self.default) | |
1807 | ||
1808 | for item in sort_list: | |
1809 | match = re.match(self.regex, item) | |
1810 | if match is None: | |
1811 | raise ExtractorError('Invalid format sort string "%s" given by extractor' % item) | |
1812 | field = match.group('field') | |
1813 | if field is None: | |
1814 | continue | |
1815 | if self._get_field_setting(field, 'type') == 'alias': | |
ee8dd27a | 1816 | alias, field = field, self._get_field_setting(field, 'field') |
08d30158 | 1817 | if self._get_field_setting(alias, 'deprecated'): |
19188702 | 1818 | self.ydl.deprecation_warning( |
1819 | f'Format sorting alias {alias} is deprecated ' | |
1820 | f'and may be removed in a future version. Please use {field} instead') | |
eb8a4433 | 1821 | reverse = match.group('reverse') is not None |
b050d210 | 1822 | closest = match.group('separator') == '~' |
eb8a4433 | 1823 | limit_text = match.group('limit') |
1824 | ||
1825 | has_limit = limit_text is not None | |
1826 | has_multiple_fields = self._get_field_setting(field, 'type') == 'combined' | |
1827 | has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit') | |
1828 | ||
1829 | fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,) | |
b5ae35ee | 1830 | limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple() |
eb8a4433 | 1831 | limit_count = len(limits) |
1832 | for (i, f) in enumerate(fields): | |
1833 | add_item(f, reverse, closest, | |
1834 | limits[i] if i < limit_count | |
1835 | else limits[0] if has_limit and not has_multiple_limits | |
1836 | else None) | |
1837 | ||
0760b0a7 | 1838 | def print_verbose_info(self, write_debug): |
b31fdeed | 1839 | if self._sort_user: |
0760b0a7 | 1840 | write_debug('Sort order given by user: %s' % ', '.join(self._sort_user)) |
eb8a4433 | 1841 | if self._sort_extractor: |
0760b0a7 | 1842 | write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor)) |
1843 | write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % ( | |
eb8a4433 | 1844 | '+' if self._get_field_setting(field, 'reverse') else '', field, |
1845 | '%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':', | |
1846 | self._get_field_setting(field, 'limit_text'), | |
1847 | self._get_field_setting(field, 'limit')) | |
1848 | if self._get_field_setting(field, 'limit_text') is not None else '') | |
1849 | for field in self._order if self._get_field_setting(field, 'visible')])) | |
1850 | ||
1851 | def _calculate_field_preference_from_value(self, format, field, type, value): | |
1852 | reverse = self._get_field_setting(field, 'reverse') | |
1853 | closest = self._get_field_setting(field, 'closest') | |
1854 | limit = self._get_field_setting(field, 'limit') | |
1855 | ||
1856 | if type == 'extractor': | |
1857 | maximum = self._get_field_setting(field, 'max') | |
1858 | if value is None or (maximum is not None and value >= maximum): | |
f983b875 | 1859 | value = -1 |
eb8a4433 | 1860 | elif type == 'boolean': |
1861 | in_list = self._get_field_setting(field, 'in_list') | |
1862 | not_in_list = self._get_field_setting(field, 'not_in_list') | |
1863 | value = 0 if ((in_list is None or value in in_list) and (not_in_list is None or value not in not_in_list)) else -1 | |
1864 | elif type == 'ordered': | |
1865 | value = self._resolve_field_value(field, value, True) | |
1866 | ||
1867 | # try to convert to number | |
6a04a74e | 1868 | val_num = float_or_none(value, default=self._get_field_setting(field, 'default')) |
eb8a4433 | 1869 | is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None |
1870 | if is_num: | |
1871 | value = val_num | |
1872 | ||
1873 | return ((-10, 0) if value is None | |
1874 | else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher | |
1875 | else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest | |
1876 | else (0, value, 0) if not reverse and (limit is None or value <= limit) | |
1877 | else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit | |
1878 | else (-1, value, 0)) | |
1879 | ||
1880 | def _calculate_field_preference(self, format, field): | |
1881 | type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple | |
1882 | get_value = lambda f: format.get(self._get_field_setting(f, 'field')) | |
1883 | if type == 'multiple': | |
1884 | type = 'field' # Only 'field' is allowed in multiple for now | |
1885 | actual_fields = self._get_field_setting(field, 'field') | |
1886 | ||
f5510afe | 1887 | value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields) |
eb8a4433 | 1888 | else: |
1889 | value = get_value(field) | |
1890 | return self._calculate_field_preference_from_value(format, field, type, value) | |
1891 | ||
1892 | def calculate_preference(self, format): | |
1893 | # Determine missing protocol | |
1894 | if not format.get('protocol'): | |
1895 | format['protocol'] = determine_protocol(format) | |
1896 | ||
1897 | # Determine missing ext | |
1898 | if not format.get('ext') and 'url' in format: | |
1899 | format['ext'] = determine_ext(format['url']) | |
1900 | if format.get('vcodec') == 'none': | |
8326b00a | 1901 | format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none' |
eb8a4433 | 1902 | format['video_ext'] = 'none' |
1903 | else: | |
1904 | format['video_ext'] = format['ext'] | |
1905 | format['audio_ext'] = 'none' | |
1906 | # if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported? | |
1907 | # format['preference'] = -1000 | |
1908 | ||
1909 | # Determine missing bitrates | |
1910 | if format.get('tbr') is None: | |
1911 | if format.get('vbr') is not None and format.get('abr') is not None: | |
1912 | format['tbr'] = format.get('vbr', 0) + format.get('abr', 0) | |
1913 | else: | |
b5ae35ee | 1914 | if format.get('vcodec') != 'none' and format.get('vbr') is None: |
eb8a4433 | 1915 | format['vbr'] = format.get('tbr') - format.get('abr', 0) |
b5ae35ee | 1916 | if format.get('acodec') != 'none' and format.get('abr') is None: |
eb8a4433 | 1917 | format['abr'] = format.get('tbr') - format.get('vbr', 0) |
1918 | ||
1919 | return tuple(self._calculate_field_preference(format, field) for field in self._order) | |
1920 | ||
1921 | def _sort_formats(self, formats, field_preference=[]): | |
1922 | if not formats: | |
88acdbc2 | 1923 | return |
f304da8a | 1924 | format_sort = self.FormatSort(self, field_preference) |
eb8a4433 | 1925 | formats.sort(key=lambda f: format_sort.calculate_preference(f)) |
59040888 | 1926 | |
96a53167 S |
1927 | def _check_formats(self, formats, video_id): |
1928 | if formats: | |
1929 | formats[:] = filter( | |
1930 | lambda f: self._is_valid_url( | |
1931 | f['url'], video_id, | |
1932 | item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'), | |
1933 | formats) | |
1934 | ||
f5bdb444 S |
1935 | @staticmethod |
1936 | def _remove_duplicate_formats(formats): | |
1937 | format_urls = set() | |
1938 | unique_formats = [] | |
1939 | for f in formats: | |
1940 | if f['url'] not in format_urls: | |
1941 | format_urls.add(f['url']) | |
1942 | unique_formats.append(f) | |
1943 | formats[:] = unique_formats | |
1944 | ||
45024183 | 1945 | def _is_valid_url(self, url, video_id, item='video', headers={}): |
2f0f6578 S |
1946 | url = self._proto_relative_url(url, scheme='http:') |
1947 | # For now assume non HTTP(S) URLs always valid | |
1948 | if not (url.startswith('http://') or url.startswith('https://')): | |
1949 | return True | |
96a53167 | 1950 | try: |
45024183 | 1951 | self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers) |
96a53167 | 1952 | return True |
8bdd16b4 | 1953 | except ExtractorError as e: |
25e911a9 | 1954 | self.to_screen( |
8bdd16b4 | 1955 | '%s: %s URL is invalid, skipping: %s' |
1956 | % (video_id, item, error_to_compat_str(e.cause))) | |
25e911a9 | 1957 | return False |
96a53167 | 1958 | |
20991253 | 1959 | def http_scheme(self): |
1ede5b24 | 1960 | """ Either "http:" or "https:", depending on the user's preferences """ |
20991253 PH |
1961 | return ( |
1962 | 'http:' | |
a06916d9 | 1963 | if self.get_param('prefer_insecure', False) |
20991253 PH |
1964 | else 'https:') |
1965 | ||
57c7411f PH |
1966 | def _proto_relative_url(self, url, scheme=None): |
1967 | if url is None: | |
1968 | return url | |
1969 | if url.startswith('//'): | |
1970 | if scheme is None: | |
1971 | scheme = self.http_scheme() | |
1972 | return scheme + url | |
1973 | else: | |
1974 | return url | |
1975 | ||
4094b6e3 PH |
1976 | def _sleep(self, timeout, video_id, msg_template=None): |
1977 | if msg_template is None: | |
f1a9d64e | 1978 | msg_template = '%(video_id)s: Waiting for %(timeout)s seconds' |
4094b6e3 PH |
1979 | msg = msg_template % {'video_id': video_id, 'timeout': timeout} |
1980 | self.to_screen(msg) | |
1981 | time.sleep(timeout) | |
1982 | ||
f983b875 | 1983 | def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None, |
4de61310 | 1984 | transform_source=lambda s: fix_xml_ampersands(s).strip(), |
7360c06f | 1985 | fatal=True, m3u8_id=None, data=None, headers={}, query={}): |
f036a632 JMF |
1986 | manifest = self._download_xml( |
1987 | manifest_url, video_id, 'Downloading f4m manifest', | |
97f4aecf S |
1988 | 'Unable to download f4m manifest', |
1989 | # Some manifests may be malformed, e.g. prosiebensat1 generated manifests | |
067aa17e | 1990 | # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244) |
4de61310 | 1991 | transform_source=transform_source, |
7360c06f | 1992 | fatal=fatal, data=data, headers=headers, query=query) |
4de61310 S |
1993 | |
1994 | if manifest is False: | |
8d29e47f | 1995 | return [] |
31bb8d3f | 1996 | |
0fdbb332 | 1997 | return self._parse_f4m_formats( |
f983b875 | 1998 | manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id, |
448bb5f3 | 1999 | transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id) |
0fdbb332 | 2000 | |
f983b875 | 2001 | def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None, |
0fdbb332 | 2002 | transform_source=lambda s: fix_xml_ampersands(s).strip(), |
448bb5f3 | 2003 | fatal=True, m3u8_id=None): |
f9934b96 | 2004 | if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal: |
d9eb580a S |
2005 | return [] |
2006 | ||
7a5c1cfe | 2007 | # currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy |
fb72ec58 | 2008 | akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0') |
2009 | if akamai_pv is not None and ';' in akamai_pv.text: | |
2010 | playerVerificationChallenge = akamai_pv.text.split(';')[0] | |
2011 | if playerVerificationChallenge.strip() != '': | |
2012 | return [] | |
2013 | ||
31bb8d3f | 2014 | formats = [] |
7a47d07c | 2015 | manifest_version = '1.0' |
b2527359 | 2016 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media') |
34e48bed | 2017 | if not media_nodes: |
7a47d07c | 2018 | manifest_version = '2.0' |
34e48bed | 2019 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') |
b22ca762 | 2020 | # Remove unsupported DRM protected media from final formats |
067aa17e | 2021 | # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573). |
b22ca762 S |
2022 | media_nodes = remove_encrypted_media(media_nodes) |
2023 | if not media_nodes: | |
2024 | return formats | |
48107c19 S |
2025 | |
2026 | manifest_base_url = get_base_url(manifest) | |
0a5685b2 | 2027 | |
a6571f10 | 2028 | bootstrap_info = xpath_element( |
0a5685b2 YCH |
2029 | manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'], |
2030 | 'bootstrap info', default=None) | |
2031 | ||
edd6074c RA |
2032 | vcodec = None |
2033 | mime_type = xpath_text( | |
2034 | manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'], | |
2035 | 'base URL', default=None) | |
2036 | if mime_type and mime_type.startswith('audio/'): | |
2037 | vcodec = 'none' | |
2038 | ||
b2527359 | 2039 | for i, media_el in enumerate(media_nodes): |
77b8b4e6 S |
2040 | tbr = int_or_none(media_el.attrib.get('bitrate')) |
2041 | width = int_or_none(media_el.attrib.get('width')) | |
