]>
Commit | Line | Data |
---|---|---|
fd475508 | 1 | # coding: utf-8 |
6a3828fd | 2 | from __future__ import unicode_literals |
f1a9d64e | 3 | |
d6983cb4 | 4 | import base64 |
f4b1c7ad | 5 | import datetime |
3ec05685 | 6 | import hashlib |
3d3538e4 | 7 | import json |
4094b6e3 | 8 | import netrc |
d6983cb4 | 9 | import os |
773f291d | 10 | import random |
d6983cb4 PH |
11 | import re |
12 | import socket | |
f8c7bed1 | 13 | import ssl |
d6983cb4 | 14 | import sys |
4094b6e3 | 15 | import time |
1bac3455 | 16 | import math |
d6983cb4 | 17 | |
8c25f81b | 18 | from ..compat import ( |
6c22cee6 | 19 | compat_cookiejar_Cookie, |
799207e8 | 20 | compat_cookies, |
ee0ba927 | 21 | compat_etree_Element, |
e9c0cdd3 | 22 | compat_etree_fromstring, |
e64b7569 | 23 | compat_getpass, |
d391b7e2 | 24 | compat_integer_types, |
d6983cb4 | 25 | compat_http_client, |
e9c0cdd3 YCH |
26 | compat_os_name, |
27 | compat_str, | |
d6983cb4 | 28 | compat_urllib_error, |
98763ee3 | 29 | compat_urllib_parse_unquote, |
15707c7e | 30 | compat_urllib_parse_urlencode, |
41d06b04 | 31 | compat_urllib_request, |
f0b5d6af | 32 | compat_urlparse, |
e01c3d2e | 33 | compat_xml_parse_error, |
8c25f81b | 34 | ) |
48107c19 S |
35 | from ..downloader.f4m import ( |
36 | get_base_url, | |
37 | remove_encrypted_media, | |
38 | ) | |
8c25f81b | 39 | from ..utils import ( |
c342041f | 40 | NO_DEFAULT, |
05900629 | 41 | age_restricted, |
02dc0a36 | 42 | base_url, |
08f2a92c | 43 | bug_reports_message, |
d6983cb4 PH |
44 | clean_html, |
45 | compiled_regex_type, | |
70f0f5a8 | 46 | determine_ext, |
46b18f23 | 47 | determine_protocol, |
d493f15c | 48 | dict_get, |
9b9c5355 | 49 | error_to_compat_str, |
d6983cb4 | 50 | ExtractorError, |
46b18f23 | 51 | extract_attributes, |
97f4aecf | 52 | fix_xml_ampersands, |
b14f3a4c | 53 | float_or_none, |
773f291d S |
54 | GeoRestrictedError, |
55 | GeoUtils, | |
31bb8d3f | 56 | int_or_none, |
a4a554a7 | 57 | js_to_json, |
0685d972 | 58 | JSON_LD_RE, |
46b18f23 JH |
59 | mimetype2ext, |
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, |
55b3e45b | 67 | RegexNotFoundError, |
5c2266df | 68 | sanitized_Request, |
46b18f23 | 69 | sanitize_filename, |
d493f15c | 70 | str_or_none, |
ce5b9040 | 71 | str_to_int, |
f856816b | 72 | strip_or_none, |
f38de77f | 73 | unescapeHTML, |
647eab45 | 74 | unified_strdate, |
6b3a3098 | 75 | unified_timestamp, |
46b18f23 JH |
76 | update_Request, |
77 | update_url_query, | |
78 | urljoin, | |
a107193e | 79 | url_basename, |
bebef109 | 80 | url_or_none, |
a6571f10 | 81 | xpath_element, |
8d6765cf S |
82 | xpath_text, |
83 | xpath_with_ns, | |
d6983cb4 | 84 | ) |
c342041f | 85 | |
d6983cb4 PH |
86 | |
87 | class InfoExtractor(object): | |
88 | """Information Extractor class. | |
89 | ||
90 | Information extractors are the classes that, given a URL, extract | |
91 | information about the video (or videos) the URL refers to. This | |
92 | information includes the real video URL, the video title, author and | |
93 | others. The information is stored in a dictionary which is then | |
5d380852 | 94 | passed to the YoutubeDL. The YoutubeDL processes this |
d6983cb4 PH |
95 | information possibly downloading the video to the file system, among |
96 | other possible outcomes. | |
97 | ||
cf0649f8 | 98 | The type field determines the type of the result. |
fed5d032 PH |
99 | By far the most common value (and the default if _type is missing) is |
100 | "video", which indicates a single video. | |
101 | ||
102 | For a video, the dictionaries must include the following fields: | |
d6983cb4 PH |
103 | |
104 | id: Video identifier. | |
d6983cb4 | 105 | title: Video title, unescaped. |
d67b0b15 | 106 | |
f49d89ee | 107 | Additionally, it must contain either a formats entry or a url one: |
d67b0b15 | 108 | |
f49d89ee PH |
109 | formats: A list of dictionaries for each format available, ordered |
110 | from worst to best quality. | |
111 | ||
112 | Potential fields: | |
c790e93a S |
113 | * url The mandatory URL representing the media: |
114 | for plain file media - HTTP URL of this file, | |
115 | for RTMP - RTMP URL, | |
116 | for HLS - URL of the M3U8 media playlist, | |
117 | for HDS - URL of the F4M manifest, | |
79d2077e S |
118 | for DASH |
119 | - HTTP URL to plain file media (in case of | |
120 | unfragmented media) | |
121 | - URL of the MPD manifest or base URL | |
122 | representing the media if MPD manifest | |
8ed7a233 | 123 | is parsed from a string (in case of |
79d2077e | 124 | fragmented media) |
c790e93a | 125 | for MSS - URL of the ISM manifest. |
86f4d14f S |
126 | * manifest_url |
127 | The URL of the manifest file in case of | |
c790e93a S |
128 | fragmented media: |
129 | for HLS - URL of the M3U8 master playlist, | |
130 | for HDS - URL of the F4M manifest, | |
131 | for DASH - URL of the MPD manifest, | |
132 | for MSS - URL of the ISM manifest. | |
10952eb2 | 133 | * ext Will be calculated from URL if missing |
d67b0b15 PH |
134 | * format A human-readable description of the format |
135 | ("mp4 container with h264/opus"). | |
136 | Calculated from the format_id, width, height. | |
137 | and format_note fields if missing. | |
138 | * format_id A short description of the format | |
5d4f3985 PH |
139 | ("mp4_h264_opus" or "19"). |
140 | Technically optional, but strongly recommended. | |
d67b0b15 PH |
141 | * format_note Additional info about the format |
142 | ("3D" or "DASH video") | |
143 | * width Width of the video, if known | |
144 | * height Height of the video, if known | |
f49d89ee | 145 | * resolution Textual description of width and height |
7217e148 | 146 | * tbr Average bitrate of audio and video in KBit/s |
d67b0b15 PH |
147 | * abr Average audio bitrate in KBit/s |
148 | * acodec Name of the audio codec in use | |
dd27fd17 | 149 | * asr Audio sampling rate in Hertz |
d67b0b15 | 150 | * vbr Average video bitrate in KBit/s |
fbb21cf5 | 151 | * fps Frame rate |
d67b0b15 | 152 | * vcodec Name of the video codec in use |
1394ce65 | 153 | * container Name of the container format |
d67b0b15 | 154 | * filesize The number of bytes, if known in advance |
9732d77e | 155 | * filesize_approx An estimate for the number of bytes |
d67b0b15 | 156 | * player_url SWF Player URL (used for rtmpdump). |
c7deaa4c PH |
157 | * protocol The protocol that will be used for the actual |
158 | download, lower-case. | |
b04b8852 | 159 | "http", "https", "rtsp", "rtmp", "rtmpe", |
af7d5a63 | 160 | "m3u8", "m3u8_native" or "http_dash_segments". |
c58c2d63 S |
161 | * fragment_base_url |
162 | Base URL for fragments. Each fragment's path | |
163 | value (if present) will be relative to | |
164 | this URL. | |
165 | * fragments A list of fragments of a fragmented media. | |
166 | Each fragment entry must contain either an url | |
167 | or a path. If an url is present it should be | |
168 | considered by a client. Otherwise both path and | |
169 | fragment_base_url must be present. Here is | |
170 | the list of all potential fields: | |
171 | * "url" - fragment's URL | |
172 | * "path" - fragment's path relative to | |
173 | fragment_base_url | |
a0d5077c S |
174 | * "duration" (optional, int or float) |
175 | * "filesize" (optional, int) | |
f49d89ee | 176 | * preference Order number of this format. If this field is |
08d13955 | 177 | present and not None, the formats get sorted |
38d63d84 | 178 | by this field, regardless of all other values. |
f49d89ee PH |
179 | -1 for default (order by other properties), |
180 | -2 or smaller for less than default. | |
e65566a9 PH |
181 | < -1000 to hide the format (if there is |
182 | another one which is strictly better) | |
32f90364 PH |
183 | * language Language code, e.g. "de" or "en-US". |
184 | * language_preference Is this in the language mentioned in | |
185 | the URL? | |
aff2f4f4 PH |
186 | 10 if it's what the URL is about, |
187 | -1 for default (don't know), | |
188 | -10 otherwise, other values reserved for now. | |
5d73273f PH |
189 | * quality Order number of the video quality of this |
190 | format, irrespective of the file format. | |
191 | -1 for default (order by other properties), | |
192 | -2 or smaller for less than default. | |
c64ed2a3 PH |
193 | * source_preference Order number for this video source |
194 | (quality takes higher priority) | |
195 | -1 for default (order by other properties), | |
196 | -2 or smaller for less than default. | |
d769be6c PH |
197 | * http_headers A dictionary of additional HTTP headers |
198 | to add to the request. | |
6271f1ca | 199 | * stretched_ratio If given and not 1, indicates that the |
3dee7826 PH |
200 | video's pixels are not square. |
201 | width : height ratio as float. | |
202 | * no_resume The server does not support resuming the | |
203 | (HTTP or RTMP) download. Boolean. | |
00c97e3e S |
204 | * downloader_options A dictionary of downloader options as |
205 | described in FileDownloader | |
3dee7826 | 206 | |
c0ba0f48 | 207 | url: Final video URL. |
d6983cb4 | 208 | ext: Video filename extension. |
d67b0b15 PH |
209 | format: The video format, defaults to ext (used for --get-format) |
210 | player_url: SWF Player URL (used for rtmpdump). | |
2f5865cc | 211 | |
d6983cb4 PH |
212 | The following fields are optional: |
213 | ||
f5e43bc6 | 214 | alt_title: A secondary title of the video. |
0afef30b PH |
215 | display_id An alternative identifier for the video, not necessarily |
216 | unique, but available before title. Typically, id is | |
217 | something like "4234987", title "Dancing naked mole rats", | |
218 | and display_id "dancing-naked-mole-rats" | |
d5519808 | 219 | thumbnails: A list of dictionaries, with the following entries: |
cfb56d1a | 220 | * "id" (optional, string) - Thumbnail format ID |
d5519808 | 221 | * "url" |
cfb56d1a | 222 | * "preference" (optional, int) - quality of the image |
d5519808 PH |
223 | * "width" (optional, int) |
224 | * "height" (optional, int) | |
5e1c39ac | 225 | * "resolution" (optional, string "{width}x{height}", |
d5519808 | 226 | deprecated) |
2de624fd | 227 | * "filesize" (optional, int) |
d6983cb4 | 228 | thumbnail: Full URL to a video thumbnail image. |
f5e43bc6 | 229 | description: Full video description. |
d6983cb4 | 230 | uploader: Full name of the video uploader. |
2bc0c46f | 231 | license: License name the video is licensed under. |
8a92e51c | 232 | creator: The creator of the video. |
8aab976b | 233 | release_date: The date (YYYYMMDD) when the video was released. |
955c4514 | 234 | timestamp: UNIX timestamp of the moment the video became available. |
d6983cb4 | 235 | upload_date: Video upload date (YYYYMMDD). |
955c4514 | 236 | If not explicitly set, calculated from timestamp. |
d6983cb4 | 237 | uploader_id: Nickname or id of the video uploader. |
7bcd2830 | 238 | uploader_url: Full URL to a personal webpage of the video uploader. |
6f1f59f3 | 239 | channel: Full name of the channel the video is uploaded on. |
0e7b8d3e | 240 | Note that channel fields may or may not repeat uploader |
6f1f59f3 S |
241 | fields. This depends on a particular extractor. |
242 | channel_id: Id of the channel. | |
243 | channel_url: Full URL to a channel webpage. | |
da9ec3b9 | 244 | location: Physical location where the video was filmed. |
a504ced0 | 245 | subtitles: The available subtitles as a dictionary in the format |
4606c34e YCH |
246 | {tag: subformats}. "tag" is usually a language code, and |
247 | "subformats" is a list sorted from lower to higher | |
248 | preference, each element is a dictionary with the "ext" | |
249 | entry and one of: | |
a504ced0 | 250 | * "data": The subtitles file contents |
10952eb2 | 251 | * "url": A URL pointing to the subtitles file |
4bba3716 | 252 | "ext" will be calculated from URL if missing |
360e1ca5 JMF |
253 | automatic_captions: Like 'subtitles', used by the YoutubeIE for |
254 | automatically generated captions | |
62d231c0 | 255 | duration: Length of the video in seconds, as an integer or float. |
f3d29461 | 256 | view_count: How many users have watched the video on the platform. |
19e3dfc9 PH |
257 | like_count: Number of positive ratings of the video |
258 | dislike_count: Number of negative ratings of the video | |
02835c6b | 259 | repost_count: Number of reposts of the video |
2d30521a | 260 | average_rating: Average rating give by users, the scale used depends on the webpage |
19e3dfc9 | 261 | comment_count: Number of comments on the video |
dd622d7c PH |
262 | comments: A list of comments, each with one or more of the following |
263 | properties (all but one of text or html optional): | |
264 | * "author" - human-readable name of the comment author | |
265 | * "author_id" - user ID of the comment author | |
266 | * "id" - Comment ID | |
267 | * "html" - Comment as HTML | |
268 | * "text" - Plain text of the comment | |
269 | * "timestamp" - UNIX timestamp of comment | |
270 | * "parent" - ID of the comment this one is replying to. | |
271 | Set to "root" to indicate that this is a | |
272 | comment to the original video. | |
8dbe9899 | 273 | age_limit: Age restriction for the video, as an integer (years) |
cefecac1 | 274 | webpage_url: The URL to the video webpage, if given to youtube-dlc it |
9103bbc5 JMF |
275 | should allow to get the same result again. (It will be set |
276 | by YoutubeDL if it's missing) | |
ad3bc6ac PH |
277 | categories: A list of categories that the video falls in, for example |
278 | ["Sports", "Berlin"] | |
864f24bd | 279 | tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"] |
7267bd53 PH |
280 | is_live: True, False, or None (=unknown). Whether this video is a |
281 | live stream that goes on instead of a fixed-length video. | |
7c80519c | 282 | start_time: Time in seconds where the reproduction should start, as |
10952eb2 | 283 | specified in the URL. |
297a564b | 284 | end_time: Time in seconds where the reproduction should end, as |
10952eb2 | 285 | specified in the URL. |
55949fed | 286 | chapters: A list of dictionaries, with the following entries: |
287 | * "start_time" - The start time of the chapter in seconds | |
288 | * "end_time" - The end time of the chapter in seconds | |
289 | * "title" (optional, string) | |
d6983cb4 | 290 | |
7109903e S |
291 | The following fields should only be used when the video belongs to some logical |
292 | chapter or section: | |
293 | ||
294 | chapter: Name or title of the chapter the video belongs to. | |
27bfd4e5 S |
295 | chapter_number: Number of the chapter the video belongs to, as an integer. |
296 | chapter_id: Id of the chapter the video belongs to, as a unicode string. | |
7109903e S |
297 | |
298 | The following fields should only be used when the video is an episode of some | |
8d76bdf1 | 299 | series, programme or podcast: |
7109903e S |
300 | |
301 | series: Title of the series or programme the video episode belongs to. | |
302 | season: Title of the season the video episode belongs to. | |
27bfd4e5 S |
303 | season_number: Number of the season the video episode belongs to, as an integer. |
304 | season_id: Id of the season the video episode belongs to, as a unicode string. | |
7109903e S |
305 | episode: Title of the video episode. Unlike mandatory video title field, |
306 | this field should denote the exact title of the video episode | |
307 | without any kind of decoration. | |
27bfd4e5 S |
308 | episode_number: Number of the video episode within a season, as an integer. |
309 | episode_id: Id of the video episode, as a unicode string. | |
7109903e | 310 | |
7a93ab5f S |
311 | The following fields should only be used when the media is a track or a part of |
312 | a music album: | |
313 | ||
314 | track: Title of the track. | |
315 | track_number: Number of the track within an album or a disc, as an integer. | |
316 | track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii), | |
317 | as a unicode string. | |
318 | artist: Artist(s) of the track. | |
319 | genre: Genre(s) of the track. | |
320 | album: Title of the album the track belongs to. | |
321 | album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc). | |
322 | album_artist: List of all artists appeared on the album (e.g. | |
323 | "Ash Borer / Fell Voices" or "Various Artists", useful for splits | |
324 | and compilations). | |
325 | disc_number: Number of the disc or other physical medium the track belongs to, | |
326 | as an integer. | |
327 | release_year: Year (YYYY) when the album was released. | |
328 | ||
deefc05b | 329 | Unless mentioned otherwise, the fields should be Unicode strings. |
d6983cb4 | 330 | |
d838b1bd PH |
331 | Unless mentioned otherwise, None is equivalent to absence of information. |
332 | ||
fed5d032 PH |
333 | |
334 | _type "playlist" indicates multiple videos. | |
b82f815f PH |
335 | There must be a key "entries", which is a list, an iterable, or a PagedList |
336 | object, each element of which is a valid dictionary by this specification. | |
fed5d032 | 337 | |
c10c9323 S |
338 | Additionally, playlists can have "id", "title", "description", "uploader", |
339 | "uploader_id", "uploader_url" attributes with the same semantics as videos | |
340 | (see above). | |
fed5d032 PH |
341 | |
342 | ||
343 | _type "multi_video" indicates that there are multiple videos that | |
344 | form a single show, for examples multiple acts of an opera or TV episode. | |
345 | It must have an entries key like a playlist and contain all the keys | |
346 | required for a video at the same time. | |
347 | ||
348 | ||
349 | _type "url" indicates that the video must be extracted from another | |
350 | location, possibly by a different extractor. Its only required key is: | |
351 | "url" - the next URL to extract. | |
f58766ce PH |
352 | The key "ie_key" can be set to the class name (minus the trailing "IE", |
353 | e.g. "Youtube") if the extractor class is known in advance. | |
354 | Additionally, the dictionary may have any properties of the resolved entity | |
