]>
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 | |
13 | import sys | |
4094b6e3 | 14 | import time |
1bac3455 | 15 | import math |
d6983cb4 | 16 | |
8c25f81b | 17 | from ..compat import ( |
42939b61 | 18 | compat_cookiejar, |
799207e8 | 19 | compat_cookies, |
ee0ba927 | 20 | compat_etree_Element, |
e9c0cdd3 | 21 | compat_etree_fromstring, |
e64b7569 | 22 | compat_getpass, |
d391b7e2 | 23 | compat_integer_types, |
d6983cb4 | 24 | compat_http_client, |
e9c0cdd3 YCH |
25 | compat_os_name, |
26 | compat_str, | |
d6983cb4 | 27 | compat_urllib_error, |
98763ee3 | 28 | compat_urllib_parse_unquote, |
15707c7e | 29 | compat_urllib_parse_urlencode, |
41d06b04 | 30 | compat_urllib_request, |
f0b5d6af | 31 | compat_urlparse, |
e01c3d2e | 32 | compat_xml_parse_error, |
8c25f81b | 33 | ) |
48107c19 S |
34 | from ..downloader.f4m import ( |
35 | get_base_url, | |
36 | remove_encrypted_media, | |
37 | ) | |
8c25f81b | 38 | from ..utils import ( |
c342041f | 39 | NO_DEFAULT, |
05900629 | 40 | age_restricted, |
02dc0a36 | 41 | base_url, |
08f2a92c | 42 | bug_reports_message, |
d6983cb4 PH |
43 | clean_html, |
44 | compiled_regex_type, | |
70f0f5a8 | 45 | determine_ext, |
46b18f23 | 46 | determine_protocol, |
d493f15c | 47 | dict_get, |
9b9c5355 | 48 | error_to_compat_str, |
d6983cb4 | 49 | ExtractorError, |
46b18f23 | 50 | extract_attributes, |
97f4aecf | 51 | fix_xml_ampersands, |
b14f3a4c | 52 | float_or_none, |
773f291d S |
53 | GeoRestrictedError, |
54 | GeoUtils, | |
31bb8d3f | 55 | int_or_none, |
a4a554a7 | 56 | js_to_json, |
0685d972 | 57 | JSON_LD_RE, |
46b18f23 JH |
58 | mimetype2ext, |
59 | orderedSet, | |
d493f15c | 60 | parse_bitrate, |
46b18f23 JH |
61 | parse_codecs, |
62 | parse_duration, | |
4ca2a3cf | 63 | parse_iso8601, |
46b18f23 | 64 | parse_m3u8_attributes, |
d493f15c | 65 | parse_resolution, |
55b3e45b | 66 | RegexNotFoundError, |
5c2266df | 67 | sanitized_Request, |
46b18f23 | 68 | sanitize_filename, |
d493f15c | 69 | str_or_none, |
f856816b | 70 | strip_or_none, |
f38de77f | 71 | unescapeHTML, |
647eab45 | 72 | unified_strdate, |
6b3a3098 | 73 | unified_timestamp, |
46b18f23 JH |
74 | update_Request, |
75 | update_url_query, | |
76 | urljoin, | |
a107193e | 77 | url_basename, |
bebef109 | 78 | url_or_none, |
a6571f10 | 79 | xpath_element, |
8d6765cf S |
80 | xpath_text, |
81 | xpath_with_ns, | |
d6983cb4 | 82 | ) |
c342041f | 83 | |
d6983cb4 PH |
84 | |
85 | class InfoExtractor(object): | |
86 | """Information Extractor class. | |
87 | ||
88 | Information extractors are the classes that, given a URL, extract | |
89 | information about the video (or videos) the URL refers to. This | |
90 | information includes the real video URL, the video title, author and | |
91 | others. The information is stored in a dictionary which is then | |
5d380852 | 92 | passed to the YoutubeDL. The YoutubeDL processes this |
d6983cb4 PH |
93 | information possibly downloading the video to the file system, among |
94 | other possible outcomes. | |
95 | ||
cf0649f8 | 96 | The type field determines the type of the result. |
fed5d032 PH |
97 | By far the most common value (and the default if _type is missing) is |
98 | "video", which indicates a single video. | |
99 | ||
100 | For a video, the dictionaries must include the following fields: | |
d6983cb4 PH |
101 | |
102 | id: Video identifier. | |
d6983cb4 | 103 | title: Video title, unescaped. |
d67b0b15 | 104 | |
f49d89ee | 105 | Additionally, it must contain either a formats entry or a url one: |
d67b0b15 | 106 | |
f49d89ee PH |
107 | formats: A list of dictionaries for each format available, ordered |
108 | from worst to best quality. | |
109 | ||
110 | Potential fields: | |
c790e93a S |
111 | * url The mandatory URL representing the media: |
112 | for plain file media - HTTP URL of this file, | |
113 | for RTMP - RTMP URL, | |
114 | for HLS - URL of the M3U8 media playlist, | |
115 | for HDS - URL of the F4M manifest, | |
79d2077e S |
116 | for DASH |
117 | - HTTP URL to plain file media (in case of | |
118 | unfragmented media) | |
119 | - URL of the MPD manifest or base URL | |
120 | representing the media if MPD manifest | |
8ed7a233 | 121 | is parsed from a string (in case of |
79d2077e | 122 | fragmented media) |
c790e93a | 123 | for MSS - URL of the ISM manifest. |
86f4d14f S |
124 | * manifest_url |
125 | The URL of the manifest file in case of | |
c790e93a S |
126 | fragmented media: |
127 | for HLS - URL of the M3U8 master playlist, | |
128 | for HDS - URL of the F4M manifest, | |
129 | for DASH - URL of the MPD manifest, | |
130 | for MSS - URL of the ISM manifest. | |
10952eb2 | 131 | * ext Will be calculated from URL if missing |
d67b0b15 PH |
132 | * format A human-readable description of the format |
133 | ("mp4 container with h264/opus"). | |
134 | Calculated from the format_id, width, height. | |
135 | and format_note fields if missing. | |
136 | * format_id A short description of the format | |
5d4f3985 PH |
137 | ("mp4_h264_opus" or "19"). |
138 | Technically optional, but strongly recommended. | |
d67b0b15 PH |
139 | * format_note Additional info about the format |
140 | ("3D" or "DASH video") | |
141 | * width Width of the video, if known | |
142 | * height Height of the video, if known | |
f49d89ee | 143 | * resolution Textual description of width and height |
7217e148 | 144 | * tbr Average bitrate of audio and video in KBit/s |
d67b0b15 PH |
145 | * abr Average audio bitrate in KBit/s |
146 | * acodec Name of the audio codec in use | |
dd27fd17 | 147 | * asr Audio sampling rate in Hertz |
d67b0b15 | 148 | * vbr Average video bitrate in KBit/s |
fbb21cf5 | 149 | * fps Frame rate |
d67b0b15 | 150 | * vcodec Name of the video codec in use |
1394ce65 | 151 | * container Name of the container format |
d67b0b15 | 152 | * filesize The number of bytes, if known in advance |
9732d77e | 153 | * filesize_approx An estimate for the number of bytes |
d67b0b15 | 154 | * player_url SWF Player URL (used for rtmpdump). |
c7deaa4c PH |
155 | * protocol The protocol that will be used for the actual |
156 | download, lower-case. | |
b04b8852 | 157 | "http", "https", "rtsp", "rtmp", "rtmpe", |
af7d5a63 | 158 | "m3u8", "m3u8_native" or "http_dash_segments". |
c58c2d63 S |
159 | * fragment_base_url |
160 | Base URL for fragments. Each fragment's path | |
161 | value (if present) will be relative to | |
162 | this URL. | |
163 | * fragments A list of fragments of a fragmented media. | |
164 | Each fragment entry must contain either an url | |
165 | or a path. If an url is present it should be | |
166 | considered by a client. Otherwise both path and | |
167 | fragment_base_url must be present. Here is | |
168 | the list of all potential fields: | |
169 | * "url" - fragment's URL | |
170 | * "path" - fragment's path relative to | |
171 | fragment_base_url | |
a0d5077c S |
172 | * "duration" (optional, int or float) |
173 | * "filesize" (optional, int) | |