2042 | height = int_or_none(media_el.attrib.get('height')) | |
34921b43 | 2043 | format_id = join_nonempty(f4m_id, tbr or i) |
448bb5f3 YCH |
2044 | # If <bootstrapInfo> is present, the specified f4m is a |
2045 | # stream-level manifest, and only set-level manifests may refer to | |
2046 | # external resources. See section 11.4 and section 4 of F4M spec | |
2047 | if bootstrap_info is None: | |
2048 | media_url = None | |
2049 | # @href is introduced in 2.0, see section 11.6 of F4M spec | |
2050 | if manifest_version == '2.0': | |
2051 | media_url = media_el.attrib.get('href') | |
2052 | if media_url is None: | |
2053 | media_url = media_el.attrib.get('url') | |
31c746e5 S |
2054 | if not media_url: |
2055 | continue | |
cc357c4d S |
2056 | manifest_url = ( |
2057 | media_url if media_url.startswith('http://') or media_url.startswith('https://') | |
48107c19 | 2058 | else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url)) |
70f0f5a8 S |
2059 | # If media_url is itself a f4m manifest do the recursive extraction |
2060 | # since bitrates in parent manifest (this one) and media_url manifest | |
2061 | # may differ leading to inability to resolve the format by requested | |
2062 | # bitrate in f4m downloader | |
240b6045 YCH |
2063 | ext = determine_ext(manifest_url) |
2064 | if ext == 'f4m': | |
77b8b4e6 | 2065 | f4m_formats = self._extract_f4m_formats( |
f983b875 | 2066 | manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id, |
77b8b4e6 S |
2067 | transform_source=transform_source, fatal=fatal) |
2068 | # Sometimes stream-level manifest contains single media entry that | |
2069 | # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player). | |
2070 | # At the same time parent's media entry in set-level manifest may | |
2071 | # contain it. We will copy it from parent in such cases. | |
2072 | if len(f4m_formats) == 1: | |
2073 | f = f4m_formats[0] | |
2074 | f.update({ | |
2075 | 'tbr': f.get('tbr') or tbr, | |
2076 | 'width': f.get('width') or width, | |
2077 | 'height': f.get('height') or height, | |
2078 | 'format_id': f.get('format_id') if not tbr else format_id, | |
edd6074c | 2079 | 'vcodec': vcodec, |
77b8b4e6 S |
2080 | }) |
2081 | formats.extend(f4m_formats) | |
70f0f5a8 | 2082 | continue |
240b6045 YCH |
2083 | elif ext == 'm3u8': |
2084 | formats.extend(self._extract_m3u8_formats( | |
2085 | manifest_url, video_id, 'mp4', preference=preference, | |
f983b875 | 2086 | quality=quality, m3u8_id=m3u8_id, fatal=fatal)) |
240b6045 | 2087 | continue |
31bb8d3f | 2088 | formats.append({ |
77b8b4e6 | 2089 | 'format_id': format_id, |
31bb8d3f | 2090 | 'url': manifest_url, |
30d0b549 | 2091 | 'manifest_url': manifest_url, |
a6571f10 | 2092 | 'ext': 'flv' if bootstrap_info is not None else None, |
187ee66c | 2093 | 'protocol': 'f4m', |
b2527359 | 2094 | 'tbr': tbr, |
77b8b4e6 S |
2095 | 'width': width, |
2096 | 'height': height, | |
edd6074c | 2097 | 'vcodec': vcodec, |
60ca389c | 2098 | 'preference': preference, |
f983b875 | 2099 | 'quality': quality, |
31bb8d3f | 2100 | }) |
31bb8d3f JMF |
2101 | return formats |
2102 | ||
f983b875 | 2103 | def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None): |
16da9bbc | 2104 | return { |
34921b43 | 2105 | 'format_id': join_nonempty(m3u8_id, 'meta'), |
704df56d PH |
2106 | 'url': m3u8_url, |
2107 | 'ext': ext, | |
2108 | 'protocol': 'm3u8', | |
37768f92 | 2109 | 'preference': preference - 100 if preference else -100, |
f983b875 | 2110 | 'quality': quality, |
704df56d PH |
2111 | 'resolution': 'multiple', |
2112 | 'format_note': 'Quality selection URL', | |
16da9bbc YCH |
2113 | } |
2114 | ||
b5ae35ee | 2115 | def _report_ignoring_subs(self, name): |
2116 | self.report_warning(bug_reports_message( | |
2117 | f'Ignoring subtitle tracks found in the {name} manifest; ' | |
2118 | 'if any subtitle tracks are missing,' | |
2119 | ), only_once=True) | |
2120 | ||
a0c3b2d5 F |
2121 | def _extract_m3u8_formats(self, *args, **kwargs): |
2122 | fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs) | |
2123 | if subs: | |
b5ae35ee | 2124 | self._report_ignoring_subs('HLS') |
a0c3b2d5 F |
2125 | return fmts |
2126 | ||
2127 | def _extract_m3u8_formats_and_subtitles( | |
177877c5 | 2128 | self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native', |
a0c3b2d5 F |
2129 | preference=None, quality=None, m3u8_id=None, note=None, |
2130 | errnote=None, fatal=True, live=False, data=None, headers={}, | |
2131 | query={}): | |
2132 | ||
dbd82a1d | 2133 | res = self._download_webpage_handle( |
81515ad9 | 2134 | m3u8_url, video_id, |
37a3bb66 | 2135 | note='Downloading m3u8 information' if note is None else note, |
2136 | errnote='Failed to download m3u8 information' if errnote is None else errnote, | |
7360c06f | 2137 | fatal=fatal, data=data, headers=headers, query=query) |
cb252080 | 2138 | |
dbd82a1d | 2139 | if res is False: |
a0c3b2d5 | 2140 | return [], {} |
cb252080 | 2141 | |
dbd82a1d | 2142 | m3u8_doc, urlh = res |
37113045 | 2143 | m3u8_url = urlh.geturl() |
9cdffeeb | 2144 | |
a0c3b2d5 | 2145 | return self._parse_m3u8_formats_and_subtitles( |
cb252080 | 2146 | m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol, |
310c2ed2 | 2147 | preference=preference, quality=quality, m3u8_id=m3u8_id, |
2148 | note=note, errnote=errnote, fatal=fatal, live=live, data=data, | |
2149 | headers=headers, query=query, video_id=video_id) | |
cb252080 | 2150 | |
a0c3b2d5 | 2151 | def _parse_m3u8_formats_and_subtitles( |
42676437 | 2152 | self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native', |
a0c3b2d5 F |
2153 | preference=None, quality=None, m3u8_id=None, live=False, note=None, |
2154 | errnote=None, fatal=True, data=None, headers={}, query={}, | |
2155 | video_id=None): | |
60755938 | 2156 | formats, subtitles = [], {} |
a0c3b2d5 | 2157 | |
6b993ca7 | 2158 | has_drm = re.search('|'.join([ |
2159 | r'#EXT-X-FAXS-CM:', # Adobe Flash Access | |
2160 | r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd://', # Apple FairPlay | |
2161 | ]), m3u8_doc) | |
a0c3b2d5 | 2162 | |
60755938 | 2163 | def format_url(url): |
2164 | return url if re.match(r'^https?://', url) else compat_urlparse.urljoin(m3u8_url, url) | |
2165 | ||
2166 | if self.get_param('hls_split_discontinuity', False): | |
2167 | def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None): | |
2168 | if not m3u8_doc: | |
2169 | if not manifest_url: | |
2170 | return [] | |
2171 | m3u8_doc = self._download_webpage( | |
2172 | manifest_url, video_id, fatal=fatal, data=data, headers=headers, | |
2173 | note=False, errnote='Failed to download m3u8 playlist information') | |
2174 | if m3u8_doc is False: | |
2175 | return [] | |
2176 | return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines())) | |
0def7587 | 2177 | |
60755938 | 2178 | else: |
2179 | def _extract_m3u8_playlist_indices(*args, **kwargs): | |
2180 | return [None] | |
310c2ed2 | 2181 | |
cb252080 S |
2182 | # References: |
2183 | # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21 | |
067aa17e S |
2184 | # 2. https://github.com/ytdl-org/youtube-dl/issues/12211 |
2185 | # 3. https://github.com/ytdl-org/youtube-dl/issues/18923 | |
cb252080 S |
2186 | |
2187 | # We should try extracting formats only from master playlists [1, 4.3.4], | |
2188 | # i.e. playlists that describe available qualities. On the other hand | |
2189 | # media playlists [1, 4.3.3] should be returned as is since they contain | |
2190 | # just the media without qualities renditions. | |
9cdffeeb | 2191 | # Fortunately, master playlist can be easily distinguished from media |
cb252080 | 2192 | # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4] |
a0566bbf | 2193 | # master playlist tags MUST NOT appear in a media playlist and vice versa. |
cb252080 S |
2194 | # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every |
2195 | # media playlist and MUST NOT appear in master playlist thus we can | |
2196 | # clearly detect media playlist with this criterion. | |
2197 | ||
9cdffeeb | 2198 | if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is |
60755938 | 2199 | formats = [{ |
34921b43 | 2200 | 'format_id': join_nonempty(m3u8_id, idx), |
60755938 | 2201 | 'format_index': idx, |
42676437 | 2202 | 'url': m3u8_url or encode_data_uri(m3u8_doc.encode('utf-8'), 'application/x-mpegurl'), |
60755938 | 2203 | 'ext': ext, |
2204 | 'protocol': entry_protocol, | |
2205 | 'preference': preference, | |
2206 | 'quality': quality, | |
88acdbc2 | 2207 | 'has_drm': has_drm, |
60755938 | 2208 | } for idx in _extract_m3u8_playlist_indices(m3u8_doc=m3u8_doc)] |
310c2ed2 | 2209 | |
a0c3b2d5 | 2210 | return formats, subtitles |
cb252080 S |
2211 | |
2212 | groups = {} | |
2213 | last_stream_inf = {} | |
2214 | ||
2215 | def extract_media(x_media_line): | |
2216 | media = parse_m3u8_attributes(x_media_line) | |
2217 | # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED | |
2218 | media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME') | |
2219 | if not (media_type and group_id and name): | |
2220 | return | |
2221 | groups.setdefault(group_id, []).append(media) | |
a0c3b2d5 F |
2222 | # <https://tools.ietf.org/html/rfc8216#section-4.3.4.1> |
2223 | if media_type == 'SUBTITLES': | |
3907333c | 2224 | # According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the |
2225 | # EXT-X-MEDIA tag if the media type is SUBTITLES. | |
2226 | # However, lack of URI has been spotted in the wild. | |
2227 | # e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339 | |
2228 | if not media.get('URI'): | |
2229 | return | |
a0c3b2d5 F |
2230 | url = format_url(media['URI']) |
2231 | sub_info = { | |
2232 | 'url': url, | |
2233 | 'ext': determine_ext(url), | |
2234 | } | |
4a2f19ab F |
2235 | if sub_info['ext'] == 'm3u8': |
2236 | # Per RFC 8216 §3.1, the only possible subtitle format m3u8 | |
2237 | # files may contain is WebVTT: | |
2238 | # <https://tools.ietf.org/html/rfc8216#section-3.1> | |
2239 | sub_info['ext'] = 'vtt' | |
2240 | sub_info['protocol'] = 'm3u8_native' | |
37a3bb66 | 2241 | lang = media.get('LANGUAGE') or 'und' |
a0c3b2d5 | 2242 | subtitles.setdefault(lang, []).append(sub_info) |
cb252080 S |
2243 | if media_type not in ('VIDEO', 'AUDIO'): |
2244 | return | |
2245 | media_url = media.get('URI') | |
2246 | if media_url: | |
310c2ed2 | 2247 | manifest_url = format_url(media_url) |
60755938 | 2248 | formats.extend({ |
34921b43 | 2249 | 'format_id': join_nonempty(m3u8_id, group_id, name, idx), |
60755938 | 2250 | 'format_note': name, |
2251 | 'format_index': idx, | |
2252 | 'url': manifest_url, | |
2253 | 'manifest_url': m3u8_url, | |
2254 | 'language': media.get('LANGUAGE'), | |
2255 | 'ext': ext, | |
2256 | 'protocol': entry_protocol, | |
2257 | 'preference': preference, | |
2258 | 'quality': quality, | |
2259 | 'vcodec': 'none' if media_type == 'AUDIO' else None, | |
2260 | } for idx in _extract_m3u8_playlist_indices(manifest_url)) | |
cb252080 S |
2261 | |
2262 | def build_stream_name(): | |
2263 | # Despite specification does not mention NAME attribute for | |
3019cb0c S |
2264 | # EXT-X-STREAM-INF tag it still sometimes may be present (see [1] |
2265 | # or vidio test in TestInfoExtractor.test_parse_m3u8_formats) | |
ddd258f9 | 2266 | # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015 |
cb252080 S |
2267 | stream_name = last_stream_inf.get('NAME') |
2268 | if stream_name: | |
2269 | return stream_name | |
2270 | # If there is no NAME in EXT-X-STREAM-INF it will be obtained | |
2271 | # from corresponding rendition group | |
2272 | stream_group_id = last_stream_inf.get('VIDEO') | |
2273 | if not stream_group_id: | |
2274 | return | |
2275 | stream_group = groups.get(stream_group_id) | |
2276 | if not stream_group: | |
2277 | return stream_group_id | |