355 | known in advance, for example "title" if the title of the referred video is | |
fed5d032 PH |
356 | known ahead of time. |
357 | ||
358 | ||
359 | _type "url_transparent" entities have the same specification as "url", but | |
360 | indicate that the given additional information is more precise than the one | |
361 | associated with the resolved URL. | |
362 | This is useful when a site employs a video service that hosts the video and | |
363 | its technical metadata, but that video service does not embed a useful | |
364 | title, description etc. | |
365 | ||
366 | ||
d6983cb4 PH |
367 | Subclasses of this one should re-define the _real_initialize() and |
368 | _real_extract() methods and define a _VALID_URL regexp. | |
369 | Probably, they should also be added to the list of extractors. | |
370 | ||
4248dad9 | 371 | _GEO_BYPASS attribute may be set to False in order to disable |
773f291d S |
372 | geo restriction bypass mechanisms for a particular extractor. |
373 | Though it won't disable explicit geo restriction bypass based on | |
504f20dd | 374 | country code provided with geo_bypass_country. |
4248dad9 S |
375 | |
376 | _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted | |
377 | countries for this extractor. One of these countries will be used by | |
378 | geo restriction bypass mechanism right away in order to bypass | |
504f20dd | 379 | geo restriction, of course, if the mechanism is not disabled. |
773f291d | 380 | |
5f95927a S |
381 | _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted |
382 | IP blocks in CIDR notation for this extractor. One of these IP blocks | |
383 | will be used by geo restriction bypass mechanism similarly | |
504f20dd | 384 | to _GEO_COUNTRIES. |
3ccdde8c | 385 | |
d6983cb4 PH |
386 | Finally, the _WORKING attribute should be set to False for broken IEs |
387 | in order to warn the users and skip the tests. | |
388 | """ | |
389 | ||
390 | _ready = False | |
391 | _downloader = None | |
773f291d | 392 | _x_forwarded_for_ip = None |
4248dad9 S |
393 | _GEO_BYPASS = True |
394 | _GEO_COUNTRIES = None | |
5f95927a | 395 | _GEO_IP_BLOCKS = None |
d6983cb4 PH |
396 | _WORKING = True |
397 | ||
398 | def __init__(self, downloader=None): | |
399 | """Constructor. Receives an optional downloader.""" | |
400 | self._ready = False | |
773f291d | 401 | self._x_forwarded_for_ip = None |
d6983cb4 PH |
402 | self.set_downloader(downloader) |
403 | ||
404 | @classmethod | |
405 | def suitable(cls, url): | |
406 | """Receives a URL and returns True if suitable for this IE.""" | |
79cb2577 PH |
407 | |
408 | # This does not use has/getattr intentionally - we want to know whether | |
409 | # we have cached the regexp for *this* class, whereas getattr would also | |
410 | # match the superclass | |
411 | if '_VALID_URL_RE' not in cls.__dict__: | |
412 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) | |
413 | return cls._VALID_URL_RE.match(url) is not None | |
d6983cb4 | 414 | |
ed9266db PH |
415 | @classmethod |
416 | def _match_id(cls, url): | |
417 | if '_VALID_URL_RE' not in cls.__dict__: | |
418 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) | |
419 | m = cls._VALID_URL_RE.match(url) | |
420 | assert m | |
1afd0b0d | 421 | return compat_str(m.group('id')) |
ed9266db | 422 | |
d6983cb4 PH |
423 | @classmethod |
424 | def working(cls): | |
425 | """Getter method for _WORKING.""" | |
426 | return cls._WORKING | |
427 | ||
428 | def initialize(self): | |
429 | """Initializes an instance (authentication, etc).""" | |
5f95927a S |
430 | self._initialize_geo_bypass({ |
431 | 'countries': self._GEO_COUNTRIES, | |
432 | 'ip_blocks': self._GEO_IP_BLOCKS, | |
433 | }) | |
4248dad9 S |
434 | if not self._ready: |
435 | self._real_initialize() | |
436 | self._ready = True | |
437 | ||
5f95927a | 438 | def _initialize_geo_bypass(self, geo_bypass_context): |
e39b5d4a S |
439 | """ |
440 | Initialize geo restriction bypass mechanism. | |
441 | ||
442 | This method is used to initialize geo bypass mechanism based on faking | |
443 | X-Forwarded-For HTTP header. A random country from provided country list | |
dc0a869e | 444 | is selected and a random IP belonging to this country is generated. This |
e39b5d4a S |
445 | IP will be passed as X-Forwarded-For HTTP header in all subsequent |
446 | HTTP requests. | |
e39b5d4a S |
447 | |
448 | This method will be used for initial geo bypass mechanism initialization | |
5f95927a S |
449 | during the instance initialization with _GEO_COUNTRIES and |
450 | _GEO_IP_BLOCKS. | |
e39b5d4a | 451 | |
5f95927a | 452 | You may also manually call it from extractor's code if geo bypass |
e39b5d4a | 453 | information is not available beforehand (e.g. obtained during |
5f95927a S |
454 | extraction) or due to some other reason. In this case you should pass |
455 | this information in geo bypass context passed as first argument. It may | |
456 | contain following fields: | |
457 | ||
458 | countries: List of geo unrestricted countries (similar | |
459 | to _GEO_COUNTRIES) | |
460 | ip_blocks: List of geo unrestricted IP blocks in CIDR notation | |
461 | (similar to _GEO_IP_BLOCKS) | |
462 | ||
e39b5d4a | 463 | """ |
773f291d | 464 | if not self._x_forwarded_for_ip: |
5f95927a S |
465 | |
466 | # Geo bypass mechanism is explicitly disabled by user | |
467 | if not self._downloader.params.get('geo_bypass', True): | |
468 | return | |
469 | ||
470 | if not geo_bypass_context: | |
471 | geo_bypass_context = {} | |
472 | ||
473 | # Backward compatibility: previously _initialize_geo_bypass | |
474 | # expected a list of countries, some 3rd party code may still use | |
475 | # it this way | |
476 | if isinstance(geo_bypass_context, (list, tuple)): | |
477 | geo_bypass_context = { | |
478 | 'countries': geo_bypass_context, | |
479 | } | |
480 | ||
481 | # The whole point of geo bypass mechanism is to fake IP | |
482 | # as X-Forwarded-For HTTP header based on some IP block or | |
483 | # country code. | |
484 | ||
485 | # Path 1: bypassing based on IP block in CIDR notation | |
486 | ||
487 | # Explicit IP block specified by user, use it right away | |
488 | # regardless of whether extractor is geo bypassable or not | |
489 | ip_block = self._downloader.params.get('geo_bypass_ip_block', None) | |
490 | ||
491 | # Otherwise use random IP block from geo bypass context but only | |
492 | # if extractor is known as geo bypassable | |
493 | if not ip_block: | |
494 | ip_blocks = geo_bypass_context.get('ip_blocks') | |
495 | if self._GEO_BYPASS and ip_blocks: | |
496 | ip_block = random.choice(ip_blocks) | |
497 | ||
498 | if ip_block: | |
499 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(ip_block) | |
500 | if self._downloader.params.get('verbose', False): | |
501 | self._downloader.to_screen( | |
502 | '[debug] Using fake IP %s as X-Forwarded-For.' | |
503 | % self._x_forwarded_for_ip) | |
504 | return | |
505 | ||
506 | # Path 2: bypassing based on country code | |
507 | ||
508 | # Explicit country code specified by user, use it right away | |
509 | # regardless of whether extractor is geo bypassable or not | |
510 | country = self._downloader.params.get('geo_bypass_country', None) | |
511 | ||
512 | # Otherwise use random country code from geo bypass context but | |
513 | # only if extractor is known as geo bypassable | |
514 | if not country: | |
515 | countries = geo_bypass_context.get('countries') | |
516 | if self._GEO_BYPASS and countries: | |
517 | country = random.choice(countries) | |
518 | ||
519 | if country: | |
520 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country) | |
4248dad9 | 521 | if self._downloader.params.get('verbose', False): |
6a9cb295 | 522 | self._downloader.to_screen( |
eea0716c | 523 | '[debug] Using fake IP %s (%s) as X-Forwarded-For.' |
5f95927a | 524 | % (self._x_forwarded_for_ip, country.upper())) |
d6983cb4 PH |
525 | |
526 | def extract(self, url): | |
527 | """Extracts URL information and returns it in list of dicts.""" | |
3a5bcd03 | 528 | try: |
773f291d S |
529 | for _ in range(2): |
530 | try: | |
531 | self.initialize() | |
0016b84e S |
532 | ie_result = self._real_extract(url) |
533 | if self._x_forwarded_for_ip: | |
534 | ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip | |
535 | return ie_result | |
773f291d | 536 | except GeoRestrictedError as e: |
4248dad9 S |
537 | if self.__maybe_fake_ip_and_retry(e.countries): |
538 | continue | |
773f291d | 539 | raise |
3a5bcd03 PH |
540 | except ExtractorError: |
541 | raise | |
542 | except compat_http_client.IncompleteRead as e: | |
dfb1b146 | 543 | raise ExtractorError('A network error has occurred.', cause=e, expected=True) |
9650885b | 544 | except (KeyError, StopIteration) as e: |
dfb1b146 | 545 | raise ExtractorError('An extractor error has occurred.', cause=e) |
d6983cb4 | 546 | |
4248dad9 | 547 | def __maybe_fake_ip_and_retry(self, countries): |
3089bc74 S |
548 | if (not self._downloader.params.get('geo_bypass_country', None) |
549 | and self._GEO_BYPASS | |
550 | and self._downloader.params.get('geo_bypass', True) | |
551 | and not self._x_forwarded_for_ip | |
552 | and countries): | |
eea0716c S |
553 | country_code = random.choice(countries) |
554 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) | |
4248dad9 S |
555 | if self._x_forwarded_for_ip: |
556 | self.report_warning( | |
eea0716c S |
557 | 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.' |
558 | % (self._x_forwarded_for_ip, country_code.upper())) | |
4248dad9 S |
559 | return True |
560 | return False | |
561 | ||
d6983cb4 PH |
562 | def set_downloader(self, downloader): |
563 | """Sets the downloader for this IE.""" | |
564 | self._downloader = downloader | |
565 | ||
566 | def _real_initialize(self): | |
567 | """Real initialization process. Redefine in subclasses.""" | |
568 | pass | |
569 | ||
570 | def _real_extract(self, url): | |
571 | """Real extraction process. Redefine in subclasses.""" | |
572 | pass | |
573 | ||
56c73665 JMF |
574 | @classmethod |
575 | def ie_key(cls): | |
576 | """A string for getting the InfoExtractor with get_info_extractor""" | |
dc519b54 | 577 | return compat_str(cls.__name__[:-2]) |
56c73665 | 578 | |
d6983cb4 PH |
579 | @property |
580 | def IE_NAME(self): | |
dc519b54 | 581 | return compat_str(type(self).__name__[:-2]) |
d6983cb4 | 582 | |
d391b7e2 S |
583 | @staticmethod |
584 | def __can_accept_status_code(err, expected_status): | |
585 | assert isinstance(err, compat_urllib_error.HTTPError) | |
586 | if expected_status is None: | |
587 | return False | |
588 | if isinstance(expected_status, compat_integer_types): | |
589 | return err.code == expected_status | |
590 | elif isinstance(expected_status, (list, tuple)): | |
591 | return err.code in expected_status | |
592 | elif callable(expected_status): | |
593 | return expected_status(err.code) is True | |
594 | else: | |
595 | assert False | |
596 | ||
597 | def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers={}, query={}, expected_status=None): | |
598 | """ | |
599 | Return the response handle. | |
600 | ||
601 | See _download_webpage docstring for arguments specification. | |
602 | """ | |
d6983cb4 PH |
603 | if note is None: |
604 | self.report_download_webpage(video_id) | |
605 | elif note is not False: | |
7cc3570e | 606 | if video_id is None: |
f1a9d64e | 607 | self.to_screen('%s' % (note,)) |
7cc3570e | 608 | else: |
f1a9d64e | 609 | self.to_screen('%s: %s' % (video_id, note)) |
2132edaa S |
610 | |
611 | # Some sites check X-Forwarded-For HTTP header in order to figure out | |
612 | # the origin of the client behind proxy. This allows bypassing geo | |
613 | # restriction by faking this header's value to IP that belongs to some | |
614 | # geo unrestricted country. We will do so once we encounter any | |
615 | # geo restriction error. | |
616 | if self._x_forwarded_for_ip: | |
617 | if 'X-Forwarded-For' not in headers: | |
618 | headers['X-Forwarded-For'] = self._x_forwarded_for_ip | |
619 | ||
41d06b04 S |
620 | if isinstance(url_or_request, compat_urllib_request.Request): |
621 | url_or_request = update_Request( | |
622 | url_or_request, data=data, headers=headers, query=query) | |
623 | else: | |
cdfee168 | 624 | if query: |
625 | url_or_request = update_url_query(url_or_request, query) | |
2c0d9c62 | 626 | if data is not None or headers: |
41d06b04 | 627 | url_or_request = sanitized_Request(url_or_request, data, headers) |
f8c7bed1 S |
628 | exceptions = [compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error] |
629 | if hasattr(ssl, 'CertificateError'): | |
630 | exceptions.append(ssl.CertificateError) | |
d6983cb4 | 631 | try: |
dca08720 | 632 | return self._downloader.urlopen(url_or_request) |
f8c7bed1 | 633 | except tuple(exceptions) as err: |
d391b7e2 S |
634 | if isinstance(err, compat_urllib_error.HTTPError): |
635 | if self.__can_accept_status_code(err, expected_status): | |
95e42d73 XDG |
636 | # Retain reference to error to prevent file object from |
637 | # being closed before it can be read. Works around the | |
638 | # effects of <https://bugs.python.org/issue15002> | |
639 | # introduced in Python 3.4.1. | |
640 | err.fp._error = err | |
d391b7e2 S |
641 | return err.fp |
642 | ||
aa94a6d3 PH |
643 | if errnote is False: |
644 | return False | |
d6983cb4 | 645 | if errnote is None: |
f1a9d64e | 646 | errnote = 'Unable to download webpage' |
7f8b2714 | 647 | |
9b9c5355 | 648 | errmsg = '%s: %s' % (errnote, error_to_compat_str(err)) |
7cc3570e PH |
649 | if fatal: |
650 | raise ExtractorError(errmsg, sys.exc_info()[2], cause=err) | |
651 | else: | |
652 | self._downloader.report_warning(errmsg) | |
653 | return False | |
d6983cb4 | 654 | |
d391b7e2 S |
655 | 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): |
656 | """ | |
657 | Return a tuple (page content as string, URL handle). | |
658 | ||
659 | See _download_webpage docstring for arguments specification. | |
660 | """ | |
b9d3e163 PH |
661 | # Strip hashes from the URL (#1038) |
662 | if isinstance(url_or_request, (compat_str, str)): | |
663 | url_or_request = url_or_request.partition('#')[0] | |
664 | ||
d391b7e2 | 665 | urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query, expected_status=expected_status) |
7cc3570e PH |
666 | if urlh is False: |
667 | assert not fatal | |
668 | return False | |
c9a77969 | 669 | content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding) |
23be51d8 PH |
670 | return (content, urlh) |
671 | ||
c9a77969 YCH |
672 | @staticmethod |
673 | def _guess_encoding_from_content(content_type, webpage_bytes): | |
d6983cb4 PH |
674 | m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type) |
675 | if m: | |
676 | encoding = m.group(1) | |
677 | else: | |
0d75ae2c | 678 | m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]', |
f143d86a PH |
679 | webpage_bytes[:1024]) |
680 | if m: | |
681 | encoding = m.group(1).decode('ascii') | |
b60016e8 PH |
682 | elif webpage_bytes.startswith(b'\xff\xfe'): |
683 | encoding = 'utf-16' | |
f143d86a PH |
684 | else: |
685 | encoding = 'utf-8' | |
c9a77969 YCH |
686 | |
687 | return encoding | |
688 | ||
4457823d S |
689 | def __check_blocked(self, content): |
690 | first_block = content[:512] | |
3089bc74 S |
691 | if ('<title>Access to this site is blocked</title>' in content |
692 | and 'Websense' in first_block): | |
4457823d S |
693 | msg = 'Access to this webpage has been blocked by Websense filtering software in your network.' |
694 | blocked_iframe = self._html_search_regex( | |
695 | r'<iframe src="([^"]+)"', content, | |
696 | 'Websense information URL', default=None) | |
697 | if blocked_iframe: | |
698 | msg += ' Visit %s for more details' % blocked_iframe | |
699 | raise ExtractorError(msg, expected=True) | |
700 | if '<title>The URL you requested has been blocked</title>' in first_block: | |
701 | msg = ( | |
702 | 'Access to this webpage has been blocked by Indian censorship. ' | |
703 | 'Use a VPN or proxy server (with --proxy) to route around it.') | |
704 | block_msg = self._html_search_regex( | |
705 | r'</h1><p>(.*?)</p>', | |
706 | content, 'block message', default=None) | |
707 | if block_msg: | |
708 | msg += ' (Message: "%s")' % block_msg.replace('\n', ' ') | |
709 | raise ExtractorError(msg, expected=True) | |
3089bc74 S |
710 | if ('<title>TTK :: Доступ к ресурсу ограничен</title>' in content |
711 | and 'blocklist.rkn.gov.ru' in content): | |
4457823d S |
712 | raise ExtractorError( |
713 | 'Access to this webpage has been blocked by decision of the Russian government. ' | |
714 | 'Visit http://blocklist.rkn.gov.ru/ for a block reason.', | |
715 | expected=True) | |
716 | ||
c9a77969 YCH |
717 | def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None): |
718 | content_type = urlh.headers.get('Content-Type', '') | |
719 | webpage_bytes = urlh.read() | |
720 | if prefix is not None: | |
721 | webpage_bytes = prefix + webpage_bytes | |
722 | if not encoding: | |
723 | encoding = self._guess_encoding_from_content(content_type, webpage_bytes) | |
d6983cb4 | 724 | if self._downloader.params.get('dump_intermediate_pages', False): |
f610dbb0 | 725 | self.to_screen('Dumping request to ' + urlh.geturl()) |
d6983cb4 PH |
726 | dump = base64.b64encode(webpage_bytes).decode('ascii') |
727 | self._downloader.to_screen(dump) | |
d41e6efc | 728 | if self._downloader.params.get('write_pages', False): |
f610dbb0 | 729 | basen = '%s_%s' % (video_id, urlh.geturl()) |
c1bce22f | 730 | if len(basen) > 240: |
f1a9d64e | 731 | h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest() |
c1bce22f PH |
732 | basen = basen[:240 - len(h)] + h |
733 | raw_filename = basen + '.dump' | |
d41e6efc | 734 | filename = sanitize_filename(raw_filename, restricted=True) |
f1a9d64e | 735 | self.to_screen('Saving request to ' + filename) |
5f58165d S |
736 | # Working around MAX_PATH limitation on Windows (see |
737 | # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx) | |