f49d89ee | 174 | * preference Order number of this format. If this field is |
08d13955 | 175 | present and not None, the formats get sorted |
38d63d84 | 176 | by this field, regardless of all other values. |
f49d89ee PH |
177 | -1 for default (order by other properties), |
178 | -2 or smaller for less than default. | |
e65566a9 PH |
179 | < -1000 to hide the format (if there is |
180 | another one which is strictly better) | |
32f90364 PH |
181 | * language Language code, e.g. "de" or "en-US". |
182 | * language_preference Is this in the language mentioned in | |
183 | the URL? | |
aff2f4f4 PH |
184 | 10 if it's what the URL is about, |
185 | -1 for default (don't know), | |
186 | -10 otherwise, other values reserved for now. | |
5d73273f PH |
187 | * quality Order number of the video quality of this |
188 | format, irrespective of the file format. | |
189 | -1 for default (order by other properties), | |
190 | -2 or smaller for less than default. | |
c64ed2a3 PH |
191 | * source_preference Order number for this video source |
192 | (quality takes higher priority) | |
193 | -1 for default (order by other properties), | |
194 | -2 or smaller for less than default. | |
d769be6c PH |
195 | * http_headers A dictionary of additional HTTP headers |
196 | to add to the request. | |
6271f1ca | 197 | * stretched_ratio If given and not 1, indicates that the |
3dee7826 PH |
198 | video's pixels are not square. |
199 | width : height ratio as float. | |
200 | * no_resume The server does not support resuming the | |
201 | (HTTP or RTMP) download. Boolean. | |
00c97e3e S |
202 | * downloader_options A dictionary of downloader options as |
203 | described in FileDownloader | |
3dee7826 | 204 | |
c0ba0f48 | 205 | url: Final video URL. |
d6983cb4 | 206 | ext: Video filename extension. |
d67b0b15 PH |
207 | format: The video format, defaults to ext (used for --get-format) |
208 | player_url: SWF Player URL (used for rtmpdump). | |
2f5865cc | 209 | |
d6983cb4 PH |
210 | The following fields are optional: |
211 | ||
f5e43bc6 | 212 | alt_title: A secondary title of the video. |
0afef30b PH |
213 | display_id An alternative identifier for the video, not necessarily |
214 | unique, but available before title. Typically, id is | |
215 | something like "4234987", title "Dancing naked mole rats", | |
216 | and display_id "dancing-naked-mole-rats" | |
d5519808 | 217 | thumbnails: A list of dictionaries, with the following entries: |
cfb56d1a | 218 | * "id" (optional, string) - Thumbnail format ID |
d5519808 | 219 | * "url" |
cfb56d1a | 220 | * "preference" (optional, int) - quality of the image |
d5519808 PH |
221 | * "width" (optional, int) |
222 | * "height" (optional, int) | |
5e1c39ac | 223 | * "resolution" (optional, string "{width}x{height}", |
d5519808 | 224 | deprecated) |
2de624fd | 225 | * "filesize" (optional, int) |
d6983cb4 | 226 | thumbnail: Full URL to a video thumbnail image. |
f5e43bc6 | 227 | description: Full video description. |
d6983cb4 | 228 | uploader: Full name of the video uploader. |
2bc0c46f | 229 | license: License name the video is licensed under. |
8a92e51c | 230 | creator: The creator of the video. |
8aab976b | 231 | release_date: The date (YYYYMMDD) when the video was released. |
955c4514 | 232 | timestamp: UNIX timestamp of the moment the video became available. |
d6983cb4 | 233 | upload_date: Video upload date (YYYYMMDD). |
955c4514 | 234 | If not explicitly set, calculated from timestamp. |
d6983cb4 | 235 | uploader_id: Nickname or id of the video uploader. |
7bcd2830 | 236 | uploader_url: Full URL to a personal webpage of the video uploader. |
6f1f59f3 | 237 | channel: Full name of the channel the video is uploaded on. |
0e7b8d3e | 238 | Note that channel fields may or may not repeat uploader |
6f1f59f3 S |
239 | fields. This depends on a particular extractor. |
240 | channel_id: Id of the channel. | |
241 | channel_url: Full URL to a channel webpage. | |
da9ec3b9 | 242 | location: Physical location where the video was filmed. |
a504ced0 | 243 | subtitles: The available subtitles as a dictionary in the format |
4606c34e YCH |
244 | {tag: subformats}. "tag" is usually a language code, and |
245 | "subformats" is a list sorted from lower to higher | |
246 | preference, each element is a dictionary with the "ext" | |
247 | entry and one of: | |
a504ced0 | 248 | * "data": The subtitles file contents |
10952eb2 | 249 | * "url": A URL pointing to the subtitles file |
4bba3716 | 250 | "ext" will be calculated from URL if missing |
360e1ca5 JMF |
251 | automatic_captions: Like 'subtitles', used by the YoutubeIE for |
252 | automatically generated captions | |
62d231c0 | 253 | duration: Length of the video in seconds, as an integer or float. |
f3d29461 | 254 | view_count: How many users have watched the video on the platform. |
19e3dfc9 PH |
255 | like_count: Number of positive ratings of the video |
256 | dislike_count: Number of negative ratings of the video | |
02835c6b | 257 | repost_count: Number of reposts of the video |
2d30521a | 258 | average_rating: Average rating give by users, the scale used depends on the webpage |
19e3dfc9 | 259 | comment_count: Number of comments on the video |
dd622d7c PH |
260 | comments: A list of comments, each with one or more of the following |
261 | properties (all but one of text or html optional): | |
262 | * "author" - human-readable name of the comment author | |
263 | * "author_id" - user ID of the comment author | |
264 | * "id" - Comment ID | |
265 | * "html" - Comment as HTML | |
266 | * "text" - Plain text of the comment | |
267 | * "timestamp" - UNIX timestamp of comment | |
268 | * "parent" - ID of the comment this one is replying to. | |
269 | Set to "root" to indicate that this is a | |
270 | comment to the original video. | |
8dbe9899 | 271 | age_limit: Age restriction for the video, as an integer (years) |
10952eb2 | 272 | webpage_url: The URL to the video webpage, if given to youtube-dl it |
9103bbc5 JMF |
273 | should allow to get the same result again. (It will be set |
274 | by YoutubeDL if it's missing) | |
ad3bc6ac PH |
275 | categories: A list of categories that the video falls in, for example |
276 | ["Sports", "Berlin"] | |
864f24bd | 277 | tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"] |
7267bd53 PH |
278 | is_live: True, False, or None (=unknown). Whether this video is a |
279 | live stream that goes on instead of a fixed-length video. | |
7c80519c | 280 | start_time: Time in seconds where the reproduction should start, as |
10952eb2 | 281 | specified in the URL. |
297a564b | 282 | end_time: Time in seconds where the reproduction should end, as |
10952eb2 | 283 | specified in the URL. |
55949fed | 284 | chapters: A list of dictionaries, with the following entries: |
285 | * "start_time" - The start time of the chapter in seconds | |
286 | * "end_time" - The end time of the chapter in seconds | |