2278 | rendition = stream_group[0] | |
2279 | return rendition.get('NAME') or stream_group_id | |
2280 | ||
379306ef | 2281 | # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the |
2bfc1d9d RA |
2282 | # chance to detect video only formats when EXT-X-STREAM-INF tags |
2283 | # precede EXT-X-MEDIA tags in HLS manifest such as [3]. | |
2284 | for line in m3u8_doc.splitlines(): | |
2285 | if line.startswith('#EXT-X-MEDIA:'): | |
2286 | extract_media(line) | |
2287 | ||
704df56d PH |
2288 | for line in m3u8_doc.splitlines(): |
2289 | if line.startswith('#EXT-X-STREAM-INF:'): | |
cb252080 | 2290 | last_stream_inf = parse_m3u8_attributes(line) |
704df56d PH |
2291 | elif line.startswith('#') or not line.strip(): |
2292 | continue | |
2293 | else: | |
9c99bef7 | 2294 | tbr = float_or_none( |
3089bc74 S |
2295 | last_stream_inf.get('AVERAGE-BANDWIDTH') |
2296 | or last_stream_inf.get('BANDWIDTH'), scale=1000) | |
30d0b549 | 2297 | manifest_url = format_url(line.strip()) |
5ef62fc4 | 2298 | |
60755938 | 2299 | for idx in _extract_m3u8_playlist_indices(manifest_url): |
2300 | format_id = [m3u8_id, None, idx] | |
310c2ed2 | 2301 | # Bandwidth of live streams may differ over time thus making |
2302 | # format_id unpredictable. So it's better to keep provided | |
2303 | # format_id intact. | |
2304 | if not live: | |
60755938 | 2305 | stream_name = build_stream_name() |
34921b43 | 2306 | format_id[1] = stream_name or '%d' % (tbr or len(formats)) |
310c2ed2 | 2307 | f = { |
34921b43 | 2308 | 'format_id': join_nonempty(*format_id), |
60755938 | 2309 | 'format_index': idx, |
310c2ed2 | 2310 | 'url': manifest_url, |
2311 | 'manifest_url': m3u8_url, | |
2312 | 'tbr': tbr, | |
2313 | 'ext': ext, | |
2314 | 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')), | |
2315 | 'protocol': entry_protocol, | |
2316 | 'preference': preference, | |
2317 | 'quality': quality, | |
2318 | } | |
2319 | resolution = last_stream_inf.get('RESOLUTION') | |
2320 | if resolution: | |
2321 | mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution) | |
2322 | if mobj: | |
2323 | f['width'] = int(mobj.group('width')) | |
2324 | f['height'] = int(mobj.group('height')) | |
2325 | # Unified Streaming Platform | |
2326 | mobj = re.search( | |
2327 | r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url']) | |
2328 | if mobj: | |
2329 | abr, vbr = mobj.groups() | |
2330 | abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000) | |
2331 | f.update({ | |
2332 | 'vbr': vbr, | |
2333 | 'abr': abr, | |
2334 | }) | |
2335 | codecs = parse_codecs(last_stream_inf.get('CODECS')) | |
2336 | f.update(codecs) | |
2337 | audio_group_id = last_stream_inf.get('AUDIO') | |
2338 | # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which | |
2339 | # references a rendition group MUST have a CODECS attribute. | |
2340 | # However, this is not always respected, for example, [2] | |
2341 | # contains EXT-X-STREAM-INF tag which references AUDIO | |
2342 | # rendition group but does not have CODECS and despite | |
2343 | # referencing an audio group it represents a complete | |
2344 | # (with audio and video) format. So, for such cases we will | |
2345 | # ignore references to rendition groups and treat them | |
2346 | # as complete formats. | |
2347 | if audio_group_id and codecs and f.get('vcodec') != 'none': | |
2348 | audio_group = groups.get(audio_group_id) | |
2349 | if audio_group and audio_group[0].get('URI'): | |
2350 | # TODO: update acodec for audio only formats with | |
2351 | # the same GROUP-ID | |
2352 | f['acodec'] = 'none' | |
fc21af50 | 2353 | if not f.get('ext'): |
2354 | f['ext'] = 'm4a' if f.get('vcodec') == 'none' else 'mp4' | |
310c2ed2 | 2355 | formats.append(f) |
2356 | ||
2357 | # for DailyMotion | |
2358 | progressive_uri = last_stream_inf.get('PROGRESSIVE-URI') | |
2359 | if progressive_uri: | |
2360 | http_f = f.copy() | |
2361 | del http_f['manifest_url'] | |
2362 | http_f.update({ | |
2363 | 'format_id': f['format_id'].replace('hls-', 'http-'), | |
2364 | 'protocol': 'http', | |
2365 | 'url': progressive_uri, | |
2366 | }) | |
2367 | formats.append(http_f) | |
5ef62fc4 | 2368 | |
cb252080 | 2369 | last_stream_inf = {} |
a0c3b2d5 | 2370 | return formats, subtitles |
704df56d | 2371 | |
3cf4b91d C |
2372 | def _extract_m3u8_vod_duration( |
2373 | self, m3u8_vod_url, video_id, note=None, errnote=None, data=None, headers={}, query={}): | |
2374 | ||
2375 | m3u8_vod = self._download_webpage( | |
2376 | m3u8_vod_url, video_id, | |
2377 | note='Downloading m3u8 VOD manifest' if note is None else note, | |
2378 | errnote='Failed to download VOD manifest' if errnote is None else errnote, | |
2379 | fatal=False, data=data, headers=headers, query=query) | |
2380 | ||
2381 | return self._parse_m3u8_vod_duration(m3u8_vod or '', video_id) | |
2382 | ||
2383 | def _parse_m3u8_vod_duration(self, m3u8_vod, video_id): | |
2384 | if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod: | |
2385 | return None | |
2386 | ||
2387 | return int(sum( | |
2388 | float(line[len('#EXTINF:'):].split(',')[0]) | |
2389 | for line in m3u8_vod.splitlines() if line.startswith('#EXTINF:'))) or None | |
2390 | ||
a107193e S |
2391 | @staticmethod |
2392 | def _xpath_ns(path, namespace=None): | |
2393 | if not namespace: | |
2394 | return path | |
2395 | out = [] | |
2396 | for c in path.split('/'): | |
2397 | if not c or c == '.': | |
2398 | out.append(c) | |
2399 | else: | |
2400 | out.append('{%s}%s' % (namespace, c)) | |
2401 | return '/'.join(out) | |
2402 | ||
da1c94ee | 2403 | def _extract_smil_formats_and_subtitles(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None): |
09f572fb | 2404 | smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source) |
a107193e | 2405 | |
995029a1 PH |
2406 | if smil is False: |
2407 | assert not fatal | |
774a46c5 | 2408 | return [], {} |
e89a2aab | 2409 | |
17712eeb | 2410 | namespace = self._parse_smil_namespace(smil) |
a107193e | 2411 | |
da1c94ee | 2412 | fmts = self._parse_smil_formats( |
a107193e | 2413 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) |
da1c94ee F |
2414 | subs = self._parse_smil_subtitles( |
2415 | smil, namespace=namespace) | |
2416 | ||
2417 | return fmts, subs | |
2418 | ||
2419 | def _extract_smil_formats(self, *args, **kwargs): | |
2420 | fmts, subs = self._extract_smil_formats_and_subtitles(*args, **kwargs) | |
2421 | if subs: | |
b5ae35ee | 2422 | self._report_ignoring_subs('SMIL') |
da1c94ee | 2423 | return fmts |
a107193e S |
2424 | |
2425 | def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None): | |
2426 | smil = self._download_smil(smil_url, video_id, fatal=fatal) | |
2427 | if smil is False: | |
2428 | return {} | |
2429 | return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params) | |
2430 | ||
09f572fb | 2431 | def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None): |
a107193e S |
2432 | return self._download_xml( |
2433 | smil_url, video_id, 'Downloading SMIL file', | |
09f572fb | 2434 | 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source) |
a107193e S |
2435 | |
2436 | def _parse_smil(self, smil, smil_url, video_id, f4m_params=None): | |
17712eeb | 2437 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
2438 | |
2439 | formats = self._parse_smil_formats( | |
2440 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
2441 | subtitles = self._parse_smil_subtitles(smil, namespace=namespace) | |
2442 | ||
2443 | video_id = os.path.splitext(url_basename(smil_url))[0] | |
2444 | title = None | |
2445 | description = None | |
647eab45 | 2446 | upload_date = None |
a107193e S |
2447 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): |
2448 | name = meta.attrib.get('name') | |
2449 | content = meta.attrib.get('content') | |
2450 | if not name or not content: | |
2451 | continue | |
2452 | if not title and name == 'title': | |
2453 | title = content | |
2454 | elif not description and name in ('description', 'abstract'): | |
2455 | description = content | |
647eab45 S |
2456 | elif not upload_date and name == 'date': |
2457 | upload_date = unified_strdate(content) | |
a107193e | 2458 | |
1e5bcdec S |
2459 | thumbnails = [{ |
2460 | 'id': image.get('type'), | |
2461 | 'url': image.get('src'), | |
2462 | 'width': int_or_none(image.get('width')), | |
2463 | 'height': int_or_none(image.get('height')), | |
2464 | } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')] | |
2465 | ||
a107193e S |
2466 | return { |
2467 | 'id': video_id, | |
2468 | 'title': title or video_id, | |
2469 | 'description': description, | |
647eab45 | 2470 | 'upload_date': upload_date, |
1e5bcdec | 2471 | 'thumbnails': thumbnails, |
a107193e S |
2472 | 'formats': formats, |
2473 | 'subtitles': subtitles, | |
2474 | } | |
2475 | ||
17712eeb S |
2476 | def _parse_smil_namespace(self, smil): |
2477 | return self._search_regex( | |
2478 | r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None) | |
2479 | ||
f877c6ae | 2480 | def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None): |
a107193e S |
2481 | base = smil_url |
2482 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): | |
2483 | b = meta.get('base') or meta.get('httpBase') | |
2484 | if b: | |
2485 | base = b | |
2486 | break | |
e89a2aab S |
2487 | |
2488 | formats = [] | |
2489 | rtmp_count = 0 | |
a107193e | 2490 | http_count = 0 |
7f32e5dc | 2491 | m3u8_count = 0 |
9359f3d4 | 2492 | imgs_count = 0 |
a107193e | 2493 | |
9359f3d4 | 2494 | srcs = set() |
ad96b4c8 YCH |
2495 | media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace)) |
2496 | for medium in media: | |
2497 | src = medium.get('src') | |
81e1c4e2 | 2498 | if not src or src in srcs: |
a107193e | 2499 | continue |
9359f3d4 | 2500 | srcs.add(src) |
a107193e | 2501 | |
ad96b4c8 YCH |
2502 | bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000) |
2503 | filesize = int_or_none(medium.get('size') or medium.get('fileSize')) | |
2504 | width = int_or_none(medium.get('width')) | |
2505 | height = int_or_none(medium.get('height')) | |
2506 | proto = medium.get('proto') | |
2507 | ext = medium.get('ext') | |
a107193e | 2508 | src_ext = determine_ext(src) |
ad96b4c8 | 2509 | streamer = medium.get('streamer') or base |
a107193e S |
2510 | |
2511 | if proto == 'rtmp' or streamer.startswith('rtmp'): | |
2512 | rtmp_count += 1 | |
2513 | formats.append({ | |
2514 | 'url': streamer, | |
2515 | 'play_path': src, | |
2516 | 'ext': 'flv', | |
2517 | 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), | |
2518 | 'tbr': bitrate, | |
2519 | 'filesize': filesize, | |
2520 | 'width': width, | |
2521 | 'height': height, | |
2522 | }) | |
f877c6ae YCH |
2523 | if transform_rtmp_url: |
2524 | streamer, src = transform_rtmp_url(streamer, src) | |
2525 | formats[-1].update({ | |
2526 | 'url': streamer, | |
2527 | 'play_path': src, | |
2528 | }) | |
a107193e S |
2529 | continue |
2530 | ||
2531 | src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src) | |
c349456e | 2532 | src_url = src_url.strip() |
a107193e S |
2533 | |
2534 | if proto == 'm3u8' or src_ext == 'm3u8': | |
7f32e5dc | 2535 | m3u8_formats = self._extract_m3u8_formats( |
2536 | src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False) | |
2537 | if len(m3u8_formats) == 1: | |
2538 | m3u8_count += 1 | |
2539 | m3u8_formats[0].update({ | |
2540 | 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate), | |
2541 | 'tbr': bitrate, | |
2542 | 'width': width, | |
2543 | 'height': height, | |
2544 | }) | |
2545 | formats.extend(m3u8_formats) | |
bd21ead2 | 2546 | elif src_ext == 'f4m': |
a107193e S |
2547 | f4m_url = src_url |
2548 | if not f4m_params: | |
2549 | f4m_params = { | |
2550 | 'hdcore': '3.2.0', | |
2551 | 'plugin': 'flowplayer-3.2.0.1', | |
2552 | } | |
2553 | f4m_url += '&' if '?' in f4m_url else '?' | |