e9c0cdd3 | 738 | if compat_os_name == 'nt': |
5f58165d S |
739 | absfilepath = os.path.abspath(filename) |
740 | if len(absfilepath) > 259: | |
741 | filename = '\\\\?\\' + absfilepath | |
d41e6efc PH |
742 | with open(filename, 'wb') as outf: |
743 | outf.write(webpage_bytes) | |
744 | ||
ec0fafbb AA |
745 | try: |
746 | content = webpage_bytes.decode(encoding, 'replace') | |
747 | except LookupError: | |
748 | content = webpage_bytes.decode('utf-8', 'replace') | |
2410c43d | 749 | |
4457823d | 750 | self.__check_blocked(content) |
2410c43d | 751 | |
23be51d8 | 752 | return content |
d6983cb4 | 753 | |
d391b7e2 S |
754 | def _download_webpage( |
755 | self, url_or_request, video_id, note=None, errnote=None, | |
756 | fatal=True, tries=1, timeout=5, encoding=None, data=None, | |
757 | headers={}, query={}, expected_status=None): | |
758 | """ | |
759 | Return the data of the page as a string. | |
760 | ||
761 | Arguments: | |
762 | url_or_request -- plain text URL as a string or | |
763 | a compat_urllib_request.Requestobject | |
764 | video_id -- Video/playlist/item identifier (string) | |
765 | ||
766 | Keyword arguments: | |
767 | note -- note printed before downloading (string) | |
768 | errnote -- note printed in case of an error (string) | |
769 | fatal -- flag denoting whether error should be considered fatal, | |
770 | i.e. whether it should cause ExtractionError to be raised, | |
771 | otherwise a warning will be reported and extraction continued | |
772 | tries -- number of tries | |
773 | timeout -- sleep interval between tries | |
774 | encoding -- encoding for a page content decoding, guessed automatically | |
775 | when not explicitly specified | |
776 | data -- POST data (bytes) | |
777 | headers -- HTTP headers (dict) | |
778 | query -- URL query (dict) | |
779 | expected_status -- allows to accept failed HTTP requests (non 2xx | |
780 | status code) by explicitly specifying a set of accepted status | |
781 | codes. Can be any of the following entities: | |
782 | - an integer type specifying an exact failed status code to | |
783 | accept | |
784 | - a list or a tuple of integer types specifying a list of | |
785 | failed status codes to accept | |
786 | - a callable accepting an actual failed status code and | |
787 | returning True if it should be accepted | |
788 | Note that this argument does not affect success status codes (2xx) | |
789 | which are always accepted. | |
790 | """ | |
791 | ||
995ad69c TF |
792 | success = False |
793 | try_count = 0 | |
794 | while success is False: | |
795 | try: | |
d391b7e2 S |
796 | res = self._download_webpage_handle( |
797 | url_or_request, video_id, note, errnote, fatal, | |
798 | encoding=encoding, data=data, headers=headers, query=query, | |
799 | expected_status=expected_status) | |
995ad69c TF |
800 | success = True |
801 | except compat_http_client.IncompleteRead as e: | |
802 | try_count += 1 | |
803 | if try_count >= tries: | |
804 | raise e | |
805 | self._sleep(timeout, video_id) | |
7cc3570e PH |
806 | if res is False: |
807 | return res | |
808 | else: | |
809 | content, _ = res | |
810 | return content | |
d6983cb4 | 811 | |
e0d198c1 S |
812 | def _download_xml_handle( |
813 | self, url_or_request, video_id, note='Downloading XML', | |
814 | errnote='Unable to download XML', transform_source=None, | |
d391b7e2 S |
815 | fatal=True, encoding=None, data=None, headers={}, query={}, |
816 | expected_status=None): | |
817 | """ | |
ee0ba927 | 818 | Return a tuple (xml as an compat_etree_Element, URL handle). |
d391b7e2 S |
819 | |
820 | See _download_webpage docstring for arguments specification. | |
821 | """ | |
e0d198c1 S |
822 | res = self._download_webpage_handle( |
823 | url_or_request, video_id, note, errnote, fatal=fatal, | |
d391b7e2 S |
824 | encoding=encoding, data=data, headers=headers, query=query, |
825 | expected_status=expected_status) | |
e0d198c1 S |
826 | if res is False: |
827 | return res | |
828 | xml_string, urlh = res | |
829 | return self._parse_xml( | |
830 | xml_string, video_id, transform_source=transform_source, | |
831 | fatal=fatal), urlh | |
832 | ||
d391b7e2 S |
833 | def _download_xml( |
834 | self, url_or_request, video_id, | |
835 | note='Downloading XML', errnote='Unable to download XML', | |
836 | transform_source=None, fatal=True, encoding=None, | |
837 | data=None, headers={}, query={}, expected_status=None): | |
838 | """ | |
ee0ba927 | 839 | Return the xml as an compat_etree_Element. |
d391b7e2 S |
840 | |
841 | See _download_webpage docstring for arguments specification. | |
842 | """ | |
e0d198c1 S |
843 | res = self._download_xml_handle( |
844 | url_or_request, video_id, note=note, errnote=errnote, | |
845 | transform_source=transform_source, fatal=fatal, encoding=encoding, | |
d391b7e2 S |
846 | data=data, headers=headers, query=query, |
847 | expected_status=expected_status) | |
e0d198c1 | 848 | return res if res is False else res[0] |
e01c3d2e S |
849 | |
850 | def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True): | |
e2b38da9 PH |
851 | if transform_source: |
852 | xml_string = transform_source(xml_string) | |
e01c3d2e S |
853 | try: |
854 | return compat_etree_fromstring(xml_string.encode('utf-8')) | |
855 | except compat_xml_parse_error as ve: | |
856 | errmsg = '%s: Failed to parse XML ' % video_id | |
857 | if fatal: | |
858 | raise ExtractorError(errmsg, cause=ve) | |
859 | else: | |
860 | self.report_warning(errmsg + str(ve)) | |
267ed0c5 | 861 | |
0fe7783e S |
862 | def _download_json_handle( |
863 | self, url_or_request, video_id, note='Downloading JSON metadata', | |
864 | errnote='Unable to download JSON metadata', transform_source=None, | |
d391b7e2 S |
865 | fatal=True, encoding=None, data=None, headers={}, query={}, |
866 | expected_status=None): | |
867 | """ | |
868 | Return a tuple (JSON object, URL handle). | |
869 | ||
870 | See _download_webpage docstring for arguments specification. | |
871 | """ | |
0fe7783e | 872 | res = self._download_webpage_handle( |
c9a77969 | 873 | url_or_request, video_id, note, errnote, fatal=fatal, |
d391b7e2 S |
874 | encoding=encoding, data=data, headers=headers, query=query, |
875 | expected_status=expected_status) | |
0fe7783e S |
876 | if res is False: |
877 | return res | |
878 | json_string, urlh = res | |
ebb64199 | 879 | return self._parse_json( |
0fe7783e S |
880 | json_string, video_id, transform_source=transform_source, |
881 | fatal=fatal), urlh | |
882 | ||
883 | def _download_json( | |
884 | self, url_or_request, video_id, note='Downloading JSON metadata', | |
885 | errnote='Unable to download JSON metadata', transform_source=None, | |
d391b7e2 S |
886 | fatal=True, encoding=None, data=None, headers={}, query={}, |
887 | expected_status=None): | |
888 | """ | |
889 | Return the JSON object as a dict. | |
890 | ||
891 | See _download_webpage docstring for arguments specification. | |
892 | """ | |
0fe7783e S |
893 | res = self._download_json_handle( |
894 | url_or_request, video_id, note=note, errnote=errnote, | |
895 | transform_source=transform_source, fatal=fatal, encoding=encoding, | |
d391b7e2 S |
896 | data=data, headers=headers, query=query, |
897 | expected_status=expected_status) | |
0fe7783e | 898 | return res if res is False else res[0] |
ebb64199 TF |
899 | |
900 | def _parse_json(self, json_string, video_id, transform_source=None, fatal=True): | |
81c2f20b PH |
901 | if transform_source: |
902 | json_string = transform_source(json_string) | |
3d3538e4 PH |
903 | try: |
904 | return json.loads(json_string) | |
905 | except ValueError as ve: | |
e7b6d122 PH |
906 | errmsg = '%s: Failed to parse JSON ' % video_id |
907 | if fatal: | |
908 | raise ExtractorError(errmsg, cause=ve) | |
909 | else: | |
910 | self.report_warning(errmsg + str(ve)) | |
3d3538e4 | 911 | |
f45f96f8 | 912 | def report_warning(self, msg, video_id=None): |
f1a9d64e | 913 | idstr = '' if video_id is None else '%s: ' % video_id |
f45f96f8 | 914 | self._downloader.report_warning( |
f1a9d64e | 915 | '[%s] %s%s' % (self.IE_NAME, idstr, msg)) |
f45f96f8 | 916 | |
d6983cb4 PH |
917 | def to_screen(self, msg): |
918 | """Print msg to screen, prefixing it with '[ie_name]'""" | |
f1a9d64e | 919 | self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg)) |
d6983cb4 PH |
920 | |
921 | def report_extraction(self, id_or_name): | |
922 | """Report information extraction.""" | |
f1a9d64e | 923 | self.to_screen('%s: Extracting information' % id_or_name) |
d6983cb4 PH |
924 | |
925 | def report_download_webpage(self, video_id): | |
926 | """Report webpage download.""" | |
f1a9d64e | 927 | self.to_screen('%s: Downloading webpage' % video_id) |
d6983cb4 PH |
928 | |
929 | def report_age_confirmation(self): | |
930 | """Report attempt to confirm age.""" | |
f1a9d64e | 931 | self.to_screen('Confirming age') |
d6983cb4 | 932 | |
fc79158d JMF |
933 | def report_login(self): |
934 | """Report attempt to log in.""" | |
f1a9d64e | 935 | self.to_screen('Logging in') |
fc79158d | 936 | |
43e7d3c9 S |
937 | @staticmethod |
938 | def raise_login_required(msg='This video is only available for registered users'): | |
939 | raise ExtractorError( | |
940 | '%s. Use --username and --password or --netrc to provide account credentials.' % msg, | |
941 | expected=True) | |
942 | ||
c430802e | 943 | @staticmethod |
773f291d S |
944 | def raise_geo_restricted(msg='This video is not available from your location due to geo restriction', countries=None): |
945 | raise GeoRestrictedError(msg, countries=countries) | |
c430802e | 946 | |
5f6a1245 | 947 | # Methods for following #608 |
c0d0b01f | 948 | @staticmethod |
830d53bf | 949 | def url_result(url, ie=None, video_id=None, video_title=None): |
10952eb2 | 950 | """Returns a URL that points to a page that should be processed""" |
5f6a1245 | 951 | # TODO: ie should be the class used for getting the info |
d6983cb4 PH |
952 | video_info = {'_type': 'url', |
953 | 'url': url, | |
954 | 'ie_key': ie} | |
7012b23c PH |
955 | if video_id is not None: |
956 | video_info['id'] = video_id | |
830d53bf S |
957 | if video_title is not None: |
958 | video_info['title'] = video_title | |
d6983cb4 | 959 | return video_info |
5f6a1245 | 960 | |
749ca5ec S |
961 | def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None): |
962 | urls = orderedSet( | |
46b18f23 JH |
963 | self.url_result(self._proto_relative_url(getter(m) if getter else m), ie) |
964 | for m in matches) | |
965 | return self.playlist_result( | |
749ca5ec | 966 | urls, playlist_id=playlist_id, playlist_title=playlist_title) |
46b18f23 | 967 | |
c0d0b01f | 968 | @staticmethod |
acf5cbfe | 969 | def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None): |
d6983cb4 PH |
970 | """Returns a playlist""" |
971 | video_info = {'_type': 'playlist', | |
972 | 'entries': entries} | |
973 | if playlist_id: | |
974 | video_info['id'] = playlist_id | |
975 | if playlist_title: | |
976 | video_info['title'] = playlist_title | |
acf5cbfe S |
977 | if playlist_description: |
978 | video_info['description'] = playlist_description | |
d6983cb4 PH |
979 | return video_info |
980 | ||
c342041f | 981 | def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
982 | """ |
983 | Perform a regex search on the given string, using a single or a list of | |
984 | patterns returning the first matching group. | |
985 | In case of failure return a default value or raise a WARNING or a | |
55b3e45b | 986 | RegexNotFoundError, depending on fatal, specifying the field name. |
d6983cb4 PH |
987 | """ |
988 | if isinstance(pattern, (str, compat_str, compiled_regex_type)): | |
989 | mobj = re.search(pattern, string, flags) | |
990 | else: | |
991 | for p in pattern: | |
992 | mobj = re.search(p, string, flags) | |
c3415d1b PH |
993 | if mobj: |
994 | break | |
d6983cb4 | 995 | |
e9c0cdd3 | 996 | if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty(): |
f1a9d64e | 997 | _name = '\033[0;34m%s\033[0m' % name |
d6983cb4 PH |
998 | else: |
999 | _name = name | |
1000 | ||
1001 | if mobj: | |
711ede6e PH |
1002 | if group is None: |
1003 | # return the first matching group | |
1004 | return next(g for g in mobj.groups() if g is not None) | |
1005 | else: | |
1006 | return mobj.group(group) | |
c342041f | 1007 | elif default is not NO_DEFAULT: |
d6983cb4 PH |
1008 | return default |
1009 | elif fatal: | |
f1a9d64e | 1010 | raise RegexNotFoundError('Unable to extract %s' % _name) |
d6983cb4 | 1011 | else: |
08f2a92c | 1012 | self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) |
d6983cb4 PH |
1013 | return None |
1014 | ||
c342041f | 1015 | def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
1016 | """ |
1017 | Like _search_regex, but strips HTML tags and unescapes entities. | |
1018 | """ | |
711ede6e | 1019 | res = self._search_regex(pattern, string, name, default, fatal, flags, group) |
d6983cb4 PH |
1020 | if res: |
1021 | return clean_html(res).strip() | |
1022 | else: | |
1023 | return res | |
1024 | ||
2118fdd1 RA |
1025 | def _get_netrc_login_info(self, netrc_machine=None): |
1026 | username = None | |
1027 | password = None | |
1028 | netrc_machine = netrc_machine or self._NETRC_MACHINE | |
1029 | ||
1030 | if self._downloader.params.get('usenetrc', False): | |
1031 | try: | |
1032 | info = netrc.netrc().authenticators(netrc_machine) | |
1033 | if info is not None: | |
1034 | username = info[0] | |
1035 | password = info[2] | |
1036 | else: | |
dcce092e S |
1037 | raise netrc.NetrcParseError( |
1038 | 'No authenticators for %s' % netrc_machine) | |
2118fdd1 | 1039 | except (IOError, netrc.NetrcParseError) as err: |
dcce092e S |
1040 | self._downloader.report_warning( |
1041 | 'parsing .netrc: %s' % error_to_compat_str(err)) | |
2118fdd1 | 1042 | |
dcce092e | 1043 | return username, password |
2118fdd1 | 1044 | |
1b6712ab | 1045 | def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None): |
fc79158d | 1046 | """ |
cf0649f8 | 1047 | Get the login info as (username, password) |
32443dd3 S |
1048 | First look for the manually specified credentials using username_option |
1049 | and password_option as keys in params dictionary. If no such credentials | |
1050 | available look in the netrc file using the netrc_machine or _NETRC_MACHINE | |
1051 | value. | |
fc79158d JMF |
1052 | If there's no info available, return (None, None) |
1053 | """ | |
1054 | if self._downloader is None: | |
1055 | return (None, None) | |
1056 | ||
fc79158d JMF |
1057 | downloader_params = self._downloader.params |
1058 | ||
1059 | # Attempt to use provided username and password or .netrc data | |
1b6712ab RA |
1060 | if downloader_params.get(username_option) is not None: |
1061 | username = downloader_params[username_option] | |
1062 | password = downloader_params[password_option] | |
2118fdd1 | 1063 | else: |
1b6712ab | 1064 | username, password = self._get_netrc_login_info(netrc_machine) |
5f6a1245 | 1065 | |
2133565c | 1066 | return username, password |
fc79158d | 1067 | |
e64b7569 | 1068 | def _get_tfa_info(self, note='two-factor verification code'): |
83317f69 | 1069 | """ |
1070 | Get the two-factor authentication info | |
1071 | TODO - asking the user will be required for sms/phone verify | |
1072 | currently just uses the command line option | |
1073 | If there's no info available, return None | |
1074 | """ | |
1075 | if self._downloader is None: | |
83317f69 | 1076 | return None |
1077 | downloader_params = self._downloader.params | |
1078 | ||
d800609c | 1079 | if downloader_params.get('twofactor') is not None: |
83317f69 | 1080 | return downloader_params['twofactor'] |
1081 | ||
e64b7569 | 1082 | return compat_getpass('Type %s and press [Return]: ' % note) |
83317f69 | 1083 | |
46720279 JMF |
1084 | # Helper functions for extracting OpenGraph info |
1085 | @staticmethod | |
ab2d5247 | 1086 | def _og_regexes(prop): |
448ef1f3 | 1087 | content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))' |
22f5f5c6 | 1088 | property_re = (r'(?:name|property)=(?:\'og[:-]%(prop)s\'|"og[:-]%(prop)s"|\s*og[:-]%(prop)s\b)' |
7a6d76a6 | 1089 | % {'prop': re.escape(prop)}) |
78fb87b2 | 1090 | template = r'<meta[^>]+?%s[^>]+?%s' |
ab2d5247 | 1091 | return [ |
78fb87b2 JMF |
1092 | template % (property_re, content_re), |
1093 | template % (content_re, property_re), | |
ab2d5247 | 1094 | ] |
46720279 | 1095 | |
864f24bd S |
1096 | @staticmethod |
1097 | def _meta_regex(prop): | |
1098 | return r'''(?isx)<meta | |
8b9848ac | 1099 | (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1) |
864f24bd S |
1100 | [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop) |
1101 | ||
3c4e6d83 | 1102 | def _og_search_property(self, prop, html, name=None, **kargs): |
b070564e S |
1103 | if not isinstance(prop, (list, tuple)): |
1104 | prop = [prop] | |
46720279 | 1105 | if name is None: |
b070564e S |
1106 | name = 'OpenGraph %s' % prop[0] |
1107 | og_regexes = [] | |
1108 | for p in prop: | |
1109 | og_regexes.extend(self._og_regexes(p)) | |
1110 | escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs) | |
eb0a8398 PH |
1111 | if escaped is None: |
1112 | return None | |
1113 | return unescapeHTML(escaped) | |
46720279 JMF |
1114 | |
1115 | def _og_search_thumbnail(self, html, **kargs): | |
10952eb2 | 1116 | return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs) |
46720279 JMF |
1117 | |
1118 | def _og_search_description(self, html, **kargs): | |
1119 | return self._og_search_property('description', html, fatal=False, **kargs) | |
1120 | ||
1121 | def _og_search_title(self, html, **kargs): | |
1122 | return self._og_search_property('title', html, **kargs) | |
1123 | ||
8ffa13e0 | 1124 | def _og_search_video_url(self, html, name='video url', secure=True, **kargs): |
a3681973 PH |
1125 | regexes = self._og_regexes('video') + self._og_regexes('video:url') |
1126 | if secure: | |
1127 | regexes = self._og_regexes('video:secure_url') + regexes | |