287 | * "title" (optional, string) | |
d6983cb4 | 288 | |
7109903e S |
289 | The following fields should only be used when the video belongs to some logical |
290 | chapter or section: | |
291 | ||
292 | chapter: Name or title of the chapter the video belongs to. | |
27bfd4e5 S |
293 | chapter_number: Number of the chapter the video belongs to, as an integer. |
294 | chapter_id: Id of the chapter the video belongs to, as a unicode string. | |
7109903e S |
295 | |
296 | The following fields should only be used when the video is an episode of some | |
8d76bdf1 | 297 | series, programme or podcast: |
7109903e S |
298 | |
299 | series: Title of the series or programme the video episode belongs to. | |
300 | season: Title of the season the video episode belongs to. | |
27bfd4e5 S |
301 | season_number: Number of the season the video episode belongs to, as an integer. |
302 | season_id: Id of the season the video episode belongs to, as a unicode string. | |
7109903e S |
303 | episode: Title of the video episode. Unlike mandatory video title field, |
304 | this field should denote the exact title of the video episode | |
305 | without any kind of decoration. | |
27bfd4e5 S |
306 | episode_number: Number of the video episode within a season, as an integer. |
307 | episode_id: Id of the video episode, as a unicode string. | |
7109903e | 308 | |
7a93ab5f S |
309 | The following fields should only be used when the media is a track or a part of |
310 | a music album: | |
311 | ||
312 | track: Title of the track. | |
313 | track_number: Number of the track within an album or a disc, as an integer. | |
314 | track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii), | |
315 | as a unicode string. | |
316 | artist: Artist(s) of the track. | |
317 | genre: Genre(s) of the track. | |
318 | album: Title of the album the track belongs to. | |
319 | album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc). | |
320 | album_artist: List of all artists appeared on the album (e.g. | |
321 | "Ash Borer / Fell Voices" or "Various Artists", useful for splits | |
322 | and compilations). | |
323 | disc_number: Number of the disc or other physical medium the track belongs to, | |
324 | as an integer. | |
325 | release_year: Year (YYYY) when the album was released. | |
326 | ||
deefc05b | 327 | Unless mentioned otherwise, the fields should be Unicode strings. |
d6983cb4 | 328 | |
d838b1bd PH |
329 | Unless mentioned otherwise, None is equivalent to absence of information. |
330 | ||
fed5d032 PH |
331 | |
332 | _type "playlist" indicates multiple videos. | |
b82f815f PH |
333 | There must be a key "entries", which is a list, an iterable, or a PagedList |
334 | object, each element of which is a valid dictionary by this specification. | |
fed5d032 | 335 | |
c10c9323 S |
336 | Additionally, playlists can have "id", "title", "description", "uploader", |
337 | "uploader_id", "uploader_url" attributes with the same semantics as videos | |
338 | (see above). | |
fed5d032 PH |
339 | |
340 | ||
341 | _type "multi_video" indicates that there are multiple videos that | |
342 | form a single show, for examples multiple acts of an opera or TV episode. | |
343 | It must have an entries key like a playlist and contain all the keys | |
344 | required for a video at the same time. | |
345 | ||
346 | ||
347 | _type "url" indicates that the video must be extracted from another | |
348 | location, possibly by a different extractor. Its only required key is: | |
349 | "url" - the next URL to extract. | |
f58766ce PH |
350 | The key "ie_key" can be set to the class name (minus the trailing "IE", |
351 | e.g. "Youtube") if the extractor class is known in advance. | |
352 | Additionally, the dictionary may have any properties of the resolved entity | |
353 | known in advance, for example "title" if the title of the referred video is | |
fed5d032 PH |
354 | known ahead of time. |
355 | ||
356 | ||
357 | _type "url_transparent" entities have the same specification as "url", but | |
358 | indicate that the given additional information is more precise than the one | |
359 | associated with the resolved URL. | |
360 | This is useful when a site employs a video service that hosts the video and | |
361 | its technical metadata, but that video service does not embed a useful | |
362 | title, description etc. | |
363 | ||
364 | ||
d6983cb4 PH |
365 | Subclasses of this one should re-define the _real_initialize() and |
366 | _real_extract() methods and define a _VALID_URL regexp. | |
367 | Probably, they should also be added to the list of extractors. | |
368 | ||
4248dad9 | 369 | _GEO_BYPASS attribute may be set to False in order to disable |
773f291d S |
370 | geo restriction bypass mechanisms for a particular extractor. |
371 | Though it won't disable explicit geo restriction bypass based on | |
504f20dd | 372 | country code provided with geo_bypass_country. |
4248dad9 S |
373 | |
374 | _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted | |
375 | countries for this extractor. One of these countries will be used by | |
376 | geo restriction bypass mechanism right away in order to bypass | |
504f20dd | 377 | geo restriction, of course, if the mechanism is not disabled. |
773f291d | 378 | |
5f95927a S |
379 | _GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted |
380 | IP blocks in CIDR notation for this extractor. One of these IP blocks | |
381 | will be used by geo restriction bypass mechanism similarly | |
504f20dd | 382 | to _GEO_COUNTRIES. |
3ccdde8c | 383 | |
d6983cb4 PH |
384 | Finally, the _WORKING attribute should be set to False for broken IEs |
385 | in order to warn the users and skip the tests. | |
386 | """ | |
387 | ||
388 | _ready = False | |
389 | _downloader = None | |
773f291d | 390 | _x_forwarded_for_ip = None |
4248dad9 S |
391 | _GEO_BYPASS = True |
392 | _GEO_COUNTRIES = None | |
5f95927a | 393 | _GEO_IP_BLOCKS = None |
d6983cb4 PH |
394 | _WORKING = True |
395 | ||
396 | def __init__(self, downloader=None): | |
397 | """Constructor. Receives an optional downloader.""" | |
398 | self._ready = False | |
773f291d | 399 | self._x_forwarded_for_ip = None |
d6983cb4 PH |
400 | self.set_downloader(downloader) |
401 | ||
402 | @classmethod | |
403 | def suitable(cls, url): | |
404 | """Receives a URL and returns True if suitable for this IE.""" | |
79cb2577 PH |
405 | |
406 | # This does not use has/getattr intentionally - we want to know whether | |
407 | # we have cached the regexp for *this* class, whereas getattr would also | |
408 | # match the superclass | |
409 | if '_VALID_URL_RE' not in cls.__dict__: | |
410 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) | |
411 | return cls._VALID_URL_RE.match(url) is not None | |
d6983cb4 | 412 | |
ed9266db PH |
413 | @classmethod |
414 | def _match_id(cls, url): | |
415 | if '_VALID_URL_RE' not in cls.__dict__: | |
416 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) | |