15707c7e | 2554 | f4m_url += compat_urllib_parse_urlencode(f4m_params) |
7e5edcfd | 2555 | formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False)) |
bd21ead2 RA |
2556 | elif src_ext == 'mpd': |
2557 | formats.extend(self._extract_mpd_formats( | |
2558 | src_url, video_id, mpd_id='dash', fatal=False)) | |
2559 | elif re.search(r'\.ism/[Mm]anifest', src_url): | |
2560 | formats.extend(self._extract_ism_formats( | |
2561 | src_url, video_id, ism_id='mss', fatal=False)) | |
2562 | elif src_url.startswith('http') and self._is_valid_url(src, video_id): | |
a107193e S |
2563 | http_count += 1 |
2564 | formats.append({ | |
2565 | 'url': src_url, | |
2566 | 'ext': ext or src_ext or 'flv', | |
2567 | 'format_id': 'http-%d' % (bitrate or http_count), | |
2568 | 'tbr': bitrate, | |
2569 | 'filesize': filesize, | |
2570 | 'width': width, | |
2571 | 'height': height, | |
2572 | }) | |
63757032 | 2573 | |
9359f3d4 F |
2574 | for medium in smil.findall(self._xpath_ns('.//imagestream', namespace)): |
2575 | src = medium.get('src') | |
2576 | if not src or src in srcs: | |
2577 | continue | |
2578 | srcs.add(src) | |
2579 | ||
2580 | imgs_count += 1 | |
2581 | formats.append({ | |
2582 | 'format_id': 'imagestream-%d' % (imgs_count), | |
2583 | 'url': src, | |
2584 | 'ext': mimetype2ext(medium.get('type')), | |
2585 | 'acodec': 'none', | |
2586 | 'vcodec': 'none', | |
2587 | 'width': int_or_none(medium.get('width')), | |
2588 | 'height': int_or_none(medium.get('height')), | |
2589 | 'format_note': 'SMIL storyboards', | |
2590 | }) | |
2591 | ||
e89a2aab S |
2592 | return formats |
2593 | ||
ce00af87 | 2594 | def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'): |
d413095f | 2595 | urls = [] |
a107193e S |
2596 | subtitles = {} |
2597 | for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))): | |
2598 | src = textstream.get('src') | |
d413095f | 2599 | if not src or src in urls: |
a107193e | 2600 | continue |
d413095f | 2601 | urls.append(src) |
df634be2 | 2602 | ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src) |
03bc7237 | 2603 | lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang |
a107193e S |
2604 | subtitles.setdefault(lang, []).append({ |
2605 | 'url': src, | |
2606 | 'ext': ext, | |
2607 | }) | |
2608 | return subtitles | |
63757032 | 2609 | |
47a5cb77 | 2610 | def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True): |
942acef5 | 2611 | xspf = self._download_xml( |
47a5cb77 | 2612 | xspf_url, playlist_id, 'Downloading xpsf playlist', |
942acef5 S |
2613 | 'Unable to download xspf manifest', fatal=fatal) |
2614 | if xspf is False: | |
2615 | return [] | |
47a5cb77 S |
2616 | return self._parse_xspf( |
2617 | xspf, playlist_id, xspf_url=xspf_url, | |
2618 | xspf_base_url=base_url(xspf_url)) | |
8d6765cf | 2619 | |
47a5cb77 | 2620 | def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None): |
8d6765cf S |
2621 | NS_MAP = { |
2622 | 'xspf': 'http://xspf.org/ns/0/', | |
2623 | 's1': 'http://static.streamone.nl/player/ns/0', | |
2624 | } | |
2625 | ||
2626 | entries = [] | |
47a5cb77 | 2627 | for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)): |
8d6765cf | 2628 | title = xpath_text( |
98044462 | 2629 | track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id) |
8d6765cf S |
2630 | description = xpath_text( |
2631 | track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description') | |
2632 | thumbnail = xpath_text( | |
2633 | track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail') | |
2634 | duration = float_or_none( | |
2635 | xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000) | |
2636 | ||
47a5cb77 S |
2637 | formats = [] |
2638 | for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)): | |
2639 | format_url = urljoin(xspf_base_url, location.text) | |
2640 | if not format_url: | |
2641 | continue | |
2642 | formats.append({ | |
2643 | 'url': format_url, | |
2644 | 'manifest_url': xspf_url, | |
2645 | 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)), | |
2646 | 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))), | |
2647 | 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))), | |
2648 | }) | |
8d6765cf S |
2649 | self._sort_formats(formats) |
2650 | ||
2651 | entries.append({ | |
2652 | 'id': playlist_id, | |
2653 | 'title': title, | |
2654 | 'description': description, | |
2655 | 'thumbnail': thumbnail, | |
2656 | 'duration': duration, | |
2657 | 'formats': formats, | |
2658 | }) | |
2659 | return entries | |
2660 | ||
171e59ed F |
2661 | def _extract_mpd_formats(self, *args, **kwargs): |
2662 | fmts, subs = self._extract_mpd_formats_and_subtitles(*args, **kwargs) | |
2663 | if subs: | |
b5ae35ee | 2664 | self._report_ignoring_subs('DASH') |
171e59ed F |
2665 | return fmts |
2666 | ||
2667 | def _extract_mpd_formats_and_subtitles( | |
2668 | self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, | |
2669 | fatal=True, data=None, headers={}, query={}): | |
47a5cb77 | 2670 | res = self._download_xml_handle( |
1bac3455 | 2671 | mpd_url, video_id, |
37a3bb66 | 2672 | note='Downloading MPD manifest' if note is None else note, |
2673 | errnote='Failed to download MPD manifest' if errnote is None else errnote, | |
7360c06f | 2674 | fatal=fatal, data=data, headers=headers, query=query) |
1bac3455 | 2675 | if res is False: |
171e59ed | 2676 | return [], {} |
47a5cb77 | 2677 | mpd_doc, urlh = res |
c25720ef | 2678 | if mpd_doc is None: |
171e59ed | 2679 | return [], {} |
02dc0a36 | 2680 | mpd_base_url = base_url(urlh.geturl()) |
1bac3455 | 2681 | |
171e59ed | 2682 | return self._parse_mpd_formats_and_subtitles( |
545cc85d | 2683 | mpd_doc, mpd_id, mpd_base_url, mpd_url) |
2d2fa82d | 2684 | |
171e59ed F |
2685 | def _parse_mpd_formats(self, *args, **kwargs): |
2686 | fmts, subs = self._parse_mpd_formats_and_subtitles(*args, **kwargs) | |
2687 | if subs: | |
b5ae35ee | 2688 | self._report_ignoring_subs('DASH') |
171e59ed F |
2689 | return fmts |
2690 | ||
2691 | def _parse_mpd_formats_and_subtitles( | |
2692 | self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None): | |
f0948348 S |
2693 | """ |
2694 | Parse formats from MPD manifest. | |
2695 | References: | |
2696 | 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E), | |
2697 | http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip | |
2698 | 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP | |
2699 | """ | |
a06916d9 | 2700 | if not self.get_param('dynamic_mpd', True): |
78895bd3 | 2701 | if mpd_doc.get('type') == 'dynamic': |
171e59ed | 2702 | return [], {} |
2d2fa82d | 2703 | |
91cb6b50 | 2704 | namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None) |
f14be228 | 2705 | |
2706 | def _add_ns(path): | |
2707 | return self._xpath_ns(path, namespace) | |
2708 | ||
675d0016 | 2709 | def is_drm_protected(element): |
2710 | return element.find(_add_ns('ContentProtection')) is not None | |
2711 | ||
1bac3455 | 2712 | def extract_multisegment_info(element, ms_parent_info): |
2713 | ms_info = ms_parent_info.copy() | |
b4c1d6e8 S |
2714 | |
2715 | # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some | |
2716 | # common attributes and elements. We will only extract relevant | |
2717 | # for us. | |
2718 | def extract_common(source): | |
2719 | segment_timeline = source.find(_add_ns('SegmentTimeline')) | |
2720 | if segment_timeline is not None: | |
2721 | s_e = segment_timeline.findall(_add_ns('S')) | |
2722 | if s_e: | |
2723 | ms_info['total_number'] = 0 | |
2724 | ms_info['s'] = [] | |
2725 | for s in s_e: | |
2726 | r = int(s.get('r', 0)) | |
2727 | ms_info['total_number'] += 1 + r | |
2728 | ms_info['s'].append({ | |
2729 | 't': int(s.get('t', 0)), | |
2730 | # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60]) | |
2731 | 'd': int(s.attrib['d']), | |
2732 | 'r': r, | |
2733 | }) | |
2734 | start_number = source.get('startNumber') | |
2735 | if start_number: | |
2736 | ms_info['start_number'] = int(start_number) | |
2737 | timescale = source.get('timescale') | |
2738 | if timescale: | |
2739 | ms_info['timescale'] = int(timescale) | |
2740 | segment_duration = source.get('duration') | |
2741 | if segment_duration: | |
48504785 | 2742 | ms_info['segment_duration'] = float(segment_duration) |
b4c1d6e8 S |
2743 | |
2744 | def extract_Initialization(source): | |
2745 | initialization = source.find(_add_ns('Initialization')) | |
2746 | if initialization is not None: | |
2747 | ms_info['initialization_url'] = initialization.attrib['sourceURL'] | |
2748 | ||
f14be228 | 2749 | segment_list = element.find(_add_ns('SegmentList')) |
1bac3455 | 2750 | if segment_list is not None: |
b4c1d6e8 S |
2751 | extract_common(segment_list) |
2752 | extract_Initialization(segment_list) | |
f14be228 | 2753 | segment_urls_e = segment_list.findall(_add_ns('SegmentURL')) |
1bac3455 | 2754 | if segment_urls_e: |
2755 | ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e] | |
1bac3455 | 2756 | else: |
f14be228 | 2757 | segment_template = element.find(_add_ns('SegmentTemplate')) |
1bac3455 | 2758 | if segment_template is not None: |
b4c1d6e8 | 2759 | extract_common(segment_template) |
e228616c S |
2760 | media = segment_template.get('media') |
2761 | if media: | |
2762 | ms_info['media'] = media | |
1bac3455 | 2763 | initialization = segment_template.get('initialization') |
2764 | if initialization: | |
e228616c | 2765 | ms_info['initialization'] = initialization |
1bac3455 | 2766 | else: |
b4c1d6e8 | 2767 | extract_Initialization(segment_template) |
1bac3455 | 2768 | return ms_info |
b323e170 | 2769 | |
1bac3455 | 2770 | mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration')) |
6251555f | 2771 | formats, subtitles = [], {} |
234416e4 | 2772 | stream_numbers = collections.defaultdict(int) |
f14be228 | 2773 | for period in mpd_doc.findall(_add_ns('Period')): |
1bac3455 | 2774 | period_duration = parse_duration(period.get('duration')) or mpd_duration |
2775 | period_ms_info = extract_multisegment_info(period, { | |
2776 | 'start_number': 1, | |
2777 | 'timescale': 1, | |
2778 | }) | |
f14be228 | 2779 | for adaptation_set in period.findall(_add_ns('AdaptationSet')): |
1bac3455 | 2780 | adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info) |
f14be228 | 2781 | for representation in adaptation_set.findall(_add_ns('Representation')): |
1bac3455 | 2782 | representation_attrib = adaptation_set.attrib.copy() |
2783 | representation_attrib.update(representation.attrib) | |
f0948348 | 2784 | # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory |
a6c8b759 | 2785 | mime_type = representation_attrib['mimeType'] |
171e59ed F |
2786 | content_type = representation_attrib.get('contentType', mime_type.split('/')[0]) |
2787 | ||
4afa3ec4 | 2788 | codecs = parse_codecs(representation_attrib.get('codecs', '')) |
be2fc5b2 | 2789 | if content_type not in ('video', 'audio', 'text'): |
2790 | if mime_type == 'image/jpeg': | |
a8731fcc | 2791 | content_type = mime_type |
4afa3ec4 F |
2792 | elif codecs['vcodec'] != 'none': |
2793 | content_type = 'video' | |
2794 | elif codecs['acodec'] != 'none': | |
2795 | content_type = 'audio' | |
2796 | elif codecs.get('tcodec', 'none') != 'none': | |
be2fc5b2 | 2797 | content_type = 'text' |
6993f78d | 2798 | elif mimetype2ext(mime_type) in ('tt', 'dfxp', 'ttml', 'xml', 'json'): |
2799 | content_type = 'text' | |
cdb19aa4 | 2800 | else: |
be2fc5b2 | 2801 | self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) |
2802 | continue | |
2803 | ||
2804 | base_url = '' | |
2805 | for element in (representation, adaptation_set, period, mpd_doc): | |
2806 | base_url_e = element.find(_add_ns('BaseURL')) | |
2807 | if base_url_e is not None: | |
2808 | base_url = base_url_e.text + base_url | |
2809 | if re.match(r'^https?://', base_url): | |
2810 | break | |
f9cc0161 D |