8ffa13e0 | 1128 | return self._html_search_regex(regexes, html, name, **kargs) |
46720279 | 1129 | |
78338f71 JMF |
1130 | def _og_search_url(self, html, **kargs): |
1131 | return self._og_search_property('url', html, **kargs) | |
1132 | ||
40c696e5 | 1133 | def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs): |
88d9f6c0 S |
1134 | if not isinstance(name, (list, tuple)): |
1135 | name = [name] | |
59040888 | 1136 | if display_name is None: |
88d9f6c0 | 1137 | display_name = name[0] |
59040888 | 1138 | return self._html_search_regex( |
88d9f6c0 | 1139 | [self._meta_regex(n) for n in name], |
711ede6e | 1140 | html, display_name, fatal=fatal, group='content', **kwargs) |
59040888 PH |
1141 | |
1142 | def _dc_search_uploader(self, html): | |
1143 | return self._html_search_meta('dc.creator', html, 'uploader') | |
1144 | ||
8dbe9899 PH |
1145 | def _rta_search(self, html): |
1146 | # See http://www.rtalabel.org/index.php?content=howtofaq#single | |
1147 | if re.search(r'(?ix)<meta\s+name="rating"\s+' | |
1148 | r' content="RTA-5042-1996-1400-1577-RTA"', | |
1149 | html): | |
1150 | return 18 | |
1151 | return 0 | |
1152 | ||
59040888 PH |
1153 | def _media_rating_search(self, html): |
1154 | # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/ | |
1155 | rating = self._html_search_meta('rating', html) | |
1156 | ||
1157 | if not rating: | |
1158 | return None | |
1159 | ||
1160 | RATING_TABLE = { | |
1161 | 'safe for kids': 0, | |
1162 | 'general': 8, | |
1163 | '14 years': 14, | |
1164 | 'mature': 17, | |
1165 | 'restricted': 19, | |
1166 | } | |
d800609c | 1167 | return RATING_TABLE.get(rating.lower()) |
59040888 | 1168 | |
69319969 | 1169 | def _family_friendly_search(self, html): |
6ca7732d | 1170 | # See http://schema.org/VideoObject |
ac8491fc S |
1171 | family_friendly = self._html_search_meta( |
1172 | 'isFamilyFriendly', html, default=None) | |
69319969 NJ |
1173 | |
1174 | if not family_friendly: | |
1175 | return None | |
1176 | ||
1177 | RATING_TABLE = { | |
1178 | '1': 0, | |
1179 | 'true': 0, | |
1180 | '0': 18, | |
1181 | 'false': 18, | |
1182 | } | |
d800609c | 1183 | return RATING_TABLE.get(family_friendly.lower()) |
69319969 | 1184 | |
0c708f11 JMF |
1185 | def _twitter_search_player(self, html): |
1186 | return self._html_search_meta('twitter:player', html, | |
9e1a5b84 | 1187 | 'twitter card player') |
0c708f11 | 1188 | |
95b31e26 | 1189 | def _search_json_ld(self, html, video_id, expected_type=None, **kwargs): |
4433bb02 | 1190 | json_ld_list = list(re.finditer(JSON_LD_RE, html)) |
321b5e08 | 1191 | default = kwargs.get('default', NO_DEFAULT) |
321b5e08 S |
1192 | # JSON-LD may be malformed and thus `fatal` should be respected. |
1193 | # At the same time `default` may be passed that assumes `fatal=False` | |
1194 | # for _search_regex. Let's simulate the same behavior here as well. | |
1195 | fatal = kwargs.get('fatal', True) if default == NO_DEFAULT else False | |
4433bb02 S |
1196 | json_ld = [] |
1197 | for mobj in json_ld_list: | |
1198 | json_ld_item = self._parse_json( | |
1199 | mobj.group('json_ld'), video_id, fatal=fatal) | |
1200 | if not json_ld_item: | |
1201 | continue | |
1202 | if isinstance(json_ld_item, dict): | |
1203 | json_ld.append(json_ld_item) | |
1204 | elif isinstance(json_ld_item, (list, tuple)): | |
1205 | json_ld.extend(json_ld_item) | |
1206 | if json_ld: | |
1207 | json_ld = self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type) | |
1208 | if json_ld: | |
1209 | return json_ld | |
1210 | if default is not NO_DEFAULT: | |
1211 | return default | |
1212 | elif fatal: | |
1213 | raise RegexNotFoundError('Unable to extract JSON-LD') | |
1214 | else: | |
1215 | self._downloader.report_warning('unable to extract JSON-LD %s' % bug_reports_message()) | |
1216 | return {} | |
4ca2a3cf | 1217 | |
95b31e26 | 1218 | def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): |
4ca2a3cf S |
1219 | if isinstance(json_ld, compat_str): |
1220 | json_ld = self._parse_json(json_ld, video_id, fatal=fatal) | |
1221 | if not json_ld: | |
1222 | return {} | |
1223 | info = {} | |
46933a15 S |
1224 | if not isinstance(json_ld, (list, tuple, dict)): |
1225 | return info | |
1226 | if isinstance(json_ld, dict): | |
1227 | json_ld = [json_ld] | |
bae14048 | 1228 | |
e7e4a6e0 S |
1229 | INTERACTION_TYPE_MAP = { |
1230 | 'CommentAction': 'comment', | |
1231 | 'AgreeAction': 'like', | |
1232 | 'DisagreeAction': 'dislike', | |
1233 | 'LikeAction': 'like', | |
1234 | 'DislikeAction': 'dislike', | |
1235 | 'ListenAction': 'view', | |
1236 | 'WatchAction': 'view', | |
1237 | 'ViewAction': 'view', | |
1238 | } | |
1239 | ||
1240 | def extract_interaction_statistic(e): | |
1241 | interaction_statistic = e.get('interactionStatistic') | |
1242 | if not isinstance(interaction_statistic, list): | |
1243 | return | |
1244 | for is_e in interaction_statistic: | |
1245 | if not isinstance(is_e, dict): | |
1246 | continue | |
1247 | if is_e.get('@type') != 'InteractionCounter': | |
1248 | continue | |
1249 | interaction_type = is_e.get('interactionType') | |
1250 | if not isinstance(interaction_type, compat_str): | |
1251 | continue | |
ce5b9040 S |
1252 | # For interaction count some sites provide string instead of |
1253 | # an integer (as per spec) with non digit characters (e.g. ",") | |
1254 | # so extracting count with more relaxed str_to_int | |
1255 | interaction_count = str_to_int(is_e.get('userInteractionCount')) | |
e7e4a6e0 S |
1256 | if interaction_count is None: |
1257 | continue | |
1258 | count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1]) | |
1259 | if not count_kind: | |
1260 | continue | |
1261 | count_key = '%s_count' % count_kind | |
1262 | if info.get(count_key) is not None: | |
1263 | continue | |
1264 | info[count_key] = interaction_count | |
1265 | ||
bae14048 S |
1266 | def extract_video_object(e): |
1267 | assert e['@type'] == 'VideoObject' | |
1268 | info.update({ | |
bebef109 | 1269 | 'url': url_or_none(e.get('contentUrl')), |
bae14048 S |
1270 | 'title': unescapeHTML(e.get('name')), |
1271 | 'description': unescapeHTML(e.get('description')), | |
bebef109 | 1272 | 'thumbnail': url_or_none(e.get('thumbnailUrl') or e.get('thumbnailURL')), |
bae14048 S |
1273 | 'duration': parse_duration(e.get('duration')), |
1274 | 'timestamp': unified_timestamp(e.get('uploadDate')), | |
ad06b99d | 1275 | 'uploader': str_or_none(e.get('author')), |
bae14048 S |
1276 | 'filesize': float_or_none(e.get('contentSize')), |
1277 | 'tbr': int_or_none(e.get('bitrate')), | |
1278 | 'width': int_or_none(e.get('width')), | |
1279 | 'height': int_or_none(e.get('height')), | |
33a81c2c | 1280 | 'view_count': int_or_none(e.get('interactionCount')), |
bae14048 | 1281 | }) |
e7e4a6e0 | 1282 | extract_interaction_statistic(e) |
bae14048 | 1283 | |
46933a15 | 1284 | for e in json_ld: |
4433bb02 | 1285 | if '@context' in e: |
46933a15 S |
1286 | item_type = e.get('@type') |
1287 | if expected_type is not None and expected_type != item_type: | |
4433bb02 | 1288 | continue |
c69701c6 | 1289 | if item_type in ('TVEpisode', 'Episode'): |
440863ad | 1290 | episode_name = unescapeHTML(e.get('name')) |
46933a15 | 1291 | info.update({ |
440863ad | 1292 | 'episode': episode_name, |
46933a15 S |
1293 | 'episode_number': int_or_none(e.get('episodeNumber')), |
1294 | 'description': unescapeHTML(e.get('description')), | |
1295 | }) | |
440863ad S |
1296 | if not info.get('title') and episode_name: |
1297 | info['title'] = episode_name | |
46933a15 | 1298 | part_of_season = e.get('partOfSeason') |
c69701c6 | 1299 | if isinstance(part_of_season, dict) and part_of_season.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'): |
458fd30f S |
1300 | info.update({ |
1301 | 'season': unescapeHTML(part_of_season.get('name')), | |
1302 | 'season_number': int_or_none(part_of_season.get('seasonNumber')), | |
1303 | }) | |
d16b3c66 | 1304 | part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries') |
c69701c6 | 1305 | if isinstance(part_of_series, dict) and part_of_series.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'): |
46933a15 | 1306 | info['series'] = unescapeHTML(part_of_series.get('name')) |
391256dc S |
1307 | elif item_type == 'Movie': |
1308 | info.update({ | |
1309 | 'title': unescapeHTML(e.get('name')), | |
1310 | 'description': unescapeHTML(e.get('description')), | |
1311 | 'duration': parse_duration(e.get('duration')), | |
1312 | 'timestamp': unified_timestamp(e.get('dateCreated')), | |
1313 | }) | |
3931b845 | 1314 | elif item_type in ('Article', 'NewsArticle'): |
46933a15 S |
1315 | info.update({ |
1316 | 'timestamp': parse_iso8601(e.get('datePublished')), | |
1317 | 'title': unescapeHTML(e.get('headline')), | |
1318 | 'description': unescapeHTML(e.get('articleBody')), | |
1319 | }) | |
1320 | elif item_type == 'VideoObject': | |
bae14048 | 1321 | extract_video_object(e) |
4433bb02 S |
1322 | if expected_type is None: |
1323 | continue | |
1324 | else: | |
1325 | break | |
c69701c6 S |
1326 | video = e.get('video') |
1327 | if isinstance(video, dict) and video.get('@type') == 'VideoObject': | |
1328 | extract_video_object(video) | |
4433bb02 S |
1329 | if expected_type is None: |
1330 | continue | |
1331 | else: | |
1332 | break | |
4ca2a3cf S |
1333 | return dict((k, v) for k, v in info.items() if v is not None) |
1334 | ||
27713812 | 1335 | @staticmethod |
f8da79f8 | 1336 | def _hidden_inputs(html): |
586f1cc5 | 1337 | html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html) |
201ea3ee | 1338 | hidden_inputs = {} |
c8498368 S |
1339 | for input in re.findall(r'(?i)(<input[^>]+>)', html): |
1340 | attrs = extract_attributes(input) | |
1341 | if not input: | |
201ea3ee | 1342 | continue |
c8498368 | 1343 | if attrs.get('type') not in ('hidden', 'submit'): |
201ea3ee | 1344 | continue |
c8498368 S |
1345 | name = attrs.get('name') or attrs.get('id') |
1346 | value = attrs.get('value') | |
1347 | if name and value is not None: | |
1348 | hidden_inputs[name] = value | |
201ea3ee | 1349 | return hidden_inputs |
27713812 | 1350 | |
cf61d96d S |
1351 | def _form_hidden_inputs(self, form_id, html): |
1352 | form = self._search_regex( | |
73eb13df | 1353 | r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id, |
cf61d96d S |
1354 | html, '%s form' % form_id, group='form') |
1355 | return self._hidden_inputs(form) | |
1356 | ||
3ded7bac | 1357 | def _sort_formats(self, formats, field_preference=None): |
7e8caf30 | 1358 | if not formats: |
f1a9d64e | 1359 | raise ExtractorError('No video formats found') |
7e8caf30 | 1360 | |
b0d21ded S |
1361 | for f in formats: |
1362 | # Automatically determine tbr when missing based on abr and vbr (improves | |
1363 | # formats sorting in some cases) | |
350cf045 | 1364 | if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None: |
b0d21ded S |
1365 | f['tbr'] = f['abr'] + f['vbr'] |
1366 | ||
4bcc7bd1 | 1367 | def _formats_key(f): |
e6812ac9 PH |
1368 | # TODO remove the following workaround |
1369 | from ..utils import determine_ext | |
1370 | if not f.get('ext') and 'url' in f: | |
1371 | f['ext'] = determine_ext(f['url']) | |
1372 | ||
3ded7bac | 1373 | if isinstance(field_preference, (list, tuple)): |
bf8dd790 S |
1374 | return tuple( |
1375 | f.get(field) | |
1376 | if f.get(field) is not None | |
1377 | else ('' if field == 'format_id' else -1) | |
1378 | for field in field_preference) | |
3ded7bac | 1379 | |
4bcc7bd1 PH |
1380 | preference = f.get('preference') |
1381 | if preference is None: | |
d497a201 | 1382 | preference = 0 |
4bcc7bd1 PH |
1383 | if f.get('ext') in ['f4f', 'f4m']: # Not yet supported |
1384 | preference -= 0.5 | |
1385 | ||
8b408545 RA |
1386 | protocol = f.get('protocol') or determine_protocol(f) |
1387 | proto_preference = 0 if protocol in ['http', 'https'] else (-0.5 if protocol == 'rtsp' else -0.1) | |
d497a201 | 1388 | |
4bcc7bd1 | 1389 | if f.get('vcodec') == 'none': # audio only |
dd867805 | 1390 | preference -= 50 |
4bcc7bd1 | 1391 | if self._downloader.params.get('prefer_free_formats'): |
f1a9d64e | 1392 | ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus'] |
4bcc7bd1 | 1393 | else: |
f1a9d64e | 1394 | ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a'] |
4bcc7bd1 PH |
1395 | ext_preference = 0 |
1396 | try: | |
1397 | audio_ext_preference = ORDER.index(f['ext']) | |
1398 | except ValueError: | |
1399 | audio_ext_preference = -1 | |
1400 | else: | |
dd867805 | 1401 | if f.get('acodec') == 'none': # video only |
1402 | preference -= 40 | |
4bcc7bd1 | 1403 | if self._downloader.params.get('prefer_free_formats'): |
f1a9d64e | 1404 | ORDER = ['flv', 'mp4', 'webm'] |
4bcc7bd1 | 1405 | else: |
f1a9d64e | 1406 | ORDER = ['webm', 'flv', 'mp4'] |
4bcc7bd1 PH |
1407 | try: |
1408 | ext_preference = ORDER.index(f['ext']) | |
1409 | except ValueError: | |
1410 | ext_preference = -1 | |
1411 | audio_ext_preference = 0 | |
1412 | ||
1413 | return ( | |
1414 | preference, | |
aff2f4f4 | 1415 | f.get('language_preference') if f.get('language_preference') is not None else -1, |
5d73273f | 1416 | f.get('quality') if f.get('quality') is not None else -1, |
9933b574 | 1417 | f.get('tbr') if f.get('tbr') is not None else -1, |
03cd72b0 | 1418 | f.get('filesize') if f.get('filesize') is not None else -1, |
4bcc7bd1 | 1419 | f.get('vbr') if f.get('vbr') is not None else -1, |
1a6373ef PH |
1420 | f.get('height') if f.get('height') is not None else -1, |
1421 | f.get('width') if f.get('width') is not None else -1, | |
d497a201 | 1422 | proto_preference, |
1e1896f2 | 1423 | ext_preference, |
4bcc7bd1 PH |
1424 | f.get('abr') if f.get('abr') is not None else -1, |
1425 | audio_ext_preference, | |
2c8e03d9 | 1426 | f.get('fps') if f.get('fps') is not None else -1, |
9732d77e | 1427 | f.get('filesize_approx') if f.get('filesize_approx') is not None else -1, |
c64ed2a3 | 1428 | f.get('source_preference') if f.get('source_preference') is not None else -1, |
74f72824 | 1429 | f.get('format_id') if f.get('format_id') is not None else '', |
4bcc7bd1 PH |
1430 | ) |
1431 | formats.sort(key=_formats_key) | |
59040888 | 1432 | |
96a53167 S |
1433 | def _check_formats(self, formats, video_id): |
1434 | if formats: | |
1435 | formats[:] = filter( | |
1436 | lambda f: self._is_valid_url( | |
1437 | f['url'], video_id, | |
1438 | item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'), | |
1439 | formats) | |
1440 | ||
f5bdb444 S |
1441 | @staticmethod |
1442 | def _remove_duplicate_formats(formats): | |
1443 | format_urls = set() | |
1444 | unique_formats = [] | |
1445 | for f in formats: | |
1446 | if f['url'] not in format_urls: | |
1447 | format_urls.add(f['url']) | |
1448 | unique_formats.append(f) | |
1449 | formats[:] = unique_formats | |
1450 | ||
45024183 | 1451 | def _is_valid_url(self, url, video_id, item='video', headers={}): |
2f0f6578 S |
1452 | url = self._proto_relative_url(url, scheme='http:') |
1453 | # For now assume non HTTP(S) URLs always valid | |
1454 | if not (url.startswith('http://') or url.startswith('https://')): | |
1455 | return True | |
96a53167 | 1456 | try: |
45024183 | 1457 | self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers) |
96a53167 | 1458 | return True |
8bdd16b4 | 1459 | except ExtractorError as e: |
25e911a9 | 1460 | self.to_screen( |
8bdd16b4 | 1461 | '%s: %s URL is invalid, skipping: %s' |
1462 | % (video_id, item, error_to_compat_str(e.cause))) | |
25e911a9 | 1463 | return False |
96a53167 | 1464 | |
20991253 | 1465 | def http_scheme(self): |
1ede5b24 | 1466 | """ Either "http:" or "https:", depending on the user's preferences """ |
20991253 PH |
1467 | return ( |
1468 | 'http:' | |
1469 | if self._downloader.params.get('prefer_insecure', False) | |
1470 | else 'https:') | |
1471 | ||
57c7411f PH |
1472 | def _proto_relative_url(self, url, scheme=None): |
1473 | if url is None: | |
1474 | return url | |
1475 | if url.startswith('//'): | |
1476 | if scheme is None: | |
1477 | scheme = self.http_scheme() | |
1478 | return scheme + url | |
1479 | else: | |
1480 | return url | |
1481 | ||
4094b6e3 PH |
1482 | def _sleep(self, timeout, video_id, msg_template=None): |
1483 | if msg_template is None: | |
f1a9d64e | 1484 | msg_template = '%(video_id)s: Waiting for %(timeout)s seconds' |
4094b6e3 PH |
1485 | msg = msg_template % {'video_id': video_id, 'timeout': timeout} |
1486 | self.to_screen(msg) | |
1487 | time.sleep(timeout) | |
1488 | ||
a38436e8 | 1489 | def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None, |
4de61310 | 1490 | transform_source=lambda s: fix_xml_ampersands(s).strip(), |
7360c06f | 1491 | fatal=True, m3u8_id=None, data=None, headers={}, query={}): |
f036a632 JMF |
1492 | manifest = self._download_xml( |
1493 | manifest_url, video_id, 'Downloading f4m manifest', | |
97f4aecf S |
1494 | 'Unable to download f4m manifest', |
1495 | # Some manifests may be malformed, e.g. prosiebensat1 generated manifests | |
067aa17e | 1496 | # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244) |
4de61310 | 1497 | transform_source=transform_source, |
7360c06f | 1498 | fatal=fatal, data=data, headers=headers, query=query) |
4de61310 S |
1499 | |
1500 | if manifest is False: | |
8d29e47f | 1501 | return [] |
31bb8d3f | 1502 | |
0fdbb332 S |
1503 | return self._parse_f4m_formats( |
1504 | manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id, | |
448bb5f3 | 1505 | transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id) |
0fdbb332 S |
1506 | |
1507 | def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None, | |
1508 | transform_source=lambda s: fix_xml_ampersands(s).strip(), | |
448bb5f3 | 1509 | fatal=True, m3u8_id=None): |
ee0ba927 | 1510 | if not isinstance(manifest, compat_etree_Element) and not fatal: |
d9eb580a S |
1511 | return [] |
1512 | ||
cefecac1 | 1513 | # currently youtube-dlc cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy |
fb72ec58 | 1514 | akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0') |