417 | m = cls._VALID_URL_RE.match(url) | |
418 | assert m | |
1afd0b0d | 419 | return compat_str(m.group('id')) |
ed9266db | 420 | |
d6983cb4 PH |
421 | @classmethod |
422 | def working(cls): | |
423 | """Getter method for _WORKING.""" | |
424 | return cls._WORKING | |
425 | ||
426 | def initialize(self): | |
427 | """Initializes an instance (authentication, etc).""" | |
5f95927a S |
428 | self._initialize_geo_bypass({ |
429 | 'countries': self._GEO_COUNTRIES, | |
430 | 'ip_blocks': self._GEO_IP_BLOCKS, | |
431 | }) | |
4248dad9 S |
432 | if not self._ready: |
433 | self._real_initialize() | |
434 | self._ready = True | |
435 | ||
5f95927a | 436 | def _initialize_geo_bypass(self, geo_bypass_context): |
e39b5d4a S |
437 | """ |
438 | Initialize geo restriction bypass mechanism. | |
439 | ||
440 | This method is used to initialize geo bypass mechanism based on faking | |
441 | X-Forwarded-For HTTP header. A random country from provided country list | |
dc0a869e | 442 | is selected and a random IP belonging to this country is generated. This |
e39b5d4a S |
443 | IP will be passed as X-Forwarded-For HTTP header in all subsequent |
444 | HTTP requests. | |
e39b5d4a S |
445 | |
446 | This method will be used for initial geo bypass mechanism initialization | |
5f95927a S |
447 | during the instance initialization with _GEO_COUNTRIES and |
448 | _GEO_IP_BLOCKS. | |
e39b5d4a | 449 | |
5f95927a | 450 | You may also manually call it from extractor's code if geo bypass |
e39b5d4a | 451 | information is not available beforehand (e.g. obtained during |
5f95927a S |
452 | extraction) or due to some other reason. In this case you should pass |
453 | this information in geo bypass context passed as first argument. It may | |
454 | contain following fields: | |
455 | ||
456 | countries: List of geo unrestricted countries (similar | |
457 | to _GEO_COUNTRIES) | |
458 | ip_blocks: List of geo unrestricted IP blocks in CIDR notation | |
459 | (similar to _GEO_IP_BLOCKS) | |
460 | ||
e39b5d4a | 461 | """ |
773f291d | 462 | if not self._x_forwarded_for_ip: |
5f95927a S |
463 | |
464 | # Geo bypass mechanism is explicitly disabled by user | |
465 | if not self._downloader.params.get('geo_bypass', True): | |
466 | return | |
467 | ||
468 | if not geo_bypass_context: | |
469 | geo_bypass_context = {} | |
470 | ||
471 | # Backward compatibility: previously _initialize_geo_bypass | |
472 | # expected a list of countries, some 3rd party code may still use | |
473 | # it this way | |
474 | if isinstance(geo_bypass_context, (list, tuple)): | |
475 | geo_bypass_context = { | |
476 | 'countries': geo_bypass_context, | |
477 | } | |
478 | ||
479 | # The whole point of geo bypass mechanism is to fake IP | |
480 | # as X-Forwarded-For HTTP header based on some IP block or | |
481 | # country code. | |
482 | ||
483 | # Path 1: bypassing based on IP block in CIDR notation | |
484 | ||
485 | # Explicit IP block specified by user, use it right away | |
486 | # regardless of whether extractor is geo bypassable or not | |
487 | ip_block = self._downloader.params.get('geo_bypass_ip_block', None) | |
488 | ||
489 | # Otherwise use random IP block from geo bypass context but only | |
490 | # if extractor is known as geo bypassable | |
491 | if not ip_block: | |
492 | ip_blocks = geo_bypass_context.get('ip_blocks') | |
493 | if self._GEO_BYPASS and ip_blocks: | |
494 | ip_block = random.choice(ip_blocks) | |
495 | ||
496 | if ip_block: | |
497 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(ip_block) | |
498 | if self._downloader.params.get('verbose', False): | |
499 | self._downloader.to_screen( | |
500 | '[debug] Using fake IP %s as X-Forwarded-For.' | |
501 | % self._x_forwarded_for_ip) | |
502 | return | |
503 | ||
504 | # Path 2: bypassing based on country code | |
505 | ||
506 | # Explicit country code specified by user, use it right away | |
507 | # regardless of whether extractor is geo bypassable or not | |
508 | country = self._downloader.params.get('geo_bypass_country', None) | |
509 | ||
510 | # Otherwise use random country code from geo bypass context but | |
511 | # only if extractor is known as geo bypassable | |
512 | if not country: | |
513 | countries = geo_bypass_context.get('countries') | |
514 | if self._GEO_BYPASS and countries: | |
515 | country = random.choice(countries) | |
516 | ||
517 | if country: | |
518 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country) | |
4248dad9 | 519 | if self._downloader.params.get('verbose', False): |
6a9cb295 | 520 | self._downloader.to_screen( |
eea0716c | 521 | '[debug] Using fake IP %s (%s) as X-Forwarded-For.' |
5f95927a | 522 | % (self._x_forwarded_for_ip, country.upper())) |
d6983cb4 PH |
523 | |
524 | def extract(self, url): | |
525 | """Extracts URL information and returns it in list of dicts.""" | |
3a5bcd03 | 526 | try: |
773f291d S |
527 | for _ in range(2): |
528 | try: | |
529 | self.initialize() | |
0016b84e S |
530 | ie_result = self._real_extract(url) |
531 | if self._x_forwarded_for_ip: | |
532 | ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip | |
533 | return ie_result | |
773f291d | 534 | except GeoRestrictedError as e: |
4248dad9 S |
535 | if self.__maybe_fake_ip_and_retry(e.countries): |
536 | continue | |
773f291d | 537 | raise |
3a5bcd03 PH |
538 | except ExtractorError: |
539 | raise | |
540 | except compat_http_client.IncompleteRead as e: | |
dfb1b146 | 541 | raise ExtractorError('A network error has occurred.', cause=e, expected=True) |
9650885b | 542 | except (KeyError, StopIteration) as e: |
dfb1b146 | 543 | raise ExtractorError('An extractor error has occurred.', cause=e) |
d6983cb4 | 544 | |
4248dad9 | 545 | def __maybe_fake_ip_and_retry(self, countries): |
3089bc74 S |
546 | if (not self._downloader.params.get('geo_bypass_country', None) |
547 | and self._GEO_BYPASS | |
548 | and self._downloader.params.get('geo_bypass', True) | |
549 | and not self._x_forwarded_for_ip | |
550 | and countries): | |
eea0716c S |
551 | country_code = random.choice(countries) |
552 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) | |
4248dad9 S |
553 | if self._x_forwarded_for_ip: |
554 | self.report_warning( | |
eea0716c S |
555 | 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.' |
556 | % (self._x_forwarded_for_ip, country_code.upper())) | |
4248dad9 S |
557 | return True |
558 | return False | |
559 | ||
d6983cb4 PH |
560 | def set_downloader(self, downloader): |
561 | """Sets the downloader for this IE.""" | |
562 | self._downloader = downloader | |
563 | ||
564 | def _real_initialize(self): | |
565 | """Real initialization process. Redefine in subclasses.""" | |
566 | pass | |
567 | ||
568 | def _real_extract(self, url): | |
569 | """Real extraction process. Redefine in subclasses.""" | |
570 | pass | |
571 | ||
56c73665 JMF |
572 | @classmethod |
573 | def ie_key(cls): | |