2811 | if mpd_base_url and base_url.startswith('/'): |
2812 | base_url = compat_urlparse.urljoin(mpd_base_url, base_url) | |
2813 | elif mpd_base_url and not re.match(r'^https?://', base_url): | |
2814 | if not mpd_base_url.endswith('/'): | |
be2fc5b2 | 2815 | mpd_base_url += '/' |
2816 | base_url = mpd_base_url + base_url | |
2817 | representation_id = representation_attrib.get('id') | |
2818 | lang = representation_attrib.get('lang') | |
2819 | url_el = representation.find(_add_ns('BaseURL')) | |
2820 | filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None) | |
2821 | bandwidth = int_or_none(representation_attrib.get('bandwidth')) | |
2822 | if representation_id is not None: | |
2823 | format_id = representation_id | |
2824 | else: | |
2825 | format_id = content_type | |
2826 | if mpd_id: | |
2827 | format_id = mpd_id + '-' + format_id | |
2828 | if content_type in ('video', 'audio'): | |
2829 | f = { | |
2830 | 'format_id': format_id, | |
2831 | 'manifest_url': mpd_url, | |
2832 | 'ext': mimetype2ext(mime_type), | |
2833 | 'width': int_or_none(representation_attrib.get('width')), | |
2834 | 'height': int_or_none(representation_attrib.get('height')), | |
2835 | 'tbr': float_or_none(bandwidth, 1000), | |
2836 | 'asr': int_or_none(representation_attrib.get('audioSamplingRate')), | |
2837 | 'fps': int_or_none(representation_attrib.get('frameRate')), | |
2838 | 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None, | |
2839 | 'format_note': 'DASH %s' % content_type, | |
2840 | 'filesize': filesize, | |
2841 | 'container': mimetype2ext(mime_type) + '_dash', | |
4afa3ec4 | 2842 | **codecs |
be2fc5b2 | 2843 | } |
be2fc5b2 | 2844 | elif content_type == 'text': |
2845 | f = { | |
2846 | 'ext': mimetype2ext(mime_type), | |
2847 | 'manifest_url': mpd_url, | |
2848 | 'filesize': filesize, | |
2849 | } | |
2850 | elif content_type == 'image/jpeg': | |
2851 | # See test case in VikiIE | |
2852 | # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1 | |
2853 | f = { | |
2854 | 'format_id': format_id, | |
2855 | 'ext': 'mhtml', | |
2856 | 'manifest_url': mpd_url, | |
2857 | 'format_note': 'DASH storyboards (jpeg)', | |
2858 | 'acodec': 'none', | |
2859 | 'vcodec': 'none', | |
2860 | } | |
88acdbc2 | 2861 | if is_drm_protected(adaptation_set) or is_drm_protected(representation): |
2862 | f['has_drm'] = True | |
be2fc5b2 | 2863 | representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info) |
2864 | ||
2865 | def prepare_template(template_name, identifiers): | |
2866 | tmpl = representation_ms_info[template_name] | |
2867 | # First of, % characters outside $...$ templates | |
2868 | # must be escaped by doubling for proper processing | |
2869 | # by % operator string formatting used further (see | |
2870 | # https://github.com/ytdl-org/youtube-dl/issues/16867). | |
2871 | t = '' | |
2872 | in_template = False | |
2873 | for c in tmpl: | |
2874 | t += c | |
2875 | if c == '$': | |
2876 | in_template = not in_template | |
2877 | elif c == '%' and not in_template: | |
eca1f0d1 | 2878 | t += c |
be2fc5b2 | 2879 | # Next, $...$ templates are translated to their |
2880 | # %(...) counterparts to be used with % operator | |
2881 | if representation_id is not None: | |
2882 | t = t.replace('$RepresentationID$', representation_id) | |
2883 | t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t) | |
2884 | t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t) | |
2885 | t.replace('$$', '$') | |
2886 | return t | |
2887 | ||
2888 | # @initialization is a regular template like @media one | |
2889 | # so it should be handled just the same way (see | |
2890 | # https://github.com/ytdl-org/youtube-dl/issues/11605) | |
2891 | if 'initialization' in representation_ms_info: | |
2892 | initialization_template = prepare_template( | |
2893 | 'initialization', | |
2894 | # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and | |
2895 | # $Time$ shall not be included for @initialization thus | |
2896 | # only $Bandwidth$ remains | |
2897 | ('Bandwidth', )) | |
2898 | representation_ms_info['initialization_url'] = initialization_template % { | |
2899 | 'Bandwidth': bandwidth, | |
2900 | } | |
2901 | ||
2902 | def location_key(location): | |
2903 | return 'url' if re.match(r'^https?://', location) else 'path' | |
2904 | ||
2905 | if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info: | |
2906 | ||
2907 | media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time')) | |
2908 | media_location_key = location_key(media_template) | |
2909 | ||
2910 | # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$ | |
2911 | # can't be used at the same time | |
2912 | if '%(Number' in media_template and 's' not in representation_ms_info: | |
2913 | segment_duration = None | |
2914 | if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info: | |
2915 | segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale']) | |
ffa89477 | 2916 | representation_ms_info['total_number'] = int(math.ceil( |
2917 | float_or_none(period_duration, segment_duration, default=0))) | |
be2fc5b2 | 2918 | representation_ms_info['fragments'] = [{ |
2919 | media_location_key: media_template % { | |
2920 | 'Number': segment_number, | |
2921 | 'Bandwidth': bandwidth, | |
2922 | }, | |
2923 | 'duration': segment_duration, | |
2924 | } for segment_number in range( | |
2925 | representation_ms_info['start_number'], | |
2926 | representation_ms_info['total_number'] + representation_ms_info['start_number'])] | |
2927 | else: | |
2928 | # $Number*$ or $Time$ in media template with S list available | |
2929 | # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg | |
2930 | # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411 | |
2931 | representation_ms_info['fragments'] = [] | |
2932 | segment_time = 0 | |
2933 | segment_d = None | |
2934 | segment_number = representation_ms_info['start_number'] | |
2935 | ||
2936 | def add_segment_url(): | |
2937 | segment_url = media_template % { | |
2938 | 'Time': segment_time, | |
2939 | 'Bandwidth': bandwidth, | |
2940 | 'Number': segment_number, | |
2941 | } | |
2942 | representation_ms_info['fragments'].append({ | |
2943 | media_location_key: segment_url, | |
2944 | 'duration': float_or_none(segment_d, representation_ms_info['timescale']), | |
2945 | }) | |
2946 | ||
2947 | for num, s in enumerate(representation_ms_info['s']): | |
2948 | segment_time = s.get('t') or segment_time | |
2949 | segment_d = s['d'] | |
2950 | add_segment_url() | |
2951 | segment_number += 1 | |
2952 | for r in range(s.get('r', 0)): | |
2953 | segment_time += segment_d | |
f0948348 | 2954 | add_segment_url() |
b4c1d6e8 | 2955 | segment_number += 1 |
be2fc5b2 | 2956 | segment_time += segment_d |
2957 | elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info: | |
2958 | # No media template | |
2959 | # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI | |
2960 | # or any YouTube dashsegments video | |
2961 | fragments = [] | |
2962 | segment_index = 0 | |
2963 | timescale = representation_ms_info['timescale'] | |
2964 | for s in representation_ms_info['s']: | |
2965 | duration = float_or_none(s['d'], timescale) | |
2966 | for r in range(s.get('r', 0) + 1): | |
2967 | segment_uri = representation_ms_info['segment_urls'][segment_index] | |
2968 | fragments.append({ | |
2969 | location_key(segment_uri): segment_uri, | |
2970 | 'duration': duration, | |
2971 | }) | |
2972 | segment_index += 1 | |
2973 | representation_ms_info['fragments'] = fragments | |
2974 | elif 'segment_urls' in representation_ms_info: | |
2975 | # Segment URLs with no SegmentTimeline | |
2976 | # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091 | |
2977 | # https://github.com/ytdl-org/youtube-dl/pull/14844 | |
2978 | fragments = [] | |
2979 | segment_duration = float_or_none( | |
2980 | representation_ms_info['segment_duration'], | |
2981 | representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None | |
2982 | for segment_url in representation_ms_info['segment_urls']: | |
2983 | fragment = { | |
2984 | location_key(segment_url): segment_url, | |
2985 | } | |
2986 | if segment_duration: | |
2987 | fragment['duration'] = segment_duration | |
2988 | fragments.append(fragment) | |
2989 | representation_ms_info['fragments'] = fragments | |
2990 | # If there is a fragments key available then we correctly recognized fragmented media. | |
2991 | # Otherwise we will assume unfragmented media with direct access. Technically, such | |
2992 | # assumption is not necessarily correct since we may simply have no support for | |
2993 | # some forms of fragmented media renditions yet, but for now we'll use this fallback. | |
2994 | if 'fragments' in representation_ms_info: | |
2995 | f.update({ | |
2996 | # NB: mpd_url may be empty when MPD manifest is parsed from a string | |
2997 | 'url': mpd_url or base_url, | |
2998 | 'fragment_base_url': base_url, | |
2999 | 'fragments': [], | |
3000 | 'protocol': 'http_dash_segments' if mime_type != 'image/jpeg' else 'mhtml', | |
3001 | }) | |
3002 | if 'initialization_url' in representation_ms_info: | |
3003 | initialization_url = representation_ms_info['initialization_url'] | |
3004 | if not f.get('url'): | |
3005 | f['url'] = initialization_url | |
3006 | f['fragments'].append({location_key(initialization_url): initialization_url}) | |
3007 | f['fragments'].extend(representation_ms_info['fragments']) | |
ffa89477 | 3008 | if not period_duration: |
3009 | period_duration = try_get( | |
3010 | representation_ms_info, | |
3011 | lambda r: sum(frag['duration'] for frag in r['fragments']), float) | |
17b598d3 | 3012 | else: |
be2fc5b2 | 3013 | # Assuming direct URL to unfragmented media. |
3014 | f['url'] = base_url | |
234416e4 | 3015 | if content_type in ('video', 'audio', 'image/jpeg'): |
3016 | f['manifest_stream_number'] = stream_numbers[f['url']] | |
3017 | stream_numbers[f['url']] += 1 | |
be2fc5b2 | 3018 | formats.append(f) |
3019 | elif content_type == 'text': | |
3020 | subtitles.setdefault(lang or 'und', []).append(f) | |
3021 | ||
171e59ed | 3022 | return formats, subtitles |
17b598d3 | 3023 | |
fd76a142 F |
3024 | def _extract_ism_formats(self, *args, **kwargs): |
3025 | fmts, subs = self._extract_ism_formats_and_subtitles(*args, **kwargs) | |
3026 | if subs: | |
b5ae35ee | 3027 | self._report_ignoring_subs('ISM') |
fd76a142 F |
3028 | return fmts |
3029 | ||
3030 | 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 | 3031 | res = self._download_xml_handle( |
b2758123 | 3032 | ism_url, video_id, |
37a3bb66 | 3033 | note='Downloading ISM manifest' if note is None else note, |
3034 | errnote='Failed to download ISM manifest' if errnote is None else errnote, | |
7360c06f | 3035 | fatal=fatal, data=data, headers=headers, query=query) |
b2758123 | 3036 | if res is False: |
fd76a142 | 3037 | return [], {} |
47a5cb77 | 3038 | ism_doc, urlh = res |
13b08034 | 3039 | if ism_doc is None: |
fd76a142 | 3040 | return [], {} |
b2758123 | 3041 | |
fd76a142 | 3042 | return self._parse_ism_formats_and_subtitles(ism_doc, urlh.geturl(), ism_id) |
b2758123 | 3043 | |
fd76a142 | 3044 | def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None): |
76d5a363 S |
3045 | """ |
3046 | Parse formats from ISM manifest. | |
3047 | References: | |
3048 | 1. [MS-SSTR]: Smooth Streaming Protocol, | |
3049 | https://msdn.microsoft.com/en-us/library/ff469518.aspx | |
3050 | """ | |
06869367 | 3051 | if ism_doc.get('IsLive') == 'TRUE': |
fd76a142 | 3052 | return [], {} |
b2758123 | 3053 | |
b2758123 RA |
3054 | duration = int(ism_doc.attrib['Duration']) |
3055 | timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000 | |
3056 | ||
3057 | formats = [] | |
fd76a142 | 3058 | subtitles = {} |
b2758123 RA |
3059 | for stream in ism_doc.findall('StreamIndex'): |