1515 | if akamai_pv is not None and ';' in akamai_pv.text: | |
1516 | playerVerificationChallenge = akamai_pv.text.split(';')[0] | |
1517 | if playerVerificationChallenge.strip() != '': | |
1518 | return [] | |
1519 | ||
31bb8d3f | 1520 | formats = [] |
7a47d07c | 1521 | manifest_version = '1.0' |
b2527359 | 1522 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media') |
34e48bed | 1523 | if not media_nodes: |
7a47d07c | 1524 | manifest_version = '2.0' |
34e48bed | 1525 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') |
b22ca762 | 1526 | # Remove unsupported DRM protected media from final formats |
067aa17e | 1527 | # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573). |
b22ca762 S |
1528 | media_nodes = remove_encrypted_media(media_nodes) |
1529 | if not media_nodes: | |
1530 | return formats | |
48107c19 S |
1531 | |
1532 | manifest_base_url = get_base_url(manifest) | |
0a5685b2 | 1533 | |
a6571f10 | 1534 | bootstrap_info = xpath_element( |
0a5685b2 YCH |
1535 | manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'], |
1536 | 'bootstrap info', default=None) | |
1537 | ||
edd6074c RA |
1538 | vcodec = None |
1539 | mime_type = xpath_text( | |
1540 | manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'], | |
1541 | 'base URL', default=None) | |
1542 | if mime_type and mime_type.startswith('audio/'): | |
1543 | vcodec = 'none' | |
1544 | ||
b2527359 | 1545 | for i, media_el in enumerate(media_nodes): |
77b8b4e6 S |
1546 | tbr = int_or_none(media_el.attrib.get('bitrate')) |
1547 | width = int_or_none(media_el.attrib.get('width')) | |
1548 | height = int_or_none(media_el.attrib.get('height')) | |
1549 | format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])) | |
448bb5f3 YCH |
1550 | # If <bootstrapInfo> is present, the specified f4m is a |
1551 | # stream-level manifest, and only set-level manifests may refer to | |
1552 | # external resources. See section 11.4 and section 4 of F4M spec | |
1553 | if bootstrap_info is None: | |
1554 | media_url = None | |
1555 | # @href is introduced in 2.0, see section 11.6 of F4M spec | |
1556 | if manifest_version == '2.0': | |
1557 | media_url = media_el.attrib.get('href') | |
1558 | if media_url is None: | |
1559 | media_url = media_el.attrib.get('url') | |
31c746e5 S |
1560 | if not media_url: |
1561 | continue | |
cc357c4d S |
1562 | manifest_url = ( |
1563 | media_url if media_url.startswith('http://') or media_url.startswith('https://') | |
48107c19 | 1564 | else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url)) |
70f0f5a8 S |
1565 | # If media_url is itself a f4m manifest do the recursive extraction |
1566 | # since bitrates in parent manifest (this one) and media_url manifest | |
1567 | # may differ leading to inability to resolve the format by requested | |
1568 | # bitrate in f4m downloader | |
240b6045 YCH |
1569 | ext = determine_ext(manifest_url) |
1570 | if ext == 'f4m': | |
77b8b4e6 | 1571 | f4m_formats = self._extract_f4m_formats( |
0fdbb332 | 1572 | manifest_url, video_id, preference=preference, f4m_id=f4m_id, |
77b8b4e6 S |
1573 | transform_source=transform_source, fatal=fatal) |
1574 | # Sometimes stream-level manifest contains single media entry that | |
1575 | # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player). | |
1576 | # At the same time parent's media entry in set-level manifest may | |
1577 | # contain it. We will copy it from parent in such cases. | |
1578 | if len(f4m_formats) == 1: | |
1579 | f = f4m_formats[0] | |
1580 | f.update({ | |
1581 | 'tbr': f.get('tbr') or tbr, | |
1582 | 'width': f.get('width') or width, | |
1583 | 'height': f.get('height') or height, | |
1584 | 'format_id': f.get('format_id') if not tbr else format_id, | |
edd6074c | 1585 | 'vcodec': vcodec, |
77b8b4e6 S |
1586 | }) |
1587 | formats.extend(f4m_formats) | |
70f0f5a8 | 1588 | continue |
240b6045 YCH |
1589 | elif ext == 'm3u8': |
1590 | formats.extend(self._extract_m3u8_formats( | |
1591 | manifest_url, video_id, 'mp4', preference=preference, | |
fac2af3c | 1592 | m3u8_id=m3u8_id, fatal=fatal)) |
240b6045 | 1593 | continue |
31bb8d3f | 1594 | formats.append({ |
77b8b4e6 | 1595 | 'format_id': format_id, |
31bb8d3f | 1596 | 'url': manifest_url, |
30d0b549 | 1597 | 'manifest_url': manifest_url, |
a6571f10 | 1598 | 'ext': 'flv' if bootstrap_info is not None else None, |
187ee66c | 1599 | 'protocol': 'f4m', |
b2527359 | 1600 | 'tbr': tbr, |
77b8b4e6 S |
1601 | 'width': width, |
1602 | 'height': height, | |
edd6074c | 1603 | 'vcodec': vcodec, |
60ca389c | 1604 | 'preference': preference, |
31bb8d3f | 1605 | }) |
31bb8d3f JMF |
1606 | return formats |
1607 | ||
16da9bbc YCH |
1608 | def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, m3u8_id=None): |
1609 | return { | |
f207019c | 1610 | 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])), |
704df56d PH |
1611 | 'url': m3u8_url, |
1612 | 'ext': ext, | |
1613 | 'protocol': 'm3u8', | |
37768f92 | 1614 | 'preference': preference - 100 if preference else -100, |
704df56d PH |
1615 | 'resolution': 'multiple', |
1616 | 'format_note': 'Quality selection URL', | |
16da9bbc YCH |
1617 | } |
1618 | ||
1619 | def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None, | |
1620 | entry_protocol='m3u8', preference=None, | |
1621 | m3u8_id=None, note=None, errnote=None, | |
7360c06f S |
1622 | fatal=True, live=False, data=None, headers={}, |
1623 | query={}): | |
dbd82a1d | 1624 | res = self._download_webpage_handle( |
81515ad9 | 1625 | m3u8_url, video_id, |
621ed9f5 | 1626 | note=note or 'Downloading m3u8 information', |
13af92fd | 1627 | errnote=errnote or 'Failed to download m3u8 information', |
7360c06f | 1628 | fatal=fatal, data=data, headers=headers, query=query) |
cb252080 | 1629 | |
dbd82a1d | 1630 | if res is False: |
8d29e47f | 1631 | return [] |
cb252080 | 1632 | |
dbd82a1d | 1633 | m3u8_doc, urlh = res |
37113045 | 1634 | m3u8_url = urlh.geturl() |
9cdffeeb | 1635 | |
cb252080 S |
1636 | return self._parse_m3u8_formats( |
1637 | m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol, | |
1638 | preference=preference, m3u8_id=m3u8_id, live=live) | |
1639 | ||
1640 | def _parse_m3u8_formats(self, m3u8_doc, m3u8_url, ext=None, | |
1641 | entry_protocol='m3u8', preference=None, | |
1642 | m3u8_id=None, live=False): | |
08a00eef RA |
1643 | if '#EXT-X-FAXS-CM:' in m3u8_doc: # Adobe Flash Access |
1644 | return [] | |
1645 | ||
ea229584 RA |
1646 | if re.search(r'#EXT-X-SESSION-KEY:.*?URI="skd://', m3u8_doc): # Apple FairPlay |
1647 | return [] | |
1648 | ||
ff99fe52 | 1649 | formats = [] |
0def7587 RA |
1650 | |
1651 | format_url = lambda u: ( | |
1652 | u | |
1653 | if re.match(r'^https?://', u) | |
1654 | else compat_urlparse.urljoin(m3u8_url, u)) | |
1655 | ||
cb252080 S |
1656 | # References: |
1657 | # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21 | |
067aa17e S |
1658 | # 2. https://github.com/ytdl-org/youtube-dl/issues/12211 |
1659 | # 3. https://github.com/ytdl-org/youtube-dl/issues/18923 | |
cb252080 S |
1660 | |
1661 | # We should try extracting formats only from master playlists [1, 4.3.4], | |
1662 | # i.e. playlists that describe available qualities. On the other hand | |
1663 | # media playlists [1, 4.3.3] should be returned as is since they contain | |
1664 | # just the media without qualities renditions. | |
9cdffeeb | 1665 | # Fortunately, master playlist can be easily distinguished from media |
cb252080 | 1666 | # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4] |
a0566bbf | 1667 | # master playlist tags MUST NOT appear in a media playlist and vice versa. |
cb252080 S |
1668 | # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every |
1669 | # media playlist and MUST NOT appear in master playlist thus we can | |
1670 | # clearly detect media playlist with this criterion. | |
1671 | ||
9cdffeeb | 1672 | if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is |
7f32e5dc | 1673 | return [{ |
1674 | 'url': m3u8_url, | |
1675 | 'format_id': m3u8_id, | |
1676 | 'ext': ext, | |
1677 | 'protocol': entry_protocol, | |
1678 | 'preference': preference, | |
1679 | }] | |
cb252080 S |
1680 | |
1681 | groups = {} | |
1682 | last_stream_inf = {} | |
1683 | ||
1684 | def extract_media(x_media_line): | |
1685 | media = parse_m3u8_attributes(x_media_line) | |
1686 | # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED | |
1687 | media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME') | |
1688 | if not (media_type and group_id and name): | |
1689 | return | |
1690 | groups.setdefault(group_id, []).append(media) | |
1691 | if media_type not in ('VIDEO', 'AUDIO'): | |
1692 | return | |
1693 | media_url = media.get('URI') | |
1694 | if media_url: | |
1695 | format_id = [] | |
9211e331 | 1696 | for v in (m3u8_id, group_id, name): |
cb252080 S |
1697 | if v: |
1698 | format_id.append(v) | |
1699 | f = { | |
1700 | 'format_id': '-'.join(format_id), | |
1701 | 'url': format_url(media_url), | |
c89b49f7 | 1702 | 'manifest_url': m3u8_url, |
cb252080 S |
1703 | 'language': media.get('LANGUAGE'), |
1704 | 'ext': ext, | |
1705 | 'protocol': entry_protocol, | |
1706 | 'preference': preference, | |
1707 | } | |
1708 | if media_type == 'AUDIO': | |
1709 | f['vcodec'] = 'none' | |
1710 | formats.append(f) | |
1711 | ||
1712 | def build_stream_name(): | |
1713 | # Despite specification does not mention NAME attribute for | |
3019cb0c S |
1714 | # EXT-X-STREAM-INF tag it still sometimes may be present (see [1] |
1715 | # or vidio test in TestInfoExtractor.test_parse_m3u8_formats) | |
ddd258f9 | 1716 | # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015 |
cb252080 S |
1717 | stream_name = last_stream_inf.get('NAME') |
1718 | if stream_name: | |
1719 | return stream_name | |
1720 | # If there is no NAME in EXT-X-STREAM-INF it will be obtained | |
1721 | # from corresponding rendition group | |
1722 | stream_group_id = last_stream_inf.get('VIDEO') | |
1723 | if not stream_group_id: | |
1724 | return | |
1725 | stream_group = groups.get(stream_group_id) | |
1726 | if not stream_group: | |
1727 | return stream_group_id | |
1728 | rendition = stream_group[0] | |
1729 | return rendition.get('NAME') or stream_group_id | |
1730 | ||
379306ef | 1731 | # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the |
2bfc1d9d RA |
1732 | # chance to detect video only formats when EXT-X-STREAM-INF tags |
1733 | # precede EXT-X-MEDIA tags in HLS manifest such as [3]. | |
1734 | for line in m3u8_doc.splitlines(): | |
1735 | if line.startswith('#EXT-X-MEDIA:'): | |
1736 | extract_media(line) | |
1737 | ||
704df56d PH |
1738 | for line in m3u8_doc.splitlines(): |
1739 | if line.startswith('#EXT-X-STREAM-INF:'): | |
cb252080 | 1740 | last_stream_inf = parse_m3u8_attributes(line) |
704df56d PH |
1741 | elif line.startswith('#') or not line.strip(): |
1742 | continue | |
1743 | else: | |
9c99bef7 | 1744 | tbr = float_or_none( |
3089bc74 S |
1745 | last_stream_inf.get('AVERAGE-BANDWIDTH') |
1746 | or last_stream_inf.get('BANDWIDTH'), scale=1000) | |
8dc9d361 S |
1747 | format_id = [] |
1748 | if m3u8_id: | |
1749 | format_id.append(m3u8_id) | |
cb252080 | 1750 | stream_name = build_stream_name() |
b24d6336 KH |
1751 | # Bandwidth of live streams may differ over time thus making |
1752 | # format_id unpredictable. So it's better to keep provided | |
1753 | # format_id intact. | |
e9c6cdf4 | 1754 | if not live: |
ed56f260 | 1755 | format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats))) |
30d0b549 | 1756 | manifest_url = format_url(line.strip()) |
704df56d | 1757 | f = { |
8dc9d361 | 1758 | 'format_id': '-'.join(format_id), |
30d0b549 | 1759 | 'url': manifest_url, |
ff99fe52 | 1760 | 'manifest_url': m3u8_url, |
704df56d PH |
1761 | 'tbr': tbr, |
1762 | 'ext': ext, | |
cb252080 | 1763 | 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')), |
f0b5d6af PH |
1764 | 'protocol': entry_protocol, |
1765 | 'preference': preference, | |
704df56d | 1766 | } |
cb252080 | 1767 | resolution = last_stream_inf.get('RESOLUTION') |
704df56d | 1768 | if resolution: |
c4c9b844 S |
1769 | mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution) |
1770 | if mobj: | |
1771 | f['width'] = int(mobj.group('width')) | |
1772 | f['height'] = int(mobj.group('height')) | |
00f4764c RA |
1773 | # Unified Streaming Platform |
1774 | mobj = re.search( | |
1775 | r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url']) | |
1776 | if mobj: | |
1777 | abr, vbr = mobj.groups() | |
1778 | abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000) | |
fbb6edd2 | 1779 | f.update({ |
00f4764c RA |
1780 | 'vbr': vbr, |
1781 | 'abr': abr, | |
fbb6edd2 | 1782 | }) |
cb252080 S |
1783 | codecs = parse_codecs(last_stream_inf.get('CODECS')) |
1784 | f.update(codecs) | |
1785 | audio_group_id = last_stream_inf.get('AUDIO') | |
1786 | # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which | |
1787 | # references a rendition group MUST have a CODECS attribute. | |
1788 | # However, this is not always respected, for example, [2] | |
1789 | # contains EXT-X-STREAM-INF tag which references AUDIO | |
1790 | # rendition group but does not have CODECS and despite | |
0e7b8d3e S |
1791 | # referencing an audio group it represents a complete |
1792 | # (with audio and video) format. So, for such cases we will | |
1793 | # ignore references to rendition groups and treat them | |
cb252080 S |
1794 | # as complete formats. |
1795 | if audio_group_id and codecs and f.get('vcodec') != 'none': | |
1796 | audio_group = groups.get(audio_group_id) | |
1797 | if audio_group and audio_group[0].get('URI'): | |
1798 | # TODO: update acodec for audio only formats with | |
1799 | # the same GROUP-ID | |
1800 | f['acodec'] = 'none' | |
704df56d | 1801 | formats.append(f) |
5ef62fc4 RA |
1802 | |
1803 | # for DailyMotion | |
1804 | progressive_uri = last_stream_inf.get('PROGRESSIVE-URI') | |
1805 | if progressive_uri: | |
1806 | http_f = f.copy() | |
1807 | del http_f['manifest_url'] | |
1808 | http_f.update({ | |
1809 | 'format_id': f['format_id'].replace('hls-', 'http-'), | |
1810 | 'protocol': 'http', | |
1811 | 'url': progressive_uri, | |
1812 | }) | |
1813 | formats.append(http_f) | |
1814 | ||
cb252080 | 1815 | last_stream_inf = {} |
704df56d PH |
1816 | return formats |
1817 | ||
a107193e S |
1818 | @staticmethod |
1819 | def _xpath_ns(path, namespace=None): | |
1820 | if not namespace: | |
1821 | return path | |
1822 | out = [] | |
1823 | for c in path.split('/'): | |
1824 | if not c or c == '.': | |
1825 | out.append(c) | |
1826 | else: | |
1827 | out.append('{%s}%s' % (namespace, c)) | |
1828 | return '/'.join(out) | |
1829 | ||
09f572fb | 1830 | def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None): |
1831 | smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source) | |
a107193e | 1832 | |
995029a1 PH |
1833 | if smil is False: |
1834 | assert not fatal | |
1835 | return [] | |
e89a2aab | 1836 | |
17712eeb | 1837 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
1838 | |
1839 | return self._parse_smil_formats( | |
1840 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
1841 | ||
1842 | def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None): | |
1843 | smil = self._download_smil(smil_url, video_id, fatal=fatal) | |
1844 | if smil is False: | |
1845 | return {} | |
1846 | return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params) | |
1847 | ||
09f572fb | 1848 | def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None): |
a107193e S |
1849 | return self._download_xml( |
1850 | smil_url, video_id, 'Downloading SMIL file', | |
09f572fb | 1851 | 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source) |
a107193e S |
1852 | |
1853 | def _parse_smil(self, smil, smil_url, video_id, f4m_params=None): | |
17712eeb | 1854 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
1855 | |
1856 | formats = self._parse_smil_formats( | |
1857 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
1858 | subtitles = self._parse_smil_subtitles(smil, namespace=namespace) | |
1859 | ||
1860 | video_id = os.path.splitext(url_basename(smil_url))[0] | |
1861 | title = None | |
1862 | description = None | |
647eab45 | 1863 | upload_date = None |
a107193e S |
1864 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): |
1865 | name = meta.attrib.get('name') | |
1866 | content = meta.attrib.get('content') | |
1867 | if not name or not content: | |
1868 | continue | |
1869 | if not title and name == 'title': | |
1870 | title = content | |
1871 | elif not description and name in ('description', 'abstract'): | |
1872 | description = content | |
647eab45 S |
1873 | elif not upload_date and name == 'date': |
1874 | upload_date = unified_strdate(content) | |
a107193e | 1875 | |
1e5bcdec S |
1876 | thumbnails = [{ |
1877 | 'id': image.get('type'), | |
1878 | 'url': image.get('src'), | |
1879 | 'width': int_or_none(image.get('width')), | |
1880 | 'height': int_or_none(image.get('height')), | |
1881 | } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')] | |
1882 | ||
a107193e S |
1883 | return { |
1884 | 'id': video_id, | |
1885 | 'title': title or video_id, | |
1886 | 'description': description, | |
647eab45 | 1887 | 'upload_date': upload_date, |
1e5bcdec | 1888 | 'thumbnails': thumbnails, |
a107193e S |
1889 | 'formats': formats, |
1890 | 'subtitles': subtitles, | |
1891 | } | |
1892 | ||
17712eeb S |
1893 | def _parse_smil_namespace(self, smil): |
1894 | return self._search_regex( | |
1895 | r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None) | |
1896 | ||
f877c6ae | 1897 | def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None): |
a107193e S |
1898 | base = smil_url |
1899 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): | |
1900 | b = meta.get('base') or meta.get('httpBase') | |
1901 | if b: | |
1902 | base = b | |
1903 | break | |
e89a2aab S |
1904 | |
1905 | formats = [] | |
1906 | rtmp_count = 0 | |