574 | """A string for getting the InfoExtractor with get_info_extractor""" | |
dc519b54 | 575 | return compat_str(cls.__name__[:-2]) |
56c73665 | 576 | |
d6983cb4 PH |
577 | @property |
578 | def IE_NAME(self): | |
dc519b54 | 579 | return compat_str(type(self).__name__[:-2]) |
d6983cb4 | 580 | |
d391b7e2 S |
581 | @staticmethod |
582 | def __can_accept_status_code(err, expected_status): | |
583 | assert isinstance(err, compat_urllib_error.HTTPError) | |
584 | if expected_status is None: | |
585 | return False | |
586 | if isinstance(expected_status, compat_integer_types): | |
587 | return err.code == expected_status | |
588 | elif isinstance(expected_status, (list, tuple)): | |
589 | return err.code in expected_status | |
590 | elif callable(expected_status): | |
591 | return expected_status(err.code) is True | |
592 | else: | |
593 | assert False | |
594 | ||
595 | def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers={}, query={}, expected_status=None): | |
596 | """ | |
597 | Return the response handle. | |
598 | ||
599 | See _download_webpage docstring for arguments specification. | |
600 | """ | |
d6983cb4 PH |
601 | if note is None: |
602 | self.report_download_webpage(video_id) | |
603 | elif note is not False: | |
7cc3570e | 604 | if video_id is None: |
f1a9d64e | 605 | self.to_screen('%s' % (note,)) |
7cc3570e | 606 | else: |
f1a9d64e | 607 | self.to_screen('%s: %s' % (video_id, note)) |
2132edaa S |
608 | |
609 | # Some sites check X-Forwarded-For HTTP header in order to figure out | |
610 | # the origin of the client behind proxy. This allows bypassing geo | |
611 | # restriction by faking this header's value to IP that belongs to some | |
612 | # geo unrestricted country. We will do so once we encounter any | |
613 | # geo restriction error. | |
614 | if self._x_forwarded_for_ip: | |
615 | if 'X-Forwarded-For' not in headers: | |
616 | headers['X-Forwarded-For'] = self._x_forwarded_for_ip | |
617 | ||
41d06b04 S |
618 | if isinstance(url_or_request, compat_urllib_request.Request): |
619 | url_or_request = update_Request( | |
620 | url_or_request, data=data, headers=headers, query=query) | |
621 | else: | |
cdfee168 | 622 | if query: |
623 | url_or_request = update_url_query(url_or_request, query) | |
2c0d9c62 | 624 | if data is not None or headers: |
41d06b04 | 625 | url_or_request = sanitized_Request(url_or_request, data, headers) |
d6983cb4 | 626 | try: |
dca08720 | 627 | return self._downloader.urlopen(url_or_request) |
d6983cb4 | 628 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: |
d391b7e2 S |
629 | if isinstance(err, compat_urllib_error.HTTPError): |
630 | if self.__can_accept_status_code(err, expected_status): | |
95e42d73 XDG |
631 | # Retain reference to error to prevent file object from |
632 | # being closed before it can be read. Works around the | |
633 | # effects of <https://bugs.python.org/issue15002> | |
634 | # introduced in Python 3.4.1. | |
635 | err.fp._error = err | |
d391b7e2 S |
636 | return err.fp |
637 | ||
aa94a6d3 PH |
638 | if errnote is False: |
639 | return False | |
d6983cb4 | 640 | if errnote is None: |
f1a9d64e | 641 | errnote = 'Unable to download webpage' |
7f8b2714 | 642 | |
9b9c5355 | 643 | errmsg = '%s: %s' % (errnote, error_to_compat_str(err)) |
7cc3570e PH |
644 | if fatal: |
645 | raise ExtractorError(errmsg, sys.exc_info()[2], cause=err) | |
646 | else: | |
647 | self._downloader.report_warning(errmsg) | |
648 | return False | |
d6983cb4 | 649 | |
d391b7e2 S |
650 | 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): |
651 | """ | |
652 | Return a tuple (page content as string, URL handle). | |
653 | ||
654 | See _download_webpage docstring for arguments specification. | |
655 | """ | |
b9d3e163 PH |
656 | # Strip hashes from the URL (#1038) |
657 | if isinstance(url_or_request, (compat_str, str)): | |
658 | url_or_request = url_or_request.partition('#')[0] | |
659 | ||
d391b7e2 | 660 | urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query, expected_status=expected_status) |
7cc3570e PH |
661 | if urlh is False: |
662 | assert not fatal | |
663 | return False | |
c9a77969 | 664 | content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding) |
23be51d8 PH |
665 | return (content, urlh) |
666 | ||
c9a77969 YCH |
667 | @staticmethod |
668 | def _guess_encoding_from_content(content_type, webpage_bytes): | |
d6983cb4 PH |
669 | m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type) |
670 | if m: | |
671 | encoding = m.group(1) | |
672 | else: | |
0d75ae2c | 673 | m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]', |
f143d86a PH |
674 | webpage_bytes[:1024]) |
675 | if m: | |
676 | encoding = m.group(1).decode('ascii') | |
b60016e8 PH |
677 | elif webpage_bytes.startswith(b'\xff\xfe'): |
678 | encoding = 'utf-16' | |
f143d86a PH |
679 | else: |
680 | encoding = 'utf-8' | |
c9a77969 YCH |
681 | |
682 | return encoding | |
683 | ||
4457823d S |
684 | def __check_blocked(self, content): |
685 | first_block = content[:512] | |
3089bc74 S |
686 | if ('<title>Access to this site is blocked</title>' in content |
687 | and 'Websense' in first_block): | |
4457823d S |
688 | msg = 'Access to this webpage has been blocked by Websense filtering software in your network.' |
689 | blocked_iframe = self._html_search_regex( | |
690 | r'<iframe src="([^"]+)"', content, | |
691 | 'Websense information URL', default=None) | |
692 | if blocked_iframe: | |
693 | msg += ' Visit %s for more details' % blocked_iframe | |
694 | raise ExtractorError(msg, expected=True) | |
695 | if '<title>The URL you requested has been blocked</title>' in first_block: | |
696 | msg = ( | |
697 | 'Access to this webpage has been blocked by Indian censorship. ' | |
698 | 'Use a VPN or proxy server (with --proxy) to route around it.') | |
699 | block_msg = self._html_search_regex( | |
700 | r'</h1><p>(.*?)</p>', | |
701 | content, 'block message', default=None) | |
702 | if block_msg: | |
703 | msg += ' (Message: "%s")' % block_msg.replace('\n', ' ') | |
704 | raise ExtractorError(msg, expected=True) | |
3089bc74 S |
705 | if ('<title>TTK :: Доступ к ресурсу ограничен</title>' in content |
706 | and 'blocklist.rkn.gov.ru' in content): | |
4457823d S |
707 | raise ExtractorError( |
708 | 'Access to this webpage has been blocked by decision of the Russian government. ' | |
709 | 'Visit http://blocklist.rkn.gov.ru/ for a block reason.', | |
710 | expected=True) | |
711 | ||
c9a77969 YCH |
712 | def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None): |
713 | content_type = urlh.headers.get('Content-Type', '') | |
714 | webpage_bytes = urlh.read() | |
715 | if prefix is not None: | |
716 | webpage_bytes = prefix + webpage_bytes | |
717 | if not encoding: | |
718 | encoding = self._guess_encoding_from_content(content_type, webpage_bytes) | |
d6983cb4 | 719 | if self._downloader.params.get('dump_intermediate_pages', False): |