3060 | stream_type = stream.get('Type') | |
fd76a142 | 3061 | if stream_type not in ('video', 'audio', 'text'): |
b2758123 RA |
3062 | continue |
3063 | url_pattern = stream.attrib['Url'] | |
3064 | stream_timescale = int_or_none(stream.get('TimeScale')) or timescale | |
3065 | stream_name = stream.get('Name') | |
fd76a142 | 3066 | stream_language = stream.get('Language', 'und') |
b2758123 | 3067 | for track in stream.findall('QualityLevel'): |
e2efe599 | 3068 | fourcc = track.get('FourCC') or ('AACL' if track.get('AudioTag') == '255' else None) |
b2758123 | 3069 | # TODO: add support for WVC1 and WMAP |
66a1b864 | 3070 | if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML'): |
b2758123 RA |
3071 | self.report_warning('%s is not a supported codec' % fourcc) |
3072 | continue | |
3073 | tbr = int(track.attrib['Bitrate']) // 1000 | |
76d5a363 S |
3074 | # [1] does not mention Width and Height attributes. However, |
3075 | # they're often present while MaxWidth and MaxHeight are | |
3076 | # missing, so should be used as fallbacks | |
3077 | width = int_or_none(track.get('MaxWidth') or track.get('Width')) | |
3078 | height = int_or_none(track.get('MaxHeight') or track.get('Height')) | |
b2758123 RA |
3079 | sampling_rate = int_or_none(track.get('SamplingRate')) |
3080 | ||
3081 | track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern) | |
3082 | track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern) | |
3083 | ||
3084 | fragments = [] | |
3085 | fragment_ctx = { | |
3086 | 'time': 0, | |
3087 | } | |
3088 | stream_fragments = stream.findall('c') | |
3089 | for stream_fragment_index, stream_fragment in enumerate(stream_fragments): | |
3090 | fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time'] | |
3091 | fragment_repeat = int_or_none(stream_fragment.get('r')) or 1 | |
3092 | fragment_ctx['duration'] = int_or_none(stream_fragment.get('d')) | |
3093 | if not fragment_ctx['duration']: | |
3094 | try: | |
3095 | next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t']) | |
3096 | except IndexError: | |
3097 | next_fragment_time = duration | |
1616f9b4 | 3098 | fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat |
b2758123 RA |
3099 | for _ in range(fragment_repeat): |
3100 | fragments.append({ | |
1616f9b4 | 3101 | 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern), |
b2758123 RA |
3102 | 'duration': fragment_ctx['duration'] / stream_timescale, |
3103 | }) | |
3104 | fragment_ctx['time'] += fragment_ctx['duration'] | |
3105 | ||
fd76a142 F |
3106 | if stream_type == 'text': |
3107 | subtitles.setdefault(stream_language, []).append({ | |
3108 | 'ext': 'ismt', | |
3109 | 'protocol': 'ism', | |
3110 | 'url': ism_url, | |
3111 | 'manifest_url': ism_url, | |
3112 | 'fragments': fragments, | |
3113 | '_download_params': { | |
3114 | 'stream_type': stream_type, | |
3115 | 'duration': duration, | |
3116 | 'timescale': stream_timescale, | |
3117 | 'fourcc': fourcc, | |
3118 | 'language': stream_language, | |
3119 | 'codec_private_data': track.get('CodecPrivateData'), | |
3120 | } | |
3121 | }) | |
3122 | elif stream_type in ('video', 'audio'): | |
3123 | formats.append({ | |
34921b43 | 3124 | 'format_id': join_nonempty(ism_id, stream_name, tbr), |
fd76a142 F |
3125 | 'url': ism_url, |
3126 | 'manifest_url': ism_url, | |
3127 | 'ext': 'ismv' if stream_type == 'video' else 'isma', | |
3128 | 'width': width, | |
3129 | 'height': height, | |
3130 | 'tbr': tbr, | |
3131 | 'asr': sampling_rate, | |
3132 | 'vcodec': 'none' if stream_type == 'audio' else fourcc, | |
3133 | 'acodec': 'none' if stream_type == 'video' else fourcc, | |
3134 | 'protocol': 'ism', | |
3135 | 'fragments': fragments, | |
88acdbc2 | 3136 | 'has_drm': ism_doc.find('Protection') is not None, |
fd76a142 F |
3137 | '_download_params': { |
3138 | 'stream_type': stream_type, | |
3139 | 'duration': duration, | |
3140 | 'timescale': stream_timescale, | |
3141 | 'width': width or 0, | |
3142 | 'height': height or 0, | |
3143 | 'fourcc': fourcc, | |
3144 | 'language': stream_language, | |
3145 | 'codec_private_data': track.get('CodecPrivateData'), | |
3146 | 'sampling_rate': sampling_rate, | |
3147 | 'channels': int_or_none(track.get('Channels', 2)), | |
3148 | 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)), | |
3149 | 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)), | |
3150 | }, | |
3151 | }) | |
3152 | return formats, subtitles | |
b2758123 | 3153 | |
079a7cfc | 3154 | 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 |
3155 | def absolute_url(item_url): |
3156 | return urljoin(base_url, item_url) | |
59bbe491 | 3157 | |
3158 | def parse_content_type(content_type): | |
3159 | if not content_type: | |
3160 | return {} | |
3161 | ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type) | |
3162 | if ctr: | |
3163 | mimetype, codecs = ctr.groups() | |
3164 | f = parse_codecs(codecs) | |
3165 | f['ext'] = mimetype2ext(mimetype) | |
3166 | return f | |
3167 | return {} | |
3168 | ||
868f79db | 3169 | def _media_formats(src, cur_media_type, type_info={}): |
520251c0 | 3170 | full_url = absolute_url(src) |
82889d4a | 3171 | ext = type_info.get('ext') or determine_ext(full_url) |
87a449c1 | 3172 | if ext == 'm3u8': |
520251c0 YCH |
3173 | is_plain_url = False |
3174 | formats = self._extract_m3u8_formats( | |
ad120ae1 | 3175 | full_url, video_id, ext='mp4', |
eeb0a956 | 3176 | entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id, |
f983b875 | 3177 | preference=preference, quality=quality, fatal=False) |
87a449c1 S |
3178 | elif ext == 'mpd': |
3179 | is_plain_url = False | |
3180 | formats = self._extract_mpd_formats( | |
b359e977 | 3181 | full_url, video_id, mpd_id=mpd_id, fatal=False) |
520251c0 YCH |
3182 | else: |
3183 | is_plain_url = True | |
3184 | formats = [{ | |
3185 | 'url': full_url, | |
3186 | 'vcodec': 'none' if cur_media_type == 'audio' else None, | |
3187 | }] | |
3188 | return is_plain_url, formats | |
3189 | ||
59bbe491 | 3190 | entries = [] |
4328ddf8 S |
3191 | # amp-video and amp-audio are very similar to their HTML5 counterparts |
3192 | # so we wll include them right here (see | |
3193 | # https://www.ampproject.org/docs/reference/components/amp-video) | |
29f7c58a | 3194 | # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/ |
3195 | _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)' | |
3196 | media_tags = [(media_tag, media_tag_name, media_type, '') | |
3197 | for media_tag, media_tag_name, media_type | |
3198 | in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)] | |
2aec7256 S |
3199 | media_tags.extend(re.findall( |
3200 | # We only allow video|audio followed by a whitespace or '>'. | |
3201 | # Allowing more characters may end up in significant slow down (see | |
067aa17e | 3202 | # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL: |
2aec7256 | 3203 | # http://www.porntrex.com/maps/videositemap.xml). |
29f7c58a | 3204 | r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage)) |
3205 | for media_tag, _, media_type, media_content in media_tags: | |
59bbe491 | 3206 | media_info = { |
3207 | 'formats': [], | |
3208 | 'subtitles': {}, | |
3209 | } | |
3210 | media_attributes = extract_attributes(media_tag) | |
f856816b | 3211 | src = strip_or_none(media_attributes.get('src')) |
59bbe491 | 3212 | if src: |
dedb1770 | 3213 | _, formats = _media_formats(src, media_type) |
520251c0 | 3214 | media_info['formats'].extend(formats) |
6780154e | 3215 | media_info['thumbnail'] = absolute_url(media_attributes.get('poster')) |
59bbe491 | 3216 | if media_content: |
3217 | for source_tag in re.findall(r'<source[^>]+>', media_content): | |
d493f15c S |
3218 | s_attr = extract_attributes(source_tag) |
3219 | # data-video-src and data-src are non standard but seen | |
3220 | # several times in the wild | |
f856816b | 3221 | src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src'))) |
59bbe491 | 3222 | if not src: |
3223 | continue | |
d493f15c | 3224 | f = parse_content_type(s_attr.get('type')) |
868f79db | 3225 | is_plain_url, formats = _media_formats(src, media_type, f) |
520251c0 | 3226 | if is_plain_url: |
d493f15c S |
3227 | # width, height, res, label and title attributes are |
3228 | # all not standard but seen several times in the wild | |
3229 | labels = [ | |
3230 | s_attr.get(lbl) | |
3231 | for lbl in ('label', 'title') | |
3232 | if str_or_none(s_attr.get(lbl)) | |
3233 | ] | |
3234 | width = int_or_none(s_attr.get('width')) | |
3089bc74 S |
3235 | height = (int_or_none(s_attr.get('height')) |
3236 | or int_or_none(s_attr.get('res'))) | |
d493f15c S |
3237 | if not width or not height: |
3238 | for lbl in labels: | |
3239 | resolution = parse_resolution(lbl) | |
3240 | if not resolution: | |
3241 | continue | |
3242 | width = width or resolution.get('width') | |
3243 | height = height or resolution.get('height') | |
3244 | for lbl in labels: | |
3245 | tbr = parse_bitrate(lbl) | |
3246 | if tbr: | |
3247 | break | |
3248 | else: | |
3249 | tbr = None | |
1ed45499 | 3250 | f.update({ |
d493f15c S |
3251 | 'width': width, |
3252 | 'height': height, | |
3253 | 'tbr': tbr, | |
3254 | 'format_id': s_attr.get('label') or s_attr.get('title'), | |
1ed45499 | 3255 | }) |
520251c0 YCH |
3256 | f.update(formats[0]) |
3257 | media_info['formats'].append(f) | |
3258 | else: | |
3259 | media_info['formats'].extend(formats) | |
59bbe491 | 3260 | for track_tag in re.findall(r'<track[^>]+>', media_content): |
3261 | track_attributes = extract_attributes(track_tag) | |
3262 | kind = track_attributes.get('kind') | |
5968d7d2 | 3263 | if not kind or kind in ('subtitles', 'captions'): |
f856816b | 3264 | src = strip_or_none(track_attributes.get('src')) |
59bbe491 | 3265 | if not src: |
3266 | continue | |
3267 | lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label') | |
3268 | media_info['subtitles'].setdefault(lang, []).append({ | |
3269 | 'url': absolute_url(src), | |
3270 | }) | |
5e8e2fa5 S |
3271 | for f in media_info['formats']: |
3272 | f.setdefault('http_headers', {})['Referer'] = base_url | |
5968d7d2 | 3273 | if media_info['formats'] or media_info['subtitles']: |
59bbe491 | 3274 | entries.append(media_info) |
3275 | return entries | |
3276 | ||
f6a1d69a F |
3277 | def _extract_akamai_formats(self, *args, **kwargs): |
3278 | fmts, subs = self._extract_akamai_formats_and_subtitles(*args, **kwargs) | |
3279 | if subs: | |
b5ae35ee | 3280 | self._report_ignoring_subs('akamai') |
f6a1d69a F |
3281 | return fmts |
3282 | ||
3283 | def _extract_akamai_formats_and_subtitles(self, manifest_url, video_id, hosts={}): | |
29f7c58a | 3284 | signed = 'hdnea=' in manifest_url |
3285 | if not signed: | |
3286 | # https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html | |
3287 | manifest_url = re.sub( | |
3288 | r'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?', | |
3289 | '', manifest_url).strip('?') | |
3290 | ||
c7c43a93 | 3291 | formats = [] |
f6a1d69a | 3292 | subtitles = {} |
70c5802b | 3293 | |
e71a4509 | 3294 | hdcore_sign = 'hdcore=3.7.0' |
ff6f9a67 | 3295 | f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') |
c4251b9a RA |
3296 | hds_host = hosts.get('hds') |
3297 | if hds_host: | |
3298 | f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url) | |
e71a4509 RA |
3299 | if 'hdcore=' not in f4m_url: |
3300 | f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign | |
3301 | f4m_formats = self._extract_f4m_formats( | |
3302 | f4m_url, video_id, f4m_id='hds', fatal=False) | |
3303 | for entry in f4m_formats: | |
3304 | entry.update({'extra_param_to_segment_url': hdcore_sign}) | |
3305 | formats.extend(f4m_formats) | |
70c5802b | 3306 | |
c4251b9a RA |
3307 | m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8') |
3308 | hls_host = hosts.get('hls') | |
3309 | if hls_host: | |