a107193e | 1907 | http_count = 0 |
7f32e5dc | 1908 | m3u8_count = 0 |
a107193e | 1909 | |
81e1c4e2 | 1910 | srcs = [] |
ad96b4c8 YCH |
1911 | media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace)) |
1912 | for medium in media: | |
1913 | src = medium.get('src') | |
81e1c4e2 | 1914 | if not src or src in srcs: |
a107193e | 1915 | continue |
81e1c4e2 | 1916 | srcs.append(src) |
a107193e | 1917 | |
ad96b4c8 YCH |
1918 | bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000) |
1919 | filesize = int_or_none(medium.get('size') or medium.get('fileSize')) | |
1920 | width = int_or_none(medium.get('width')) | |
1921 | height = int_or_none(medium.get('height')) | |
1922 | proto = medium.get('proto') | |
1923 | ext = medium.get('ext') | |
a107193e | 1924 | src_ext = determine_ext(src) |
ad96b4c8 | 1925 | streamer = medium.get('streamer') or base |
a107193e S |
1926 | |
1927 | if proto == 'rtmp' or streamer.startswith('rtmp'): | |
1928 | rtmp_count += 1 | |
1929 | formats.append({ | |
1930 | 'url': streamer, | |
1931 | 'play_path': src, | |
1932 | 'ext': 'flv', | |
1933 | 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), | |
1934 | 'tbr': bitrate, | |
1935 | 'filesize': filesize, | |
1936 | 'width': width, | |
1937 | 'height': height, | |
1938 | }) | |
f877c6ae YCH |
1939 | if transform_rtmp_url: |
1940 | streamer, src = transform_rtmp_url(streamer, src) | |
1941 | formats[-1].update({ | |
1942 | 'url': streamer, | |
1943 | 'play_path': src, | |
1944 | }) | |
a107193e S |
1945 | continue |
1946 | ||
1947 | src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src) | |
c349456e | 1948 | src_url = src_url.strip() |
a107193e S |
1949 | |
1950 | if proto == 'm3u8' or src_ext == 'm3u8': | |
7f32e5dc | 1951 | m3u8_formats = self._extract_m3u8_formats( |
1952 | src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False) | |
1953 | if len(m3u8_formats) == 1: | |
1954 | m3u8_count += 1 | |
1955 | m3u8_formats[0].update({ | |
1956 | 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate), | |
1957 | 'tbr': bitrate, | |
1958 | 'width': width, | |
1959 | 'height': height, | |
1960 | }) | |
1961 | formats.extend(m3u8_formats) | |
bd21ead2 | 1962 | elif src_ext == 'f4m': |
a107193e S |
1963 | f4m_url = src_url |
1964 | if not f4m_params: | |
1965 | f4m_params = { | |
1966 | 'hdcore': '3.2.0', | |
1967 | 'plugin': 'flowplayer-3.2.0.1', | |
1968 | } | |
1969 | f4m_url += '&' if '?' in f4m_url else '?' | |
15707c7e | 1970 | f4m_url += compat_urllib_parse_urlencode(f4m_params) |
7e5edcfd | 1971 | formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False)) |
bd21ead2 RA |
1972 | elif src_ext == 'mpd': |
1973 | formats.extend(self._extract_mpd_formats( | |
1974 | src_url, video_id, mpd_id='dash', fatal=False)) | |
1975 | elif re.search(r'\.ism/[Mm]anifest', src_url): | |
1976 | formats.extend(self._extract_ism_formats( | |
1977 | src_url, video_id, ism_id='mss', fatal=False)) | |
1978 | elif src_url.startswith('http') and self._is_valid_url(src, video_id): | |
a107193e S |
1979 | http_count += 1 |
1980 | formats.append({ | |
1981 | 'url': src_url, | |
1982 | 'ext': ext or src_ext or 'flv', | |
1983 | 'format_id': 'http-%d' % (bitrate or http_count), | |
1984 | 'tbr': bitrate, | |
1985 | 'filesize': filesize, | |
1986 | 'width': width, | |
1987 | 'height': height, | |
1988 | }) | |
63757032 | 1989 | |
e89a2aab S |
1990 | return formats |
1991 | ||
ce00af87 | 1992 | def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'): |
d413095f | 1993 | urls = [] |
a107193e S |
1994 | subtitles = {} |
1995 | for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))): | |
1996 | src = textstream.get('src') | |
d413095f | 1997 | if not src or src in urls: |
a107193e | 1998 | continue |
d413095f | 1999 | urls.append(src) |
df634be2 | 2000 | ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src) |
03bc7237 | 2001 | lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang |
a107193e S |
2002 | subtitles.setdefault(lang, []).append({ |
2003 | 'url': src, | |
2004 | 'ext': ext, | |
2005 | }) | |
2006 | return subtitles | |
63757032 | 2007 | |
47a5cb77 | 2008 | def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True): |
942acef5 | 2009 | xspf = self._download_xml( |
47a5cb77 | 2010 | xspf_url, playlist_id, 'Downloading xpsf playlist', |
942acef5 S |
2011 | 'Unable to download xspf manifest', fatal=fatal) |
2012 | if xspf is False: | |
2013 | return [] | |
47a5cb77 S |
2014 | return self._parse_xspf( |
2015 | xspf, playlist_id, xspf_url=xspf_url, | |
2016 | xspf_base_url=base_url(xspf_url)) | |
8d6765cf | 2017 | |
47a5cb77 | 2018 | def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None): |
8d6765cf S |
2019 | NS_MAP = { |
2020 | 'xspf': 'http://xspf.org/ns/0/', | |
2021 | 's1': 'http://static.streamone.nl/player/ns/0', | |
2022 | } | |
2023 | ||
2024 | entries = [] | |
47a5cb77 | 2025 | for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)): |
8d6765cf | 2026 | title = xpath_text( |
98044462 | 2027 | track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id) |
8d6765cf S |
2028 | description = xpath_text( |
2029 | track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description') | |
2030 | thumbnail = xpath_text( | |
2031 | track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail') | |
2032 | duration = float_or_none( | |
2033 | xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000) | |
2034 | ||
47a5cb77 S |
2035 | formats = [] |
2036 | for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)): | |
2037 | format_url = urljoin(xspf_base_url, location.text) | |
2038 | if not format_url: | |
2039 | continue | |
2040 | formats.append({ | |
2041 | 'url': format_url, | |
2042 | 'manifest_url': xspf_url, | |
2043 | 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)), | |
2044 | 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))), | |
2045 | 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))), | |
2046 | }) | |
8d6765cf S |
2047 | self._sort_formats(formats) |
2048 | ||
2049 | entries.append({ | |
2050 | 'id': playlist_id, | |
2051 | 'title': title, | |
2052 | 'description': description, | |
2053 | 'thumbnail': thumbnail, | |
2054 | 'duration': duration, | |
2055 | 'formats': formats, | |
2056 | }) | |
2057 | return entries | |
2058 | ||
7360c06f | 2059 | def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}, data=None, headers={}, query={}): |
47a5cb77 | 2060 | res = self._download_xml_handle( |
1bac3455 | 2061 | mpd_url, video_id, |
2062 | note=note or 'Downloading MPD manifest', | |
2063 | errnote=errnote or 'Failed to download MPD manifest', | |
7360c06f | 2064 | fatal=fatal, data=data, headers=headers, query=query) |
1bac3455 | 2065 | if res is False: |
2d2fa82d | 2066 | return [] |
47a5cb77 | 2067 | mpd_doc, urlh = res |
c25720ef RA |
2068 | if mpd_doc is None: |
2069 | return [] | |
02dc0a36 | 2070 | mpd_base_url = base_url(urlh.geturl()) |
1bac3455 | 2071 | |
91cb6b50 | 2072 | return self._parse_mpd_formats( |
47a5cb77 | 2073 | mpd_doc, mpd_id=mpd_id, mpd_base_url=mpd_base_url, |
86f4d14f | 2074 | formats_dict=formats_dict, mpd_url=mpd_url) |
2d2fa82d | 2075 | |
86f4d14f | 2076 | def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}, mpd_url=None): |
f0948348 S |
2077 | """ |
2078 | Parse formats from MPD manifest. | |
2079 | References: | |
2080 | 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E), | |
2081 | http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip | |
2082 | 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP | |
2083 | """ | |
78895bd3 U |
2084 | if not self._downloader.params.get('dynamic_mpd'): |
2085 | if mpd_doc.get('type') == 'dynamic': | |
2086 | return [] | |
2d2fa82d | 2087 | |
91cb6b50 | 2088 | namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None) |
f14be228 | 2089 | |
2090 | def _add_ns(path): | |
2091 | return self._xpath_ns(path, namespace) | |
2092 | ||
675d0016 | 2093 | def is_drm_protected(element): |
2094 | return element.find(_add_ns('ContentProtection')) is not None | |
2095 | ||
1bac3455 | 2096 | def extract_multisegment_info(element, ms_parent_info): |
2097 | ms_info = ms_parent_info.copy() | |
b4c1d6e8 S |
2098 | |
2099 | # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some | |
2100 | # common attributes and elements. We will only extract relevant | |
2101 | # for us. | |
2102 | def extract_common(source): | |
2103 | segment_timeline = source.find(_add_ns('SegmentTimeline')) | |
2104 | if segment_timeline is not None: | |
2105 | s_e = segment_timeline.findall(_add_ns('S')) | |
2106 | if s_e: | |
2107 | ms_info['total_number'] = 0 | |
2108 | ms_info['s'] = [] | |
2109 | for s in s_e: | |
2110 | r = int(s.get('r', 0)) | |
2111 | ms_info['total_number'] += 1 + r | |
2112 | ms_info['s'].append({ | |
2113 | 't': int(s.get('t', 0)), | |
2114 | # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60]) | |
2115 | 'd': int(s.attrib['d']), | |
2116 | 'r': r, | |
2117 | }) | |
2118 | start_number = source.get('startNumber') | |
2119 | if start_number: | |
2120 | ms_info['start_number'] = int(start_number) | |
2121 | timescale = source.get('timescale') | |
2122 | if timescale: | |
2123 | ms_info['timescale'] = int(timescale) | |
2124 | segment_duration = source.get('duration') | |
2125 | if segment_duration: | |
48504785 | 2126 | ms_info['segment_duration'] = float(segment_duration) |
b4c1d6e8 S |
2127 | |
2128 | def extract_Initialization(source): | |
2129 | initialization = source.find(_add_ns('Initialization')) | |
2130 | if initialization is not None: | |
2131 | ms_info['initialization_url'] = initialization.attrib['sourceURL'] | |
2132 | ||
f14be228 | 2133 | segment_list = element.find(_add_ns('SegmentList')) |
1bac3455 | 2134 | if segment_list is not None: |
b4c1d6e8 S |
2135 | extract_common(segment_list) |
2136 | extract_Initialization(segment_list) | |
f14be228 | 2137 | segment_urls_e = segment_list.findall(_add_ns('SegmentURL')) |
1bac3455 | 2138 | if segment_urls_e: |
2139 | ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e] | |
1bac3455 | 2140 | else: |
f14be228 | 2141 | segment_template = element.find(_add_ns('SegmentTemplate')) |
1bac3455 | 2142 | if segment_template is not None: |
b4c1d6e8 | 2143 | extract_common(segment_template) |
e228616c S |
2144 | media = segment_template.get('media') |
2145 | if media: | |
2146 | ms_info['media'] = media | |
1bac3455 | 2147 | initialization = segment_template.get('initialization') |
2148 | if initialization: | |
e228616c | 2149 | ms_info['initialization'] = initialization |
1bac3455 | 2150 | else: |
b4c1d6e8 | 2151 | extract_Initialization(segment_template) |
1bac3455 | 2152 | return ms_info |
b323e170 | 2153 | |
1bac3455 | 2154 | mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration')) |
17b598d3 | 2155 | formats = [] |
f14be228 | 2156 | for period in mpd_doc.findall(_add_ns('Period')): |
1bac3455 | 2157 | period_duration = parse_duration(period.get('duration')) or mpd_duration |
2158 | period_ms_info = extract_multisegment_info(period, { | |
2159 | 'start_number': 1, | |
2160 | 'timescale': 1, | |
2161 | }) | |
f14be228 | 2162 | for adaptation_set in period.findall(_add_ns('AdaptationSet')): |
675d0016 | 2163 | if is_drm_protected(adaptation_set): |
2164 | continue | |
1bac3455 | 2165 | adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info) |
f14be228 | 2166 | for representation in adaptation_set.findall(_add_ns('Representation')): |
675d0016 | 2167 | if is_drm_protected(representation): |
2168 | continue | |
1bac3455 | 2169 | representation_attrib = adaptation_set.attrib.copy() |
2170 | representation_attrib.update(representation.attrib) | |
f0948348 | 2171 | # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory |
a6c8b759 YCH |
2172 | mime_type = representation_attrib['mimeType'] |
2173 | content_type = mime_type.split('/')[0] | |
1bac3455 | 2174 | if content_type == 'text': |
2175 | # TODO implement WebVTT downloading | |
2176 | pass | |
40fcba5e | 2177 | elif content_type in ('video', 'audio'): |
1bac3455 | 2178 | base_url = '' |
2179 | for element in (representation, adaptation_set, period, mpd_doc): | |
f14be228 | 2180 | base_url_e = element.find(_add_ns('BaseURL')) |
1bac3455 | 2181 | if base_url_e is not None: |
2182 | base_url = base_url_e.text + base_url | |
2183 | if re.match(r'^https?://', base_url): | |
2184 | break | |
bb20526b S |
2185 | if mpd_base_url and not re.match(r'^https?://', base_url): |
2186 | if not mpd_base_url.endswith('/') and not base_url.startswith('/'): | |
2187 | mpd_base_url += '/' | |
1bac3455 | 2188 | base_url = mpd_base_url + base_url |
2189 | representation_id = representation_attrib.get('id') | |
d577c796 | 2190 | lang = representation_attrib.get('lang') |
51e9094f | 2191 | url_el = representation.find(_add_ns('BaseURL')) |
2192 | filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None) | |
e228616c | 2193 | bandwidth = int_or_none(representation_attrib.get('bandwidth')) |
1bac3455 | 2194 | f = { |
154c209e | 2195 | 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id, |
86f4d14f | 2196 | 'manifest_url': mpd_url, |
a6c8b759 | 2197 | 'ext': mimetype2ext(mime_type), |
1bac3455 | 2198 | 'width': int_or_none(representation_attrib.get('width')), |
2199 | 'height': int_or_none(representation_attrib.get('height')), | |
9c99bef7 | 2200 | 'tbr': float_or_none(bandwidth, 1000), |
1bac3455 | 2201 | 'asr': int_or_none(representation_attrib.get('audioSamplingRate')), |
2202 | 'fps': int_or_none(representation_attrib.get('frameRate')), | |
d577c796 | 2203 | 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None, |
1bac3455 | 2204 | 'format_note': 'DASH %s' % content_type, |
51e9094f | 2205 | 'filesize': filesize, |
126f225b | 2206 | 'container': mimetype2ext(mime_type) + '_dash', |
1bac3455 | 2207 | } |
7fe15920 | 2208 | f.update(parse_codecs(representation_attrib.get('codecs'))) |
1bac3455 | 2209 | representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info) |
b4c1d6e8 | 2210 | |
e228616c | 2211 | def prepare_template(template_name, identifiers): |
eca1f0d1 S |
2212 | tmpl = representation_ms_info[template_name] |
2213 | # First of, % characters outside $...$ templates | |
2214 | # must be escaped by doubling for proper processing | |
2215 | # by % operator string formatting used further (see | |
067aa17e | 2216 | # https://github.com/ytdl-org/youtube-dl/issues/16867). |
eca1f0d1 S |
2217 | t = '' |
2218 | in_template = False | |
2219 | for c in tmpl: | |
2220 | t += c | |
2221 | if c == '$': | |
2222 | in_template = not in_template | |
2223 | elif c == '%' and not in_template: | |
2224 | t += c | |
2225 | # Next, $...$ templates are translated to their | |
2226 | # %(...) counterparts to be used with % operator | |
e228616c S |
2227 | t = t.replace('$RepresentationID$', representation_id) |
2228 | t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t) | |
2229 | t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t) | |
2230 | t.replace('$$', '$') | |
2231 | return t | |
2232 | ||
2233 | # @initialization is a regular template like @media one | |
2234 | # so it should be handled just the same way (see | |
067aa17e | 2235 | # https://github.com/ytdl-org/youtube-dl/issues/11605) |
e228616c S |
2236 | if 'initialization' in representation_ms_info: |
2237 | initialization_template = prepare_template( | |
2238 | 'initialization', | |
2239 | # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and | |
2240 | # $Time$ shall not be included for @initialization thus | |
2241 | # only $Bandwidth$ remains | |
2242 | ('Bandwidth', )) | |
2243 | representation_ms_info['initialization_url'] = initialization_template % { | |
2244 | 'Bandwidth': bandwidth, | |
2245 | } | |
2246 | ||
1141e910 S |
2247 | def location_key(location): |
2248 | return 'url' if re.match(r'^https?://', location) else 'path' | |
2249 | ||
e228616c S |
2250 | if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info: |
2251 | ||
2252 | media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time')) | |
1141e910 | 2253 | media_location_key = location_key(media_template) |
f0948348 S |
2254 | |
2255 | # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$ | |
2256 | # can't be used at the same time | |
b4c1d6e8 S |
2257 | if '%(Number' in media_template and 's' not in representation_ms_info: |
2258 | segment_duration = None | |
c110944f | 2259 | if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info: |
b4c1d6e8 S |
2260 | segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale']) |
2261 | representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration)) | |
b4c1d6e8 | 2262 | representation_ms_info['fragments'] = [{ |
1141e910 | 2263 | media_location_key: media_template % { |
b4c1d6e8 | 2264 | 'Number': segment_number, |
e228616c | 2265 | 'Bandwidth': bandwidth, |
b4c1d6e8 S |
2266 | }, |
2267 | 'duration': segment_duration, | |
2268 | } for segment_number in range( | |
2269 | representation_ms_info['start_number'], | |
2270 | representation_ms_info['total_number'] + representation_ms_info['start_number'])] | |
f0948348 | 2271 | else: |
b4c1d6e8 S |
2272 | # $Number*$ or $Time$ in media template with S list available |
2273 | # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg | |
2274 | # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411 | |
b4c1d6e8 | 2275 | representation_ms_info['fragments'] = [] |
f0948348 | 2276 | segment_time = 0 |
b4c1d6e8 S |
2277 | segment_d = None |
2278 | segment_number = representation_ms_info['start_number'] | |
f0948348 S |
2279 | |
2280 | def add_segment_url(): | |
b4c1d6e8 S |
2281 | segment_url = media_template % { |
2282 | 'Time': segment_time, | |
e228616c | 2283 | 'Bandwidth': bandwidth, |
b4c1d6e8 S |
2284 | 'Number': segment_number, |
2285 | } | |
b4c1d6e8 | 2286 | representation_ms_info['fragments'].append({ |
1141e910 | 2287 | media_location_key: segment_url, |
b4c1d6e8 S |