f610dbb0 | 720 | self.to_screen('Dumping request to ' + urlh.geturl()) |
d6983cb4 PH |
721 | dump = base64.b64encode(webpage_bytes).decode('ascii') |
722 | self._downloader.to_screen(dump) | |
d41e6efc | 723 | if self._downloader.params.get('write_pages', False): |
f610dbb0 | 724 | basen = '%s_%s' % (video_id, urlh.geturl()) |
c1bce22f | 725 | if len(basen) > 240: |
f1a9d64e | 726 | h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest() |
c1bce22f PH |
727 | basen = basen[:240 - len(h)] + h |
728 | raw_filename = basen + '.dump' | |
d41e6efc | 729 | filename = sanitize_filename(raw_filename, restricted=True) |
f1a9d64e | 730 | self.to_screen('Saving request to ' + filename) |
5f58165d S |
731 | # Working around MAX_PATH limitation on Windows (see |
732 | # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx) | |
e9c0cdd3 | 733 | if compat_os_name == 'nt': |
5f58165d S |
734 | absfilepath = os.path.abspath(filename) |
735 | if len(absfilepath) > 259: | |
736 | filename = '\\\\?\\' + absfilepath | |
d41e6efc PH |
737 | with open(filename, 'wb') as outf: |
738 | outf.write(webpage_bytes) | |
739 | ||
ec0fafbb AA |
740 | try: |
741 | content = webpage_bytes.decode(encoding, 'replace') | |
742 | except LookupError: | |
743 | content = webpage_bytes.decode('utf-8', 'replace') | |
2410c43d | 744 | |
4457823d | 745 | self.__check_blocked(content) |
2410c43d | 746 | |
23be51d8 | 747 | return content |
d6983cb4 | 748 | |
d391b7e2 S |
749 | def _download_webpage( |
750 | self, url_or_request, video_id, note=None, errnote=None, | |
751 | fatal=True, tries=1, timeout=5, encoding=None, data=None, | |
752 | headers={}, query={}, expected_status=None): | |
753 | """ | |
754 | Return the data of the page as a string. | |
755 | ||
756 | Arguments: | |
757 | url_or_request -- plain text URL as a string or | |
758 | a compat_urllib_request.Requestobject | |
759 | video_id -- Video/playlist/item identifier (string) | |
760 | ||
761 | Keyword arguments: | |
762 | note -- note printed before downloading (string) | |
763 | errnote -- note printed in case of an error (string) | |
764 | fatal -- flag denoting whether error should be considered fatal, | |
765 | i.e. whether it should cause ExtractionError to be raised, | |
766 | otherwise a warning will be reported and extraction continued | |
767 | tries -- number of tries | |
768 | timeout -- sleep interval between tries | |
769 | encoding -- encoding for a page content decoding, guessed automatically | |
770 | when not explicitly specified | |
771 | data -- POST data (bytes) | |
772 | headers -- HTTP headers (dict) | |
773 | query -- URL query (dict) | |
774 | expected_status -- allows to accept failed HTTP requests (non 2xx | |
775 | status code) by explicitly specifying a set of accepted status | |
776 | codes. Can be any of the following entities: | |
777 | - an integer type specifying an exact failed status code to | |
778 | accept | |
779 | - a list or a tuple of integer types specifying a list of | |
780 | failed status codes to accept | |
781 | - a callable accepting an actual failed status code and | |
782 | returning True if it should be accepted | |
783 | Note that this argument does not affect success status codes (2xx) | |
784 | which are always accepted. | |
785 | """ | |
786 | ||
995ad69c TF |
787 | success = False |
788 | try_count = 0 | |
789 | while success is False: | |
790 | try: | |
d391b7e2 S |
791 | res = self._download_webpage_handle( |
792 | url_or_request, video_id, note, errnote, fatal, | |
793 | encoding=encoding, data=data, headers=headers, query=query, | |
794 | expected_status=expected_status) | |
995ad69c TF |
795 | success = True |
796 | except compat_http_client.IncompleteRead as e: | |
797 | try_count += 1 | |
798 | if try_count >= tries: | |
799 | raise e | |
800 | self._sleep(timeout, video_id) | |
7cc3570e PH |
801 | if res is False: |
802 | return res | |
803 | else: | |
804 | content, _ = res | |
805 | return content | |
d6983cb4 | 806 | |
e0d198c1 S |
807 | def _download_xml_handle( |
808 | self, url_or_request, video_id, note='Downloading XML', | |
809 | errnote='Unable to download XML', transform_source=None, | |
d391b7e2 S |
810 | fatal=True, encoding=None, data=None, headers={}, query={}, |
811 | expected_status=None): | |
812 | """ | |
ee0ba927 | 813 | Return a tuple (xml as an compat_etree_Element, URL handle). |
d391b7e2 S |
814 | |
815 | See _download_webpage docstring for arguments specification. | |
816 | """ | |
e0d198c1 S |
817 | res = self._download_webpage_handle( |
818 | url_or_request, video_id, note, errnote, fatal=fatal, | |
d391b7e2 S |
819 | encoding=encoding, data=data, headers=headers, query=query, |
820 | expected_status=expected_status) | |
e0d198c1 S |
821 | if res is False: |
822 | return res | |
823 | xml_string, urlh = res | |
824 | return self._parse_xml( | |
825 | xml_string, video_id, transform_source=transform_source, | |
826 | fatal=fatal), urlh | |
827 | ||
d391b7e2 S |
828 | def _download_xml( |
829 | self, url_or_request, video_id, | |
830 | note='Downloading XML', errnote='Unable to download XML', | |
831 | transform_source=None, fatal=True, encoding=None, | |
832 | data=None, headers={}, query={}, expected_status=None): | |
833 | """ | |
ee0ba927 | 834 | Return the xml as an compat_etree_Element. |
d391b7e2 S |
835 | |
836 | See _download_webpage docstring for arguments specification. | |
837 | """ | |
e0d198c1 S |
838 | res = self._download_xml_handle( |
839 | url_or_request, video_id, note=note, errnote=errnote, | |
840 | transform_source=transform_source, fatal=fatal, encoding=encoding, | |
d391b7e2 S |
841 | data=data, headers=headers, query=query, |
842 | expected_status=expected_status) | |
e0d198c1 | 843 | return res if res is False else res[0] |
e01c3d2e S |
844 | |
845 | def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True): | |
e2b38da9 PH |
846 | if transform_source: |
847 | xml_string = transform_source(xml_string) | |
e01c3d2e S |
848 | try: |
849 | return compat_etree_fromstring(xml_string.encode('utf-8')) | |
850 | except compat_xml_parse_error as ve: | |
851 | errmsg = '%s: Failed to parse XML ' % video_id | |
852 | if fatal: | |
853 | raise ExtractorError(errmsg, cause=ve) | |
854 | else: | |
855 | self.report_warning(errmsg + str(ve)) | |
267ed0c5 | 856 | |
0fe7783e S |
857 | def _download_json_handle( |
858 | self, url_or_request, video_id, note='Downloading JSON metadata', | |
859 | errnote='Unable to download JSON metadata', transform_source=None, | |
d391b7e2 S |
860 | fatal=True, encoding=None, data=None, headers={}, query={}, |
861 | expected_status=None): | |
862 | """ | |
863 | Return a tuple (JSON object, URL handle). | |
864 | ||
865 | See _download_webpage docstring for arguments specification. | |
866 | """ | |
0fe7783e | 867 | res = self._download_webpage_handle( |
c9a77969 | 868 | url_or_request, video_id, note, errnote, fatal=fatal, |
d391b7e2 S |