3310 | m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url) | |
f6a1d69a | 3311 | m3u8_formats, m3u8_subtitles = self._extract_m3u8_formats_and_subtitles( |
c7c43a93 | 3312 | m3u8_url, video_id, 'mp4', 'm3u8_native', |
29f7c58a | 3313 | m3u8_id='hls', fatal=False) |
3314 | formats.extend(m3u8_formats) | |
f6a1d69a | 3315 | subtitles = self._merge_subtitles(subtitles, m3u8_subtitles) |
70c5802b | 3316 | |
3317 | http_host = hosts.get('http') | |
29f7c58a | 3318 | if http_host and m3u8_formats and not signed: |
3319 | REPL_REGEX = r'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+' | |
70c5802b | 3320 | qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',') |
3321 | qualities_length = len(qualities) | |
29f7c58a | 3322 | if len(m3u8_formats) in (qualities_length, qualities_length + 1): |
70c5802b | 3323 | i = 0 |
29f7c58a | 3324 | for f in m3u8_formats: |
3325 | if f['vcodec'] != 'none': | |
70c5802b | 3326 | for protocol in ('http', 'https'): |
3327 | http_f = f.copy() | |
3328 | del http_f['manifest_url'] | |
3329 | http_url = re.sub( | |
86e5f3ed | 3330 | REPL_REGEX, protocol + fr'://{http_host}/\g<1>{qualities[i]}\3', f['url']) |
70c5802b | 3331 | http_f.update({ |
3332 | 'format_id': http_f['format_id'].replace('hls-', protocol + '-'), | |
3333 | 'url': http_url, | |
3334 | 'protocol': protocol, | |
3335 | }) | |
29f7c58a | 3336 | formats.append(http_f) |
70c5802b | 3337 | i += 1 |
70c5802b | 3338 | |
f6a1d69a | 3339 | return formats, subtitles |
c7c43a93 | 3340 | |
6ad02195 | 3341 | def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]): |
044eeb14 | 3342 | query = compat_urlparse.urlparse(url).query |
6ad02195 | 3343 | url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url) |
240f2622 S |
3344 | mobj = re.search( |
3345 | r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url) | |
3346 | url_base = mobj.group('url') | |
3347 | http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base) | |
6ad02195 | 3348 | formats = [] |
044eeb14 S |
3349 | |
3350 | def manifest_url(manifest): | |
86e5f3ed | 3351 | m_url = f'{http_base_url}/{manifest}' |
044eeb14 S |
3352 | if query: |
3353 | m_url += '?%s' % query | |
3354 | return m_url | |
3355 | ||
6ad02195 RA |
3356 | if 'm3u8' not in skip_protocols: |
3357 | formats.extend(self._extract_m3u8_formats( | |
044eeb14 | 3358 | manifest_url('playlist.m3u8'), video_id, 'mp4', |
6ad02195 RA |
3359 | m3u8_entry_protocol, m3u8_id='hls', fatal=False)) |
3360 | if 'f4m' not in skip_protocols: | |
3361 | formats.extend(self._extract_f4m_formats( | |
044eeb14 | 3362 | manifest_url('manifest.f4m'), |
6ad02195 | 3363 | video_id, f4m_id='hds', fatal=False)) |
0384932e RA |
3364 | if 'dash' not in skip_protocols: |
3365 | formats.extend(self._extract_mpd_formats( | |
044eeb14 | 3366 | manifest_url('manifest.mpd'), |
0384932e | 3367 | video_id, mpd_id='dash', fatal=False)) |
6ad02195 | 3368 | if re.search(r'(?:/smil:|\.smil)', url_base): |
6ad02195 RA |
3369 | if 'smil' not in skip_protocols: |
3370 | rtmp_formats = self._extract_smil_formats( | |
044eeb14 | 3371 | manifest_url('jwplayer.smil'), |
6ad02195 RA |
3372 | video_id, fatal=False) |
3373 | for rtmp_format in rtmp_formats: | |
3374 | rtsp_format = rtmp_format.copy() | |
3375 | rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path']) | |
3376 | del rtsp_format['play_path'] | |
3377 | del rtsp_format['ext'] | |
3378 | rtsp_format.update({ | |
3379 | 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'), | |
3380 | 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'), | |
3381 | 'protocol': 'rtsp', | |
3382 | }) | |
3383 | formats.extend([rtmp_format, rtsp_format]) | |
3384 | else: | |
3385 | for protocol in ('rtmp', 'rtsp'): | |
3386 | if protocol not in skip_protocols: | |
3387 | formats.append({ | |
86e5f3ed | 3388 | 'url': f'{protocol}:{url_base}', |
6ad02195 RA |
3389 | 'format_id': protocol, |
3390 | 'protocol': protocol, | |
3391 | }) | |
3392 | return formats | |
3393 | ||
c73e330e | 3394 | def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json): |
a4a554a7 | 3395 | mobj = re.search( |
ac9c69ac | 3396 | r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)', |
a4a554a7 YCH |
3397 | webpage) |
3398 | if mobj: | |
c73e330e RU |
3399 | try: |
3400 | jwplayer_data = self._parse_json(mobj.group('options'), | |
3401 | video_id=video_id, | |
3402 | transform_source=transform_source) | |
3403 | except ExtractorError: | |
3404 | pass | |
3405 | else: | |
3406 | if isinstance(jwplayer_data, dict): | |
3407 | return jwplayer_data | |
a4a554a7 YCH |
3408 | |
3409 | def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs): | |
c73e330e RU |
3410 | jwplayer_data = self._find_jwplayer_data( |
3411 | webpage, video_id, transform_source=js_to_json) | |
a4a554a7 YCH |
3412 | return self._parse_jwplayer_data( |
3413 | jwplayer_data, video_id, *args, **kwargs) | |
3414 | ||
3415 | def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True, | |
3416 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
3417 | # JWPlayer backward compatibility: flattened playlists | |
3418 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96 | |
3419 | if 'playlist' not in jwplayer_data: | |
3420 | jwplayer_data = {'playlist': [jwplayer_data]} | |
3421 | ||
3422 | entries = [] | |
3423 | ||
3424 | # JWPlayer backward compatibility: single playlist item | |
3425 | # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10 | |
3426 | if not isinstance(jwplayer_data['playlist'], list): | |
3427 | jwplayer_data['playlist'] = [jwplayer_data['playlist']] | |
3428 | ||
3429 | for video_data in jwplayer_data['playlist']: | |
3430 | # JWPlayer backward compatibility: flattened sources | |
3431 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35 | |
3432 | if 'sources' not in video_data: | |
3433 | video_data['sources'] = [video_data] | |
3434 | ||
3435 | this_video_id = video_id or video_data['mediaid'] | |
3436 | ||
1a2192cb S |
3437 | formats = self._parse_jwplayer_formats( |
3438 | video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id, | |
3439 | mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url) | |
a4a554a7 YCH |
3440 | |
3441 | subtitles = {} | |
3442 | tracks = video_data.get('tracks') | |
3443 | if tracks and isinstance(tracks, list): | |
3444 | for track in tracks: | |
96a2daa1 S |
3445 | if not isinstance(track, dict): |
3446 | continue | |
f4b74272 S |
3447 | track_kind = track.get('kind') |
3448 | if not track_kind or not isinstance(track_kind, compat_str): | |
3449 | continue | |
3450 | if track_kind.lower() not in ('captions', 'subtitles'): | |
a4a554a7 YCH |
3451 | continue |
3452 | track_url = urljoin(base_url, track.get('file')) | |
3453 | if not track_url: | |
3454 | continue | |
3455 | subtitles.setdefault(track.get('label') or 'en', []).append({ | |
3456 | 'url': self._proto_relative_url(track_url) | |
3457 | }) | |
3458 | ||
50d808f5 | 3459 | entry = { |
a4a554a7 | 3460 | 'id': this_video_id, |
50d808f5 | 3461 | 'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')), |
f81dd65b | 3462 | 'description': clean_html(video_data.get('description')), |
6945b9e7 | 3463 | 'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))), |
a4a554a7 YCH |
3464 | 'timestamp': int_or_none(video_data.get('pubdate')), |
3465 | 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')), | |
3466 | 'subtitles': subtitles, | |
50d808f5 RA |
3467 | } |
3468 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32 | |
3469 | if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']): | |
3470 | entry.update({ | |
3471 | '_type': 'url_transparent', | |
3472 | 'url': formats[0]['url'], | |
3473 | }) | |
3474 | else: | |
3475 | self._sort_formats(formats) | |
3476 | entry['formats'] = formats | |
3477 | entries.append(entry) | |
a4a554a7 YCH |
3478 | if len(entries) == 1: |
3479 | return entries[0] | |
3480 | else: | |
3481 | return self.playlist_result(entries) | |
3482 | ||
ed0cf9b3 S |
3483 | def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None, |
3484 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
bf1b87cd | 3485 | urls = [] |
ed0cf9b3 | 3486 | formats = [] |
1a2192cb | 3487 | for source in jwplayer_sources_data: |
0a268c6e S |
3488 | if not isinstance(source, dict): |
3489 | continue | |
6945b9e7 RA |
3490 | source_url = urljoin( |
3491 | base_url, self._proto_relative_url(source.get('file'))) | |
3492 | if not source_url or source_url in urls: | |
bf1b87cd RA |
3493 | continue |
3494 | urls.append(source_url) | |
ed0cf9b3 S |
3495 | source_type = source.get('type') or '' |
3496 | ext = mimetype2ext(source_type) or determine_ext(source_url) | |
3497 | if source_type == 'hls' or ext == 'm3u8': | |
3498 | formats.extend(self._extract_m3u8_formats( | |
0236cd0d S |
3499 | source_url, video_id, 'mp4', entry_protocol='m3u8_native', |
3500 | m3u8_id=m3u8_id, fatal=False)) | |
0d9c48de | 3501 | elif source_type == 'dash' or ext == 'mpd': |
ed0cf9b3 S |
3502 | formats.extend(self._extract_mpd_formats( |
3503 | source_url, video_id, mpd_id=mpd_id, fatal=False)) | |
b51dc9db S |
3504 | elif ext == 'smil': |
3505 | formats.extend(self._extract_smil_formats( | |
3506 | source_url, video_id, fatal=False)) | |
ed0cf9b3 | 3507 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67 |
0236cd0d S |
3508 | elif source_type.startswith('audio') or ext in ( |
3509 | 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'): | |
ed0cf9b3 S |
3510 | formats.append({ |
3511 | 'url': source_url, | |
3512 | 'vcodec': 'none', | |
3513 | 'ext': ext, | |
3514 | }) | |
3515 | else: | |
3516 | height = int_or_none(source.get('height')) | |
3517 | if height is None: | |
3518 | # Often no height is provided but there is a label in | |
0236cd0d | 3519 | # format like "1080p", "720p SD", or 1080. |
ed0cf9b3 | 3520 | height = int_or_none(self._search_regex( |
0236cd0d | 3521 | r'^(\d{3,4})[pP]?(?:\b|$)', compat_str(source.get('label') or ''), |
ed0cf9b3 S |
3522 | 'height', default=None)) |
3523 | a_format = { | |
3524 | 'url': source_url, | |
3525 | 'width': int_or_none(source.get('width')), | |
3526 | 'height': height, | |
0236cd0d | 3527 | 'tbr': int_or_none(source.get('bitrate')), |
ed0cf9b3 S |
3528 | 'ext': ext, |
3529 | } | |
3530 | if source_url.startswith('rtmp'): | |
3531 | a_format['ext'] = 'flv' | |
ed0cf9b3 S |
3532 | # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as |
3533 | # of jwplayer.flash.swf | |
3534 | rtmp_url_parts = re.split( | |
3535 | r'((?:mp4|mp3|flv):)', source_url, 1) | |
3536 | if len(rtmp_url_parts) == 3: | |
3537 | rtmp_url, prefix, play_path = rtmp_url_parts | |
3538 | a_format.update({ | |
3539 | 'url': rtmp_url, | |
3540 | 'play_path': prefix + play_path, | |
3541 | }) | |
3542 | if rtmp_params: | |
3543 | a_format.update(rtmp_params) | |
3544 | formats.append(a_format) | |
3545 | return formats | |
3546 | ||
f4b1c7ad | 3547 | def _live_title(self, name): |
39ca3b5c | 3548 | self._downloader.deprecation_warning('yt_dlp.InfoExtractor._live_title is deprecated and does not work as expected') |
3549 | return name | |
f4b1c7ad | 3550 | |
b14f3a4c PH |
3551 | def _int(self, v, name, fatal=False, **kwargs): |
3552 | res = int_or_none(v, **kwargs) | |
b14f3a4c | 3553 | if res is None: |
86e5f3ed | 3554 | msg = f'Failed to extract {name}: Could not parse value {v!r}' |
b14f3a4c PH |
3555 | if fatal: |
3556 | raise ExtractorError(msg) | |
3557 | else: | |
6a39ee13 | 3558 | self.report_warning(msg) |
b14f3a4c PH |
3559 | return res |
3560 | ||
3561 | def _float(self, v, name, fatal=False, **kwargs): | |
3562 | res = float_or_none(v, **kwargs) | |
3563 | if res is None: | |
86e5f3ed | 3564 | msg = f'Failed to extract {name}: Could not parse value {v!r}' |
b14f3a4c PH |
3565 | if fatal: |