2288 | 'duration': float_or_none(segment_d, representation_ms_info['timescale']), |
2289 | }) | |
f0948348 S |
2290 | |
2291 | for num, s in enumerate(representation_ms_info['s']): | |
2292 | segment_time = s.get('t') or segment_time | |
b4c1d6e8 | 2293 | segment_d = s['d'] |
f0948348 | 2294 | add_segment_url() |
b4c1d6e8 | 2295 | segment_number += 1 |
f0948348 | 2296 | for r in range(s.get('r', 0)): |
b4c1d6e8 | 2297 | segment_time += segment_d |
f0948348 | 2298 | add_segment_url() |
b4c1d6e8 S |
2299 | segment_number += 1 |
2300 | segment_time += segment_d | |
2301 | elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info: | |
2302 | # No media template | |
2303 | # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI | |
2304 | # or any YouTube dashsegments video | |
2305 | fragments = [] | |
d04621da S |
2306 | segment_index = 0 |
2307 | timescale = representation_ms_info['timescale'] | |
2308 | for s in representation_ms_info['s']: | |
2309 | duration = float_or_none(s['d'], timescale) | |
b4c1d6e8 | 2310 | for r in range(s.get('r', 0) + 1): |
1141e910 | 2311 | segment_uri = representation_ms_info['segment_urls'][segment_index] |
b4c1d6e8 | 2312 | fragments.append({ |
1141e910 | 2313 | location_key(segment_uri): segment_uri, |
d04621da | 2314 | 'duration': duration, |
b4c1d6e8 | 2315 | }) |
d04621da | 2316 | segment_index += 1 |
b4c1d6e8 | 2317 | representation_ms_info['fragments'] = fragments |
41bf647e PN |
2318 | elif 'segment_urls' in representation_ms_info: |
2319 | # Segment URLs with no SegmentTimeline | |
2320 | # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091 | |
067aa17e | 2321 | # https://github.com/ytdl-org/youtube-dl/pull/14844 |
41bf647e | 2322 | fragments = [] |
603fc4e0 S |
2323 | segment_duration = float_or_none( |
2324 | representation_ms_info['segment_duration'], | |
2325 | representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None | |
41bf647e | 2326 | for segment_url in representation_ms_info['segment_urls']: |
603fc4e0 | 2327 | fragment = { |
41bf647e | 2328 | location_key(segment_url): segment_url, |
603fc4e0 S |
2329 | } |
2330 | if segment_duration: | |
2331 | fragment['duration'] = segment_duration | |
2332 | fragments.append(fragment) | |
41bf647e | 2333 | representation_ms_info['fragments'] = fragments |
79d2077e S |
2334 | # If there is a fragments key available then we correctly recognized fragmented media. |
2335 | # Otherwise we will assume unfragmented media with direct access. Technically, such | |
2336 | # assumption is not necessarily correct since we may simply have no support for | |
2337 | # some forms of fragmented media renditions yet, but for now we'll use this fallback. | |
86f4d14f | 2338 | if 'fragments' in representation_ms_info: |
1bac3455 | 2339 | f.update({ |
79d2077e S |
2340 | # NB: mpd_url may be empty when MPD manifest is parsed from a string |
2341 | 'url': mpd_url or base_url, | |
1141e910 | 2342 | 'fragment_base_url': base_url, |
b4c1d6e8 | 2343 | 'fragments': [], |
1bac3455 | 2344 | 'protocol': 'http_dash_segments', |
df374b52 | 2345 | }) |
1bac3455 | 2346 | if 'initialization_url' in representation_ms_info: |
e228616c | 2347 | initialization_url = representation_ms_info['initialization_url'] |
1bac3455 | 2348 | if not f.get('url'): |
2349 | f['url'] = initialization_url | |
1141e910 | 2350 | f['fragments'].append({location_key(initialization_url): initialization_url}) |
b4c1d6e8 | 2351 | f['fragments'].extend(representation_ms_info['fragments']) |
79d2077e S |
2352 | else: |
2353 | # Assuming direct URL to unfragmented media. | |
2354 | f['url'] = base_url | |
2355 | ||
9d6ac71c S |
2356 | # According to [1, 5.3.5.2, Table 7, page 35] @id of Representation |
2357 | # is not necessarily unique within a Period thus formats with | |
2358 | # the same `format_id` are quite possible. There are numerous examples | |
067aa17e S |
2359 | # of such manifests (see https://github.com/ytdl-org/youtube-dl/issues/15111, |
2360 | # https://github.com/ytdl-org/youtube-dl/issues/13919) | |
9d6ac71c S |
2361 | full_info = formats_dict.get(representation_id, {}).copy() |
2362 | full_info.update(f) | |
2363 | formats.append(full_info) | |
17b598d3 | 2364 | else: |
1bac3455 | 2365 | self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) |
17b598d3 YCH |
2366 | return formats |
2367 | ||
7360c06f | 2368 | def _extract_ism_formats(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}): |
47a5cb77 | 2369 | res = self._download_xml_handle( |
b2758123 RA |
2370 | ism_url, video_id, |
2371 | note=note or 'Downloading ISM manifest', | |
2372 | errnote=errnote or 'Failed to download ISM manifest', | |
7360c06f | 2373 | fatal=fatal, data=data, headers=headers, query=query) |
b2758123 RA |
2374 | if res is False: |
2375 | return [] | |
47a5cb77 | 2376 | ism_doc, urlh = res |
13b08034 S |
2377 | if ism_doc is None: |
2378 | return [] | |
b2758123 | 2379 | |
7947a1f7 | 2380 | return self._parse_ism_formats(ism_doc, urlh.geturl(), ism_id) |
b2758123 RA |
2381 | |
2382 | def _parse_ism_formats(self, ism_doc, ism_url, ism_id=None): | |
76d5a363 S |
2383 | """ |
2384 | Parse formats from ISM manifest. | |
2385 | References: | |
2386 | 1. [MS-SSTR]: Smooth Streaming Protocol, | |
2387 | https://msdn.microsoft.com/en-us/library/ff469518.aspx | |
2388 | """ | |
b2758123 RA |
2389 | if ism_doc.get('IsLive') == 'TRUE' or ism_doc.find('Protection') is not None: |
2390 | return [] | |
2391 | ||
b2758123 RA |
2392 | duration = int(ism_doc.attrib['Duration']) |
2393 | timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000 | |
2394 | ||
2395 | formats = [] | |
2396 | for stream in ism_doc.findall('StreamIndex'): | |
2397 | stream_type = stream.get('Type') | |
2398 | if stream_type not in ('video', 'audio'): | |
2399 | continue | |
2400 | url_pattern = stream.attrib['Url'] | |
2401 | stream_timescale = int_or_none(stream.get('TimeScale')) or timescale | |
2402 | stream_name = stream.get('Name') | |
2403 | for track in stream.findall('QualityLevel'): | |
2501d41e | 2404 | fourcc = track.get('FourCC', 'AACL' if track.get('AudioTag') == '255' else None) |
b2758123 RA |
2405 | # TODO: add support for WVC1 and WMAP |
2406 | if fourcc not in ('H264', 'AVC1', 'AACL'): | |
2407 | self.report_warning('%s is not a supported codec' % fourcc) | |
2408 | continue | |
2409 | tbr = int(track.attrib['Bitrate']) // 1000 | |
76d5a363 S |
2410 | # [1] does not mention Width and Height attributes. However, |
2411 | # they're often present while MaxWidth and MaxHeight are | |
2412 | # missing, so should be used as fallbacks | |
2413 | width = int_or_none(track.get('MaxWidth') or track.get('Width')) | |
2414 | height = int_or_none(track.get('MaxHeight') or track.get('Height')) | |
b2758123 RA |
2415 | sampling_rate = int_or_none(track.get('SamplingRate')) |
2416 | ||
2417 | track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern) | |
2418 | track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern) | |
2419 | ||
2420 | fragments = [] | |
2421 | fragment_ctx = { | |
2422 | 'time': 0, | |
2423 | } | |
2424 | stream_fragments = stream.findall('c') | |
2425 | for stream_fragment_index, stream_fragment in enumerate(stream_fragments): | |
2426 | fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time'] | |
2427 | fragment_repeat = int_or_none(stream_fragment.get('r')) or 1 | |
2428 | fragment_ctx['duration'] = int_or_none(stream_fragment.get('d')) | |
2429 | if not fragment_ctx['duration']: | |
2430 | try: | |
2431 | next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t']) | |
2432 | except IndexError: | |
2433 | next_fragment_time = duration | |
1616f9b4 | 2434 | fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat |
b2758123 RA |
2435 | for _ in range(fragment_repeat): |
2436 | fragments.append({ | |
1616f9b4 | 2437 | 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern), |
b2758123 RA |
2438 | 'duration': fragment_ctx['duration'] / stream_timescale, |
2439 | }) | |
2440 | fragment_ctx['time'] += fragment_ctx['duration'] | |
2441 | ||
2442 | format_id = [] | |
2443 | if ism_id: | |
2444 | format_id.append(ism_id) | |
2445 | if stream_name: | |
2446 | format_id.append(stream_name) | |
2447 | format_id.append(compat_str(tbr)) | |
2448 | ||
2449 | formats.append({ | |
2450 | 'format_id': '-'.join(format_id), | |
2451 | 'url': ism_url, | |
2452 | 'manifest_url': ism_url, | |
2453 | 'ext': 'ismv' if stream_type == 'video' else 'isma', | |
2454 | 'width': width, | |
2455 | 'height': height, | |
2456 | 'tbr': tbr, | |
2457 | 'asr': sampling_rate, | |
2458 | 'vcodec': 'none' if stream_type == 'audio' else fourcc, | |
2459 | 'acodec': 'none' if stream_type == 'video' else fourcc, | |
2460 | 'protocol': 'ism', | |
2461 | 'fragments': fragments, | |
2462 | '_download_params': { | |
2463 | 'duration': duration, | |
2464 | 'timescale': stream_timescale, | |
2465 | 'width': width or 0, | |
2466 | 'height': height or 0, | |
2467 | 'fourcc': fourcc, | |
2468 | 'codec_private_data': track.get('CodecPrivateData'), | |
2469 | 'sampling_rate': sampling_rate, | |
2470 | 'channels': int_or_none(track.get('Channels', 2)), | |
2471 | 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)), | |
2472 | 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)), | |
2473 | }, | |
2474 | }) | |
2475 | return formats | |
2476 | ||
eeb0a956 | 2477 | def _parse_html5_media_entries(self, base_url, webpage, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8', mpd_id=None, preference=None): |
6780154e S |
2478 | def absolute_url(item_url): |
2479 | return urljoin(base_url, item_url) | |
59bbe491 | 2480 | |
2481 | def parse_content_type(content_type): | |
2482 | if not content_type: | |
2483 | return {} | |
2484 | ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type) | |
2485 | if ctr: | |
2486 | mimetype, codecs = ctr.groups() | |
2487 | f = parse_codecs(codecs) | |
2488 | f['ext'] = mimetype2ext(mimetype) | |
2489 | return f | |
2490 | return {} | |
2491 | ||
868f79db | 2492 | def _media_formats(src, cur_media_type, type_info={}): |
520251c0 | 2493 | full_url = absolute_url(src) |
82889d4a | 2494 | ext = type_info.get('ext') or determine_ext(full_url) |
87a449c1 | 2495 | if ext == 'm3u8': |
520251c0 YCH |
2496 | is_plain_url = False |
2497 | formats = self._extract_m3u8_formats( | |
ad120ae1 | 2498 | full_url, video_id, ext='mp4', |
eeb0a956 | 2499 | entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id, |
b359e977 | 2500 | preference=preference, fatal=False) |
87a449c1 S |
2501 | elif ext == 'mpd': |
2502 | is_plain_url = False | |
2503 | formats = self._extract_mpd_formats( | |
b359e977 | 2504 | full_url, video_id, mpd_id=mpd_id, fatal=False) |
520251c0 YCH |
2505 | else: |
2506 | is_plain_url = True | |
2507 | formats = [{ | |
2508 | 'url': full_url, | |
2509 | 'vcodec': 'none' if cur_media_type == 'audio' else None, | |
2510 | }] | |
2511 | return is_plain_url, formats | |
2512 | ||
59bbe491 | 2513 | entries = [] |
4328ddf8 S |
2514 | # amp-video and amp-audio are very similar to their HTML5 counterparts |
2515 | # so we wll include them right here (see | |
2516 | # https://www.ampproject.org/docs/reference/components/amp-video) | |
cea364f7 YCH |
2517 | media_tags = [(media_tag, media_type, '') |
2518 | for media_tag, media_type | |
4328ddf8 | 2519 | in re.findall(r'(?s)(<(?:amp-)?(video|audio)[^>]*/>)', webpage)] |
2aec7256 S |
2520 | media_tags.extend(re.findall( |
2521 | # We only allow video|audio followed by a whitespace or '>'. | |
2522 | # Allowing more characters may end up in significant slow down (see | |
067aa17e | 2523 | # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL: |
2aec7256 | 2524 | # http://www.porntrex.com/maps/videositemap.xml). |
4328ddf8 | 2525 | r'(?s)(<(?P<tag>(?:amp-)?(?:video|audio))(?:\s+[^>]*)?>)(.*?)</(?P=tag)>', webpage)) |
cea364f7 | 2526 | for media_tag, media_type, media_content in media_tags: |
59bbe491 | 2527 | media_info = { |
2528 | 'formats': [], | |
2529 | 'subtitles': {}, | |
2530 | } | |
2531 | media_attributes = extract_attributes(media_tag) | |
f856816b | 2532 | src = strip_or_none(media_attributes.get('src')) |
59bbe491 | 2533 | if src: |
dedb1770 | 2534 | _, formats = _media_formats(src, media_type) |
520251c0 | 2535 | media_info['formats'].extend(formats) |
6780154e | 2536 | media_info['thumbnail'] = absolute_url(media_attributes.get('poster')) |
59bbe491 | 2537 | if media_content: |
2538 | for source_tag in re.findall(r'<source[^>]+>', media_content): | |
d493f15c S |
2539 | s_attr = extract_attributes(source_tag) |
2540 | # data-video-src and data-src are non standard but seen | |
2541 | # several times in the wild | |
f856816b | 2542 | src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src'))) |
59bbe491 | 2543 | if not src: |
2544 | continue | |
d493f15c | 2545 | f = parse_content_type(s_attr.get('type')) |
868f79db | 2546 | is_plain_url, formats = _media_formats(src, media_type, f) |
520251c0 | 2547 | if is_plain_url: |
d493f15c S |
2548 | # width, height, res, label and title attributes are |
2549 | # all not standard but seen several times in the wild | |
2550 | labels = [ | |
2551 | s_attr.get(lbl) | |
2552 | for lbl in ('label', 'title') | |
2553 | if str_or_none(s_attr.get(lbl)) | |
2554 | ] | |
2555 | width = int_or_none(s_attr.get('width')) | |
3089bc74 S |
2556 | height = (int_or_none(s_attr.get('height')) |
2557 | or int_or_none(s_attr.get('res'))) | |
d493f15c S |
2558 | if not width or not height: |
2559 | for lbl in labels: | |
2560 | resolution = parse_resolution(lbl) | |
2561 | if not resolution: | |
2562 | continue | |
2563 | width = width or resolution.get('width') | |
2564 | height = height or resolution.get('height') | |
2565 | for lbl in labels: | |
2566 | tbr = parse_bitrate(lbl) | |
2567 | if tbr: | |
2568 | break | |
2569 | else: | |
2570 | tbr = None | |
1ed45499 | 2571 | f.update({ |
d493f15c S |
2572 | 'width': width, |
2573 | 'height': height, | |
2574 | 'tbr': tbr, | |
2575 | 'format_id': s_attr.get('label') or s_attr.get('title'), | |
1ed45499 | 2576 | }) |
520251c0 YCH |
2577 | f.update(formats[0]) |
2578 | media_info['formats'].append(f) | |
2579 | else: | |
2580 | media_info['formats'].extend(formats) | |
59bbe491 | 2581 | for track_tag in re.findall(r'<track[^>]+>', media_content): |
2582 | track_attributes = extract_attributes(track_tag) | |
2583 | kind = track_attributes.get('kind') | |
5968d7d2 | 2584 | if not kind or kind in ('subtitles', 'captions'): |
f856816b | 2585 | src = strip_or_none(track_attributes.get('src')) |
59bbe491 | 2586 | if not src: |
2587 | continue | |
2588 | lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label') | |
2589 | media_info['subtitles'].setdefault(lang, []).append({ | |
2590 | 'url': absolute_url(src), | |
2591 | }) | |
5e8e2fa5 S |
2592 | for f in media_info['formats']: |
2593 | f.setdefault('http_headers', {})['Referer'] = base_url | |
5968d7d2 | 2594 | if media_info['formats'] or media_info['subtitles']: |
59bbe491 | 2595 | entries.append(media_info) |
2596 | return entries | |
2597 | ||
c4251b9a | 2598 | def _extract_akamai_formats(self, manifest_url, video_id, hosts={}): |
c7c43a93 | 2599 | formats = [] |
70c5802b | 2600 | |
e71a4509 | 2601 | hdcore_sign = 'hdcore=3.7.0' |
ff6f9a67 | 2602 | f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') |
c4251b9a RA |
2603 | hds_host = hosts.get('hds') |
2604 | if hds_host: | |
2605 | f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url) | |
e71a4509 RA |
2606 | if 'hdcore=' not in f4m_url: |
2607 | f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign | |
2608 | f4m_formats = self._extract_f4m_formats( | |
2609 | f4m_url, video_id, f4m_id='hds', fatal=False) | |
2610 | for entry in f4m_formats: | |
2611 | entry.update({'extra_param_to_segment_url': hdcore_sign}) | |
2612 | formats.extend(f4m_formats) | |
70c5802b | 2613 | |
c4251b9a RA |
2614 | m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8') |
2615 | hls_host = hosts.get('hls') | |
2616 | if hls_host: | |
2617 | m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url) | |
c7c43a93 RA |
2618 | formats.extend(self._extract_m3u8_formats( |
2619 | m3u8_url, video_id, 'mp4', 'm3u8_native', | |
2620 | m3u8_id='hls', fatal=False)) | |
70c5802b | 2621 | |
2622 | http_host = hosts.get('http') | |
2623 | if http_host and 'hdnea=' not in manifest_url: | |
2624 | REPL_REGEX = r'https://[^/]+/i/([^,]+),([^/]+),([^/]+).csmil/.+' | |
2625 | qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',') | |
2626 | qualities_length = len(qualities) | |
2627 | if len(formats) in (qualities_length + 1, qualities_length * 2 + 1): | |
2628 | i = 0 | |
2629 | http_formats = [] | |
2630 | for f in formats: | |
2631 | if f['protocol'] == 'm3u8_native' and f['vcodec'] != 'none': | |
2632 | for protocol in ('http', 'https'): | |
2633 | http_f = f.copy() | |
2634 | del http_f['manifest_url'] | |
2635 | http_url = re.sub( | |
2636 | REPL_REGEX, protocol + r'://%s/\1%s\3' % (http_host, qualities[i]), f['url']) | |
2637 | http_f.update({ | |
2638 | 'format_id': http_f['format_id'].replace('hls-', protocol + '-'), | |
2639 | 'url': http_url, | |
2640 | 'protocol': protocol, | |
2641 | }) | |
2642 | http_formats.append(http_f) | |
2643 | i += 1 | |
2644 | formats.extend(http_formats) | |
2645 | ||
c7c43a93 RA |
2646 | return formats |
2647 | ||
6ad02195 | 2648 | def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]): |
044eeb14 | 2649 | query = compat_urlparse.urlparse(url).query |
6ad02195 | 2650 | url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url) |
240f2622 S |
2651 | mobj = re.search( |
2652 | r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url) | |
2653 | url_base = mobj.group('url') | |
2654 | http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base) | |
6ad02195 | 2655 | formats = [] |
044eeb14 S |
2656 | |
2657 | def manifest_url(manifest): | |
2658 | m_url = '%s/%s' % (http_base_url, manifest) | |