869 | encoding=encoding, data=data, headers=headers, query=query, |
870 | expected_status=expected_status) | |
0fe7783e S |
871 | if res is False: |
872 | return res | |
873 | json_string, urlh = res | |
ebb64199 | 874 | return self._parse_json( |
0fe7783e S |
875 | json_string, video_id, transform_source=transform_source, |
876 | fatal=fatal), urlh | |
877 | ||
878 | def _download_json( | |
879 | self, url_or_request, video_id, note='Downloading JSON metadata', | |
880 | errnote='Unable to download JSON metadata', transform_source=None, | |
d391b7e2 S |
881 | fatal=True, encoding=None, data=None, headers={}, query={}, |
882 | expected_status=None): | |
883 | """ | |
884 | Return the JSON object as a dict. | |
885 | ||
886 | See _download_webpage docstring for arguments specification. | |
887 | """ | |
0fe7783e S |
888 | res = self._download_json_handle( |
889 | url_or_request, video_id, note=note, errnote=errnote, | |
890 | transform_source=transform_source, fatal=fatal, encoding=encoding, | |
d391b7e2 S |
891 | data=data, headers=headers, query=query, |
892 | expected_status=expected_status) | |
0fe7783e | 893 | return res if res is False else res[0] |
ebb64199 TF |
894 | |
895 | def _parse_json(self, json_string, video_id, transform_source=None, fatal=True): | |
81c2f20b PH |
896 | if transform_source: |
897 | json_string = transform_source(json_string) | |
3d3538e4 PH |
898 | try: |
899 | return json.loads(json_string) | |
900 | except ValueError as ve: | |
e7b6d122 PH |
901 | errmsg = '%s: Failed to parse JSON ' % video_id |
902 | if fatal: | |
903 | raise ExtractorError(errmsg, cause=ve) | |
904 | else: | |
905 | self.report_warning(errmsg + str(ve)) | |
3d3538e4 | 906 | |
f45f96f8 | 907 | def report_warning(self, msg, video_id=None): |
f1a9d64e | 908 | idstr = '' if video_id is None else '%s: ' % video_id |
f45f96f8 | 909 | self._downloader.report_warning( |
f1a9d64e | 910 | '[%s] %s%s' % (self.IE_NAME, idstr, msg)) |
f45f96f8 | 911 | |
d6983cb4 PH |
912 | def to_screen(self, msg): |
913 | """Print msg to screen, prefixing it with '[ie_name]'""" | |
f1a9d64e | 914 | self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg)) |
d6983cb4 PH |
915 | |
916 | def report_extraction(self, id_or_name): | |
917 | """Report information extraction.""" | |
f1a9d64e | 918 | self.to_screen('%s: Extracting information' % id_or_name) |
d6983cb4 PH |
919 | |
920 | def report_download_webpage(self, video_id): | |
921 | """Report webpage download.""" | |
f1a9d64e | 922 | self.to_screen('%s: Downloading webpage' % video_id) |
d6983cb4 PH |
923 | |
924 | def report_age_confirmation(self): | |
925 | """Report attempt to confirm age.""" | |
f1a9d64e | 926 | self.to_screen('Confirming age') |
d6983cb4 | 927 | |
fc79158d JMF |
928 | def report_login(self): |
929 | """Report attempt to log in.""" | |
f1a9d64e | 930 | self.to_screen('Logging in') |
fc79158d | 931 | |
43e7d3c9 S |
932 | @staticmethod |
933 | def raise_login_required(msg='This video is only available for registered users'): | |
934 | raise ExtractorError( | |
935 | '%s. Use --username and --password or --netrc to provide account credentials.' % msg, | |
936 | expected=True) | |
937 | ||
c430802e | 938 | @staticmethod |
773f291d S |
939 | def raise_geo_restricted(msg='This video is not available from your location due to geo restriction', countries=None): |
940 | raise GeoRestrictedError(msg, countries=countries) | |
c430802e | 941 | |
5f6a1245 | 942 | # Methods for following #608 |
c0d0b01f | 943 | @staticmethod |
830d53bf | 944 | def url_result(url, ie=None, video_id=None, video_title=None): |
10952eb2 | 945 | """Returns a URL that points to a page that should be processed""" |
5f6a1245 | 946 | # TODO: ie should be the class used for getting the info |
d6983cb4 PH |
947 | video_info = {'_type': 'url', |
948 | 'url': url, | |
949 | 'ie_key': ie} | |
7012b23c PH |
950 | if video_id is not None: |
951 | video_info['id'] = video_id | |
830d53bf S |
952 | if video_title is not None: |
953 | video_info['title'] = video_title | |
d6983cb4 | 954 | return video_info |
5f6a1245 | 955 | |
749ca5ec S |
956 | def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None): |
957 | urls = orderedSet( | |
46b18f23 JH |
958 | self.url_result(self._proto_relative_url(getter(m) if getter else m), ie) |
959 | for m in matches) | |
960 | return self.playlist_result( | |
749ca5ec | 961 | urls, playlist_id=playlist_id, playlist_title=playlist_title) |
46b18f23 | 962 | |
c0d0b01f | 963 | @staticmethod |
acf5cbfe | 964 | def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None): |
d6983cb4 PH |
965 | """Returns a playlist""" |
966 | video_info = {'_type': 'playlist', | |
967 | 'entries': entries} | |
968 | if playlist_id: | |
969 | video_info['id'] = playlist_id | |
970 | if playlist_title: | |
971 | video_info['title'] = playlist_title | |
acf5cbfe S |
972 | if playlist_description: |
973 | video_info['description'] = playlist_description | |
d6983cb4 PH |
974 | return video_info |
975 | ||
c342041f | 976 | def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
977 | """ |
978 | Perform a regex search on the given string, using a single or a list of | |
979 | patterns returning the first matching group. | |
980 | In case of failure return a default value or raise a WARNING or a | |
55b3e45b | 981 | RegexNotFoundError, depending on fatal, specifying the field name. |
d6983cb4 PH |
982 | """ |
983 | if isinstance(pattern, (str, compat_str, compiled_regex_type)): | |
984 | mobj = re.search(pattern, string, flags) | |
985 | else: | |
986 | for p in pattern: | |
987 | mobj = re.search(p, string, flags) | |
c3415d1b PH |
988 | if mobj: |
989 | break | |
d6983cb4 | 990 | |
e9c0cdd3 | 991 | if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty(): |
f1a9d64e | 992 | _name = '\033[0;34m%s\033[0m' % name |
d6983cb4 PH |
993 | else: |
994 | _name = name | |
995 | ||
996 | if mobj: | |
711ede6e PH |
997 | if group is None: |
998 | # return the first matching group | |
999 | return next(g for g in mobj.groups() if g is not None) | |
1000 | else: | |
1001 | return mobj.group(group) | |
c342041f | 1002 | elif default is not NO_DEFAULT: |
d6983cb4 PH |
1003 | return default |
1004 | elif fatal: | |
f1a9d64e | 1005 | raise RegexNotFoundError('Unable to extract %s' % _name) |
d6983cb4 | 1006 | else: |
08f2a92c | 1007 | self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) |
d6983cb4 PH |
1008 | return None |
1009 | ||
c342041f | 1010 | def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
1011 | """ |
1012 | Like _search_regex, but strips HTML tags and unescapes entities. | |
1013 | """ | |
711ede6e | 1014 | res = self._search_regex(pattern, string, name, default, fatal, flags, group) |
d6983cb4 PH |
1015 | if res: |
1016 | return clean_html(res).strip() | |
1017 | else: | |
1018 | return res | |
1019 | ||
2118fdd1 RA |
1020 | def _get_netrc_login_info(self, netrc_machine=None): |