3566 | raise ExtractorError(msg) | |
3567 | else: | |
6a39ee13 | 3568 | self.report_warning(msg) |
b14f3a4c PH |
3569 | return res |
3570 | ||
40e41780 TF |
3571 | def _set_cookie(self, domain, name, value, expire_time=None, port=None, |
3572 | path='/', secure=False, discard=False, rest={}, **kwargs): | |
6c22cee6 | 3573 | cookie = compat_cookiejar_Cookie( |
4ed2d7b7 | 3574 | 0, name, value, port, port is not None, domain, True, |
40e41780 TF |
3575 | domain.startswith('.'), path, True, secure, expire_time, |
3576 | discard, None, None, rest) | |
42939b61 JMF |
3577 | self._downloader.cookiejar.set_cookie(cookie) |
3578 | ||
799207e8 | 3579 | def _get_cookies(self, url): |
f7ad7160 | 3580 | """ Return a compat_cookies_SimpleCookie with the cookies for the url """ |
5c2266df | 3581 | req = sanitized_Request(url) |
799207e8 | 3582 | self._downloader.cookiejar.add_cookie_header(req) |
f7ad7160 | 3583 | return compat_cookies_SimpleCookie(req.get_header('Cookie')) |
799207e8 | 3584 | |
e3c1266f | 3585 | def _apply_first_set_cookie_header(self, url_handle, cookie): |
ce2fe4c0 S |
3586 | """ |
3587 | Apply first Set-Cookie header instead of the last. Experimental. | |
3588 | ||
3589 | Some sites (e.g. [1-3]) may serve two cookies under the same name | |
3590 | in Set-Cookie header and expect the first (old) one to be set rather | |
3591 | than second (new). However, as of RFC6265 the newer one cookie | |
3592 | should be set into cookie store what actually happens. | |
3593 | We will workaround this issue by resetting the cookie to | |
3594 | the first one manually. | |
3595 | 1. https://new.vk.com/ | |
3596 | 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201 | |
3597 | 3. https://learning.oreilly.com/ | |
3598 | """ | |
e3c1266f S |
3599 | for header, cookies in url_handle.headers.items(): |
3600 | if header.lower() != 'set-cookie': | |
3601 | continue | |
cfb0511d | 3602 | cookies = cookies.encode('iso-8859-1').decode('utf-8') |
e3c1266f S |
3603 | cookie_value = re.search( |
3604 | r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies) | |
3605 | if cookie_value: | |
3606 | value, domain = cookie_value.groups() | |
3607 | self._set_cookie(domain, cookie, value) | |
3608 | break | |
3609 | ||
05900629 PH |
3610 | def get_testcases(self, include_onlymatching=False): |
3611 | t = getattr(self, '_TEST', None) | |
3612 | if t: | |
3613 | assert not hasattr(self, '_TESTS'), \ | |
3614 | '%s has _TEST and _TESTS' % type(self).__name__ | |
3615 | tests = [t] | |
3616 | else: | |
3617 | tests = getattr(self, '_TESTS', []) | |
3618 | for t in tests: | |
3619 | if not include_onlymatching and t.get('only_matching', False): | |
3620 | continue | |
3621 | t['name'] = type(self).__name__[:-len('IE')] | |
3622 | yield t | |
3623 | ||
3624 | def is_suitable(self, age_limit): | |
3625 | """ Test whether the extractor is generally suitable for the given | |
3626 | age limit (i.e. pornographic sites are not, all others usually are) """ | |
3627 | ||
3628 | any_restricted = False | |
3629 | for tc in self.get_testcases(include_onlymatching=False): | |
40090e8d | 3630 | if tc.get('playlist', []): |
05900629 PH |
3631 | tc = tc['playlist'][0] |
3632 | is_restricted = age_restricted( | |
3633 | tc.get('info_dict', {}).get('age_limit'), age_limit) | |
3634 | if not is_restricted: | |
3635 | return True | |
3636 | any_restricted = any_restricted or is_restricted | |
3637 | return not any_restricted | |
3638 | ||
a504ced0 | 3639 | def extract_subtitles(self, *args, **kwargs): |
a06916d9 | 3640 | if (self.get_param('writesubtitles', False) |
3641 | or self.get_param('listsubtitles')): | |
9868ea49 JMF |
3642 | return self._get_subtitles(*args, **kwargs) |
3643 | return {} | |
a504ced0 JMF |
3644 | |
3645 | def _get_subtitles(self, *args, **kwargs): | |
611c1dd9 | 3646 | raise NotImplementedError('This method must be implemented by subclasses') |
a504ced0 | 3647 | |
a2160aa4 | 3648 | def extract_comments(self, *args, **kwargs): |
3649 | if not self.get_param('getcomments'): | |
3650 | return None | |
3651 | generator = self._get_comments(*args, **kwargs) | |
3652 | ||
3653 | def extractor(): | |
3654 | comments = [] | |
d2b2fca5 | 3655 | interrupted = True |
a2160aa4 | 3656 | try: |
3657 | while True: | |
3658 | comments.append(next(generator)) | |
a2160aa4 | 3659 | except StopIteration: |
3660 | interrupted = False | |
d2b2fca5 | 3661 | except KeyboardInterrupt: |
3662 | self.to_screen('Interrupted by user') | |
3663 | except Exception as e: | |
3664 | if self.get_param('ignoreerrors') is not True: | |
3665 | raise | |
3666 | self._downloader.report_error(e) | |
a2160aa4 | 3667 | comment_count = len(comments) |
3668 | self.to_screen(f'Extracted {comment_count} comments') | |
3669 | return { | |
3670 | 'comments': comments, | |
3671 | 'comment_count': None if interrupted else comment_count | |
3672 | } | |
3673 | return extractor | |
3674 | ||
3675 | def _get_comments(self, *args, **kwargs): | |
3676 | raise NotImplementedError('This method must be implemented by subclasses') | |
3677 | ||
912e0b7e YCH |
3678 | @staticmethod |
3679 | def _merge_subtitle_items(subtitle_list1, subtitle_list2): | |
a825ffbf | 3680 | """ Merge subtitle items for one language. Items with duplicated URLs/data |
912e0b7e | 3681 | will be dropped. """ |
86e5f3ed | 3682 | list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1} |
912e0b7e | 3683 | ret = list(subtitle_list1) |
a44ca5a4 | 3684 | ret.extend(item for item in subtitle_list2 if (item.get('url'), item.get('data')) not in list1_data) |
912e0b7e YCH |
3685 | return ret |
3686 | ||
3687 | @classmethod | |
46890374 | 3688 | def _merge_subtitles(cls, *dicts, target=None): |
19bb3920 | 3689 | """ Merge subtitle dictionaries, language by language. """ |
19bb3920 F |
3690 | if target is None: |
3691 | target = {} | |
3692 | for d in dicts: | |
3693 | for lang, subs in d.items(): | |
3694 | target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs) | |
3695 | return target | |
912e0b7e | 3696 | |
360e1ca5 | 3697 | def extract_automatic_captions(self, *args, **kwargs): |
a06916d9 | 3698 | if (self.get_param('writeautomaticsub', False) |
3699 | or self.get_param('listsubtitles')): | |
9868ea49 JMF |
3700 | return self._get_automatic_captions(*args, **kwargs) |
3701 | return {} | |
360e1ca5 JMF |
3702 | |
3703 | def _get_automatic_captions(self, *args, **kwargs): | |
611c1dd9 | 3704 | raise NotImplementedError('This method must be implemented by subclasses') |
360e1ca5 | 3705 | |
d77ab8e2 | 3706 | def mark_watched(self, *args, **kwargs): |
1813a6cc | 3707 | if not self.get_param('mark_watched', False): |
3708 | return | |
52efa4b3 | 3709 | if (self.supports_login() and self._get_login_info()[0] is not None |
3710 | or self.get_param('cookiefile') or self.get_param('cookiesfrombrowser')): | |
d77ab8e2 S |
3711 | self._mark_watched(*args, **kwargs) |
3712 | ||
3713 | def _mark_watched(self, *args, **kwargs): | |
3714 | raise NotImplementedError('This method must be implemented by subclasses') | |
3715 | ||
38cce791 YCH |
3716 | def geo_verification_headers(self): |
3717 | headers = {} | |
a06916d9 | 3718 | geo_verification_proxy = self.get_param('geo_verification_proxy') |
38cce791 YCH |
3719 | if geo_verification_proxy: |
3720 | headers['Ytdl-request-proxy'] = geo_verification_proxy | |
3721 | return headers | |
3722 | ||
98763ee3 YCH |
3723 | def _generic_id(self, url): |
3724 | return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0]) | |
3725 | ||
3726 | def _generic_title(self, url): | |
3727 | return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0]) | |
3728 | ||
c224251a | 3729 | @staticmethod |
b0089e89 | 3730 | def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None): |
c224251a M |
3731 | all_known = all(map( |
3732 | lambda x: x is not None, | |
3733 | (is_private, needs_premium, needs_subscription, needs_auth, is_unlisted))) | |
3734 | return ( | |
3735 | 'private' if is_private | |
3736 | else 'premium_only' if needs_premium | |
3737 | else 'subscriber_only' if needs_subscription | |
3738 | else 'needs_auth' if needs_auth | |
3739 | else 'unlisted' if is_unlisted | |
3740 | else 'public' if all_known | |
3741 | else None) | |
3742 | ||
d43de682 | 3743 | def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False): |
4bb6b02f | 3744 | ''' |
3745 | @returns A list of values for the extractor argument given by "key" | |
3746 | or "default" if no such key is present | |
3747 | @param default The default value to return when the key is not present (default: []) | |
3748 | @param casesense When false, the values are converted to lower case | |
3749 | ''' | |
3750 | val = traverse_obj( | |
d43de682 | 3751 | self._downloader.params, ('extractor_args', (ie_key or self.ie_key()).lower(), key)) |
4bb6b02f | 3752 | if val is None: |
3753 | return [] if default is NO_DEFAULT else default | |
3754 | return list(val) if casesense else [x.lower() for x in val] | |
5d3a0e79 | 3755 | |
f40ee5e9 | 3756 | def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist', video_label='video'): |
3757 | if not playlist_id or not video_id: | |
3758 | return not video_id | |
3759 | ||
3760 | no_playlist = (smuggled_data or {}).get('force_noplaylist') | |
3761 | if no_playlist is not None: | |
3762 | return not no_playlist | |
3763 | ||
3764 | video_id = '' if video_id is True else f' {video_id}' | |
3765 | playlist_id = '' if playlist_id is True else f' {playlist_id}' | |
3766 | if self.get_param('noplaylist'): | |
3767 | self.to_screen(f'Downloading just the {video_label}{video_id} because of --no-playlist') | |
3768 | return False | |
3769 | self.to_screen(f'Downloading {playlist_label}{playlist_id} - add --no-playlist to download just the {video_label}{video_id}') | |
3770 | return True | |
3771 | ||
8dbe9899 | 3772 | |
d6983cb4 PH |
3773 | class SearchInfoExtractor(InfoExtractor): |
3774 | """ | |
3775 | Base class for paged search queries extractors. | |
10952eb2 | 3776 | They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query} |
96565c7e | 3777 | Instances should define _SEARCH_KEY and optionally _MAX_RESULTS |
d6983cb4 PH |
3778 | """ |
3779 | ||
96565c7e | 3780 | _MAX_RESULTS = float('inf') |
3781 | ||
d6983cb4 PH |
3782 | @classmethod |
3783 | def _make_valid_url(cls): | |
3784 | return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY | |
3785 | ||
d6983cb4 | 3786 | def _real_extract(self, query): |
2c4aaadd | 3787 | prefix, query = self._match_valid_url(query).group('prefix', 'query') |
d6983cb4 PH |
3788 | if prefix == '': |
3789 | return self._get_n_results(query, 1) | |
3790 | elif prefix == 'all': | |
3791 | return self._get_n_results(query, self._MAX_RESULTS) | |
3792 | else: | |
3793 | n = int(prefix) | |
3794 | if n <= 0: | |
86e5f3ed | 3795 | raise ExtractorError(f'invalid download number {n} for query "{query}"') |
d6983cb4 | 3796 | elif n > self._MAX_RESULTS: |
6a39ee13 | 3797 | self.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n)) |
d6983cb4 PH |
3798 | n = self._MAX_RESULTS |
3799 | return self._get_n_results(query, n) | |
3800 | ||
3801 | def _get_n_results(self, query, n): | |
cc16383f | 3802 | """Get a specified number of results for a query. |
3803 | Either this function or _search_results must be overridden by subclasses """ | |
3804 | return self.playlist_result( | |
3805 | itertools.islice(self._search_results(query), 0, None if n == float('inf') else n), | |
3806 | query, query) | |
3807 | ||
3808 | def _search_results(self, query): | |
3809 | """Returns an iterator of search results""" | |
611c1dd9 | 3810 | raise NotImplementedError('This method must be implemented by subclasses') |
0f818663 PH |
3811 | |
3812 | @property | |
3813 | def SEARCH_KEY(self): | |
3814 | return self._SEARCH_KEY |