2659 | if query: | |
2660 | m_url += '?%s' % query | |
2661 | return m_url | |
2662 | ||
6ad02195 RA |
2663 | if 'm3u8' not in skip_protocols: |
2664 | formats.extend(self._extract_m3u8_formats( | |
044eeb14 | 2665 | manifest_url('playlist.m3u8'), video_id, 'mp4', |
6ad02195 RA |
2666 | m3u8_entry_protocol, m3u8_id='hls', fatal=False)) |
2667 | if 'f4m' not in skip_protocols: | |
2668 | formats.extend(self._extract_f4m_formats( | |
044eeb14 | 2669 | manifest_url('manifest.f4m'), |
6ad02195 | 2670 | video_id, f4m_id='hds', fatal=False)) |
0384932e RA |
2671 | if 'dash' not in skip_protocols: |
2672 | formats.extend(self._extract_mpd_formats( | |
044eeb14 | 2673 | manifest_url('manifest.mpd'), |
0384932e | 2674 | video_id, mpd_id='dash', fatal=False)) |
6ad02195 | 2675 | if re.search(r'(?:/smil:|\.smil)', url_base): |
6ad02195 RA |
2676 | if 'smil' not in skip_protocols: |
2677 | rtmp_formats = self._extract_smil_formats( | |
044eeb14 | 2678 | manifest_url('jwplayer.smil'), |
6ad02195 RA |
2679 | video_id, fatal=False) |
2680 | for rtmp_format in rtmp_formats: | |
2681 | rtsp_format = rtmp_format.copy() | |
2682 | rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path']) | |
2683 | del rtsp_format['play_path'] | |
2684 | del rtsp_format['ext'] | |
2685 | rtsp_format.update({ | |
2686 | 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'), | |
2687 | 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'), | |
2688 | 'protocol': 'rtsp', | |
2689 | }) | |
2690 | formats.extend([rtmp_format, rtsp_format]) | |
2691 | else: | |
2692 | for protocol in ('rtmp', 'rtsp'): | |
2693 | if protocol not in skip_protocols: | |
2694 | formats.append({ | |
f2e2f0c7 | 2695 | 'url': '%s:%s' % (protocol, url_base), |
6ad02195 RA |
2696 | 'format_id': protocol, |
2697 | 'protocol': protocol, | |
2698 | }) | |
2699 | return formats | |
2700 | ||
c73e330e | 2701 | def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json): |
a4a554a7 | 2702 | mobj = re.search( |
ac9c69ac | 2703 | r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)', |
a4a554a7 YCH |
2704 | webpage) |
2705 | if mobj: | |
c73e330e RU |
2706 | try: |
2707 | jwplayer_data = self._parse_json(mobj.group('options'), | |
2708 | video_id=video_id, | |
2709 | transform_source=transform_source) | |
2710 | except ExtractorError: | |
2711 | pass | |
2712 | else: | |
2713 | if isinstance(jwplayer_data, dict): | |
2714 | return jwplayer_data | |
a4a554a7 YCH |
2715 | |
2716 | def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs): | |
c73e330e RU |
2717 | jwplayer_data = self._find_jwplayer_data( |
2718 | webpage, video_id, transform_source=js_to_json) | |
a4a554a7 YCH |
2719 | return self._parse_jwplayer_data( |
2720 | jwplayer_data, video_id, *args, **kwargs) | |
2721 | ||
2722 | def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True, | |
2723 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
2724 | # JWPlayer backward compatibility: flattened playlists | |
2725 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96 | |
2726 | if 'playlist' not in jwplayer_data: | |
2727 | jwplayer_data = {'playlist': [jwplayer_data]} | |
2728 | ||
2729 | entries = [] | |
2730 | ||
2731 | # JWPlayer backward compatibility: single playlist item | |
2732 | # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10 | |
2733 | if not isinstance(jwplayer_data['playlist'], list): | |
2734 | jwplayer_data['playlist'] = [jwplayer_data['playlist']] | |
2735 | ||
2736 | for video_data in jwplayer_data['playlist']: | |
2737 | # JWPlayer backward compatibility: flattened sources | |
2738 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35 | |
2739 | if 'sources' not in video_data: | |
2740 | video_data['sources'] = [video_data] | |
2741 | ||
2742 | this_video_id = video_id or video_data['mediaid'] | |
2743 | ||
1a2192cb S |
2744 | formats = self._parse_jwplayer_formats( |
2745 | video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id, | |
2746 | mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url) | |
a4a554a7 YCH |
2747 | |
2748 | subtitles = {} | |
2749 | tracks = video_data.get('tracks') | |
2750 | if tracks and isinstance(tracks, list): | |
2751 | for track in tracks: | |
96a2daa1 S |
2752 | if not isinstance(track, dict): |
2753 | continue | |
f4b74272 S |
2754 | track_kind = track.get('kind') |
2755 | if not track_kind or not isinstance(track_kind, compat_str): | |
2756 | continue | |
2757 | if track_kind.lower() not in ('captions', 'subtitles'): | |
a4a554a7 YCH |
2758 | continue |
2759 | track_url = urljoin(base_url, track.get('file')) | |
2760 | if not track_url: | |
2761 | continue | |
2762 | subtitles.setdefault(track.get('label') or 'en', []).append({ | |
2763 | 'url': self._proto_relative_url(track_url) | |
2764 | }) | |
2765 | ||
50d808f5 | 2766 | entry = { |
a4a554a7 | 2767 | 'id': this_video_id, |
50d808f5 | 2768 | 'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')), |
f81dd65b | 2769 | 'description': clean_html(video_data.get('description')), |
6945b9e7 | 2770 | 'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))), |
a4a554a7 YCH |
2771 | 'timestamp': int_or_none(video_data.get('pubdate')), |
2772 | 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')), | |
2773 | 'subtitles': subtitles, | |
50d808f5 RA |
2774 | } |
2775 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32 | |
2776 | if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']): | |
2777 | entry.update({ | |
2778 | '_type': 'url_transparent', | |
2779 | 'url': formats[0]['url'], | |
2780 | }) | |
2781 | else: | |
2782 | self._sort_formats(formats) | |
2783 | entry['formats'] = formats | |
2784 | entries.append(entry) | |
a4a554a7 YCH |
2785 | if len(entries) == 1: |
2786 | return entries[0] | |
2787 | else: | |
2788 | return self.playlist_result(entries) | |
2789 | ||
ed0cf9b3 S |
2790 | def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None, |
2791 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
bf1b87cd | 2792 | urls = [] |
ed0cf9b3 | 2793 | formats = [] |
1a2192cb | 2794 | for source in jwplayer_sources_data: |
0a268c6e S |
2795 | if not isinstance(source, dict): |
2796 | continue | |
6945b9e7 RA |
2797 | source_url = urljoin( |
2798 | base_url, self._proto_relative_url(source.get('file'))) | |
2799 | if not source_url or source_url in urls: | |
bf1b87cd RA |
2800 | continue |
2801 | urls.append(source_url) | |
ed0cf9b3 S |
2802 | source_type = source.get('type') or '' |
2803 | ext = mimetype2ext(source_type) or determine_ext(source_url) | |
2804 | if source_type == 'hls' or ext == 'm3u8': | |
2805 | formats.extend(self._extract_m3u8_formats( | |
0236cd0d S |
2806 | source_url, video_id, 'mp4', entry_protocol='m3u8_native', |
2807 | m3u8_id=m3u8_id, fatal=False)) | |
0d9c48de | 2808 | elif source_type == 'dash' or ext == 'mpd': |
ed0cf9b3 S |
2809 | formats.extend(self._extract_mpd_formats( |
2810 | source_url, video_id, mpd_id=mpd_id, fatal=False)) | |
b51dc9db S |
2811 | elif ext == 'smil': |
2812 | formats.extend(self._extract_smil_formats( | |
2813 | source_url, video_id, fatal=False)) | |
ed0cf9b3 | 2814 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67 |
0236cd0d S |
2815 | elif source_type.startswith('audio') or ext in ( |
2816 | 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'): | |
ed0cf9b3 S |
2817 | formats.append({ |
2818 | 'url': source_url, | |
2819 | 'vcodec': 'none', | |
2820 | 'ext': ext, | |
2821 | }) | |
2822 | else: | |
2823 | height = int_or_none(source.get('height')) | |
2824 | if height is None: | |
2825 | # Often no height is provided but there is a label in | |
0236cd0d | 2826 | # format like "1080p", "720p SD", or 1080. |
ed0cf9b3 | 2827 | height = int_or_none(self._search_regex( |
0236cd0d | 2828 | r'^(\d{3,4})[pP]?(?:\b|$)', compat_str(source.get('label') or ''), |
ed0cf9b3 S |
2829 | 'height', default=None)) |
2830 | a_format = { | |
2831 | 'url': source_url, | |
2832 | 'width': int_or_none(source.get('width')), | |
2833 | 'height': height, | |
0236cd0d | 2834 | 'tbr': int_or_none(source.get('bitrate')), |
ed0cf9b3 S |
2835 | 'ext': ext, |
2836 | } | |
2837 | if source_url.startswith('rtmp'): | |
2838 | a_format['ext'] = 'flv' | |
ed0cf9b3 S |
2839 | # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as |
2840 | # of jwplayer.flash.swf | |
2841 | rtmp_url_parts = re.split( | |
2842 | r'((?:mp4|mp3|flv):)', source_url, 1) | |
2843 | if len(rtmp_url_parts) == 3: | |
2844 | rtmp_url, prefix, play_path = rtmp_url_parts | |
2845 | a_format.update({ | |
2846 | 'url': rtmp_url, | |
2847 | 'play_path': prefix + play_path, | |
2848 | }) | |
2849 | if rtmp_params: | |
2850 | a_format.update(rtmp_params) | |
2851 | formats.append(a_format) | |
2852 | return formats | |
2853 | ||
f4b1c7ad PH |
2854 | def _live_title(self, name): |
2855 | """ Generate the title for a live video """ | |
2856 | now = datetime.datetime.now() | |
611c1dd9 | 2857 | now_str = now.strftime('%Y-%m-%d %H:%M') |
f4b1c7ad PH |
2858 | return name + ' ' + now_str |
2859 | ||
b14f3a4c PH |
2860 | def _int(self, v, name, fatal=False, **kwargs): |
2861 | res = int_or_none(v, **kwargs) | |
2862 | if 'get_attr' in kwargs: | |
2863 | print(getattr(v, kwargs['get_attr'])) | |
2864 | if res is None: | |
2865 | msg = 'Failed to extract %s: Could not parse value %r' % (name, v) | |
2866 | if fatal: | |
2867 | raise ExtractorError(msg) | |
2868 | else: | |
2869 | self._downloader.report_warning(msg) | |
2870 | return res | |
2871 | ||
2872 | def _float(self, v, name, fatal=False, **kwargs): | |
2873 | res = float_or_none(v, **kwargs) | |
2874 | if res is None: | |
2875 | msg = 'Failed to extract %s: Could not parse value %r' % (name, v) | |
2876 | if fatal: | |
2877 | raise ExtractorError(msg) | |
2878 | else: | |
2879 | self._downloader.report_warning(msg) | |
2880 | return res | |
2881 | ||
40e41780 TF |
2882 | def _set_cookie(self, domain, name, value, expire_time=None, port=None, |
2883 | path='/', secure=False, discard=False, rest={}, **kwargs): | |
6c22cee6 | 2884 | cookie = compat_cookiejar_Cookie( |
4ed2d7b7 | 2885 | 0, name, value, port, port is not None, domain, True, |
40e41780 TF |
2886 | domain.startswith('.'), path, True, secure, expire_time, |
2887 | discard, None, None, rest) | |
42939b61 JMF |
2888 | self._downloader.cookiejar.set_cookie(cookie) |
2889 | ||
799207e8 | 2890 | def _get_cookies(self, url): |
2891 | """ Return a compat_cookies.SimpleCookie with the cookies for the url """ | |
5c2266df | 2892 | req = sanitized_Request(url) |
799207e8 | 2893 | self._downloader.cookiejar.add_cookie_header(req) |
2894 | return compat_cookies.SimpleCookie(req.get_header('Cookie')) | |
2895 | ||
e3c1266f | 2896 | def _apply_first_set_cookie_header(self, url_handle, cookie): |
ce2fe4c0 S |
2897 | """ |
2898 | Apply first Set-Cookie header instead of the last. Experimental. | |
2899 | ||
2900 | Some sites (e.g. [1-3]) may serve two cookies under the same name | |
2901 | in Set-Cookie header and expect the first (old) one to be set rather | |
2902 | than second (new). However, as of RFC6265 the newer one cookie | |
2903 | should be set into cookie store what actually happens. | |
2904 | We will workaround this issue by resetting the cookie to | |
2905 | the first one manually. | |
2906 | 1. https://new.vk.com/ | |
2907 | 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201 | |
2908 | 3. https://learning.oreilly.com/ | |
2909 | """ | |
e3c1266f S |
2910 | for header, cookies in url_handle.headers.items(): |
2911 | if header.lower() != 'set-cookie': | |
2912 | continue | |
2913 | if sys.version_info[0] >= 3: | |
2914 | cookies = cookies.encode('iso-8859-1') | |
2915 | cookies = cookies.decode('utf-8') | |
2916 | cookie_value = re.search( | |
2917 | r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies) | |
2918 | if cookie_value: | |
2919 | value, domain = cookie_value.groups() | |
2920 | self._set_cookie(domain, cookie, value) | |
2921 | break | |
2922 | ||
05900629 PH |
2923 | def get_testcases(self, include_onlymatching=False): |
2924 | t = getattr(self, '_TEST', None) | |
2925 | if t: | |
2926 | assert not hasattr(self, '_TESTS'), \ | |
2927 | '%s has _TEST and _TESTS' % type(self).__name__ | |
2928 | tests = [t] | |
2929 | else: | |
2930 | tests = getattr(self, '_TESTS', []) | |
2931 | for t in tests: | |
2932 | if not include_onlymatching and t.get('only_matching', False): | |
2933 | continue | |
2934 | t['name'] = type(self).__name__[:-len('IE')] | |
2935 | yield t | |
2936 | ||
2937 | def is_suitable(self, age_limit): | |
2938 | """ Test whether the extractor is generally suitable for the given | |
2939 | age limit (i.e. pornographic sites are not, all others usually are) """ | |
2940 | ||
2941 | any_restricted = False | |
2942 | for tc in self.get_testcases(include_onlymatching=False): | |
40090e8d | 2943 | if tc.get('playlist', []): |
05900629 PH |
2944 | tc = tc['playlist'][0] |
2945 | is_restricted = age_restricted( | |
2946 | tc.get('info_dict', {}).get('age_limit'), age_limit) | |
2947 | if not is_restricted: | |
2948 | return True | |
2949 | any_restricted = any_restricted or is_restricted | |
2950 | return not any_restricted | |
2951 | ||
a504ced0 | 2952 | def extract_subtitles(self, *args, **kwargs): |
3089bc74 S |
2953 | if (self._downloader.params.get('writesubtitles', False) |
2954 | or self._downloader.params.get('listsubtitles')): | |
9868ea49 JMF |
2955 | return self._get_subtitles(*args, **kwargs) |
2956 | return {} | |
a504ced0 JMF |
2957 | |
2958 | def _get_subtitles(self, *args, **kwargs): | |
611c1dd9 | 2959 | raise NotImplementedError('This method must be implemented by subclasses') |
a504ced0 | 2960 | |
912e0b7e YCH |
2961 | @staticmethod |
2962 | def _merge_subtitle_items(subtitle_list1, subtitle_list2): | |
2963 | """ Merge subtitle items for one language. Items with duplicated URLs | |
2964 | will be dropped. """ | |
2965 | list1_urls = set([item['url'] for item in subtitle_list1]) | |
2966 | ret = list(subtitle_list1) | |
2967 | ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls]) | |
2968 | return ret | |
2969 | ||
2970 | @classmethod | |
8c97f819 | 2971 | def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2): |
912e0b7e | 2972 | """ Merge two subtitle dictionaries, language by language. """ |
912e0b7e YCH |
2973 | ret = dict(subtitle_dict1) |
2974 | for lang in subtitle_dict2: | |
8c97f819 | 2975 | ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang]) |
912e0b7e YCH |
2976 | return ret |
2977 | ||
360e1ca5 | 2978 | def extract_automatic_captions(self, *args, **kwargs): |
3089bc74 S |
2979 | if (self._downloader.params.get('writeautomaticsub', False) |
2980 | or self._downloader.params.get('listsubtitles')): | |
9868ea49 JMF |
2981 | return self._get_automatic_captions(*args, **kwargs) |
2982 | return {} | |
360e1ca5 JMF |
2983 | |
2984 | def _get_automatic_captions(self, *args, **kwargs): | |
611c1dd9 | 2985 | raise NotImplementedError('This method must be implemented by subclasses') |
360e1ca5 | 2986 | |
d77ab8e2 | 2987 | def mark_watched(self, *args, **kwargs): |
3089bc74 S |
2988 | if (self._downloader.params.get('mark_watched', False) |
2989 | and (self._get_login_info()[0] is not None | |
2990 | or self._downloader.params.get('cookiefile') is not None)): | |
d77ab8e2 S |
2991 | self._mark_watched(*args, **kwargs) |
2992 | ||
2993 | def _mark_watched(self, *args, **kwargs): | |
2994 | raise NotImplementedError('This method must be implemented by subclasses') | |
2995 | ||
38cce791 YCH |
2996 | def geo_verification_headers(self): |
2997 | headers = {} | |
2998 | geo_verification_proxy = self._downloader.params.get('geo_verification_proxy') | |
2999 | if geo_verification_proxy: | |
3000 | headers['Ytdl-request-proxy'] = geo_verification_proxy | |
3001 | return headers | |
3002 | ||
98763ee3 YCH |
3003 | def _generic_id(self, url): |
3004 | return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0]) | |
3005 | ||
3006 | def _generic_title(self, url): | |
3007 | return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0]) | |
3008 | ||
8dbe9899 | 3009 | |
d6983cb4 PH |
3010 | class SearchInfoExtractor(InfoExtractor): |
3011 | """ | |
3012 | Base class for paged search queries extractors. | |
10952eb2 | 3013 | They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query} |
d6983cb4 PH |
3014 | Instances should define _SEARCH_KEY and _MAX_RESULTS. |
3015 | """ | |
3016 | ||
3017 | @classmethod | |
3018 | def _make_valid_url(cls): | |
3019 | return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY | |
3020 | ||
3021 | @classmethod | |
3022 | def suitable(cls, url): | |
3023 | return re.match(cls._make_valid_url(), url) is not None | |
3024 | ||
3025 | def _real_extract(self, query): | |
3026 | mobj = re.match(self._make_valid_url(), query) | |
3027 | if mobj is None: | |
f1a9d64e | 3028 | raise ExtractorError('Invalid search query "%s"' % query) |
d6983cb4 PH |
3029 | |
3030 | prefix = mobj.group('prefix') | |
3031 | query = mobj.group('query') | |
3032 | if prefix == '': | |
3033 | return self._get_n_results(query, 1) | |
3034 | elif prefix == 'all': | |
3035 | return self._get_n_results(query, self._MAX_RESULTS) | |
3036 | else: | |
3037 | n = int(prefix) | |
3038 | if n <= 0: | |
f1a9d64e | 3039 | raise ExtractorError('invalid download number %s for query "%s"' % (n, query)) |
d6983cb4 | 3040 | elif n > self._MAX_RESULTS: |
f1a9d64e | 3041 | self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n)) |
d6983cb4 PH |
3042 | n = self._MAX_RESULTS |
3043 | return self._get_n_results(query, n) | |
3044 | ||
3045 | def _get_n_results(self, query, n): | |
3046 | """Get a specified number of results for a query""" | |
611c1dd9 | 3047 | raise NotImplementedError('This method must be implemented by subclasses') |
0f818663 PH |
3048 | |
3049 | @property | |
3050 | def SEARCH_KEY(self): | |
3051 | return self._SEARCH_KEY |