1021 | username = None | |
1022 | password = None | |
1023 | netrc_machine = netrc_machine or self._NETRC_MACHINE | |
1024 | ||
1025 | if self._downloader.params.get('usenetrc', False): | |
1026 | try: | |
1027 | info = netrc.netrc().authenticators(netrc_machine) | |
1028 | if info is not None: | |
1029 | username = info[0] | |
1030 | password = info[2] | |
1031 | else: | |
dcce092e S |
1032 | raise netrc.NetrcParseError( |
1033 | 'No authenticators for %s' % netrc_machine) | |
2118fdd1 | 1034 | except (IOError, netrc.NetrcParseError) as err: |
dcce092e S |
1035 | self._downloader.report_warning( |
1036 | 'parsing .netrc: %s' % error_to_compat_str(err)) | |
2118fdd1 | 1037 | |
dcce092e | 1038 | return username, password |
2118fdd1 | 1039 | |
1b6712ab | 1040 | def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None): |
fc79158d | 1041 | """ |
cf0649f8 | 1042 | Get the login info as (username, password) |
32443dd3 S |
1043 | First look for the manually specified credentials using username_option |
1044 | and password_option as keys in params dictionary. If no such credentials | |
1045 | available look in the netrc file using the netrc_machine or _NETRC_MACHINE | |
1046 | value. | |
fc79158d JMF |
1047 | If there's no info available, return (None, None) |
1048 | """ | |
1049 | if self._downloader is None: | |
1050 | return (None, None) | |
1051 | ||
fc79158d JMF |
1052 | downloader_params = self._downloader.params |
1053 | ||
1054 | # Attempt to use provided username and password or .netrc data | |
1b6712ab RA |
1055 | if downloader_params.get(username_option) is not None: |
1056 | username = downloader_params[username_option] | |
1057 | password = downloader_params[password_option] | |
2118fdd1 | 1058 | else: |
1b6712ab | 1059 | username, password = self._get_netrc_login_info(netrc_machine) |
5f6a1245 | 1060 | |
2133565c | 1061 | return username, password |
fc79158d | 1062 | |
e64b7569 | 1063 | def _get_tfa_info(self, note='two-factor verification code'): |
83317f69 | 1064 | """ |
1065 | Get the two-factor authentication info | |
1066 | TODO - asking the user will be required for sms/phone verify | |
1067 | currently just uses the command line option | |
1068 | If there's no info available, return None | |
1069 | """ | |
1070 | if self._downloader is None: | |
83317f69 | 1071 | return None |
1072 | downloader_params = self._downloader.params | |
1073 | ||
d800609c | 1074 | if downloader_params.get('twofactor') is not None: |
83317f69 | 1075 | return downloader_params['twofactor'] |
1076 | ||
e64b7569 | 1077 | return compat_getpass('Type %s and press [Return]: ' % note) |
83317f69 | 1078 | |
46720279 JMF |
1079 | # Helper functions for extracting OpenGraph info |
1080 | @staticmethod | |
ab2d5247 | 1081 | def _og_regexes(prop): |
448ef1f3 | 1082 | content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))' |
22f5f5c6 | 1083 | property_re = (r'(?:name|property)=(?:\'og[:-]%(prop)s\'|"og[:-]%(prop)s"|\s*og[:-]%(prop)s\b)' |
7a6d76a6 | 1084 | % {'prop': re.escape(prop)}) |
78fb87b2 | 1085 | template = r'<meta[^>]+?%s[^>]+?%s' |
ab2d5247 | 1086 | return [ |
78fb87b2 JMF |
1087 | template % (property_re, content_re), |
1088 | template % (content_re, property_re), | |
ab2d5247 | 1089 | ] |
46720279 | 1090 | |
864f24bd S |
1091 | @staticmethod |
1092 | def _meta_regex(prop): | |
1093 | return r'''(?isx)<meta | |
8b9848ac | 1094 | (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1) |
864f24bd S |
1095 | [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop) |
1096 | ||
3c4e6d83 | 1097 | def _og_search_property(self, prop, html, name=None, **kargs): |
b070564e S |
1098 | if not isinstance(prop, (list, tuple)): |
1099 | prop = [prop] | |
46720279 | 1100 | if name is None: |
b070564e S |
1101 | name = 'OpenGraph %s' % prop[0] |
1102 | og_regexes = [] | |
1103 | for p in prop: | |
1104 | og_regexes.extend(self._og_regexes(p)) | |
1105 | escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs) | |
eb0a8398 PH |
1106 | if escaped is None: |
1107 | return None | |
1108 | return unescapeHTML(escaped) | |
46720279 JMF |
1109 | |
1110 | def _og_search_thumbnail(self, html, **kargs): | |
10952eb2 | 1111 | return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs) |
46720279 JMF |
1112 | |
1113 | def _og_search_description(self, html, **kargs): | |
1114 | return self._og_search_property('description', html, fatal=False, **kargs) | |
1115 | ||
1116 | def _og_search_title(self, html, **kargs): | |
1117 | return self._og_search_property('title', html, **kargs) | |
1118 | ||
8ffa13e0 | 1119 | def _og_search_video_url(self, html, name='video url', secure=True, **kargs): |
a3681973 PH |
1120 | regexes = self._og_regexes('video') + self._og_regexes('video:url') |
1121 | if secure: | |
1122 | regexes = self._og_regexes('video:secure_url') + regexes | |
8ffa13e0 | 1123 | return self._html_search_regex(regexes, html, name, **kargs) |
46720279 | 1124 | |
78338f71 JMF |
1125 | def _og_search_url(self, html, **kargs): |
1126 | return self._og_search_property('url', html, **kargs) | |
1127 | ||
40c696e5 | 1128 | def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs): |
88d9f6c0 S |
1129 | if not isinstance(name, (list, tuple)): |
1130 | name = [name] | |
59040888 | 1131 | if display_name is None: |
88d9f6c0 | 1132 | display_name = name[0] |
59040888 | 1133 | return self._html_search_regex( |
88d9f6c0 | 1134 | [self._meta_regex(n) for n in name], |
711ede6e | 1135 | html, display_name, fatal=fatal, group='content', **kwargs) |
59040888 PH |
1136 | |
1137 | def _dc_search_uploader(self, html): | |
1138 | return self._html_search_meta('dc.creator', html, 'uploader') | |
1139 | ||
8dbe9899 PH |
1140 | def _rta_search(self, html): |
1141 | # See http://www.rtalabel.org/index.php?content=howtofaq#single | |
1142 | if re.search(r'(?ix)<meta\s+name="rating"\s+' | |
1143 | r' content="RTA-5042-1996-1400-1577-RTA"', | |
1144 | html): | |
1145 | return 18 | |
1146 | return 0 | |
1147 | ||
59040888 PH |
1148 | def _media_rating_search(self, html): |
1149 | # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/ | |
1150 | rating = self._html_search_meta('rating', html) | |
1151 | ||
1152 | if not rating: | |
1153 | return None | |
1154 | ||
1155 | RATING_TABLE = { | |
1156 | 'safe for kids': 0, | |
1157 | 'general': 8, | |
1158 | '14 years': 14, | |
1159 | 'mature': 17, | |
1160 | 'restricted': 19, | |
1161 | } | |
d800609c | 1162 | return RATING_TABLE.get(rating.lower()) |
59040888 | 1163 | |
69319969 | 1164 | def _family_friendly_search(self, html): |
6ca7732d | 1165 | # See http://schema.org/VideoObject |
ac8491fc S |
1166 | family_friendly = self._html_search_meta( |
1167 | 'isFamilyFriendly', html, default=None) | |
69319969 NJ |
1168 | |
1169 | if not family_friendly: | |
1170 | return None | |
1171 | ||
1172 | RATING_TABLE = { | |
1173 | '1': 0, | |
1174 | 'true': 0, | |
1175 | '0': 18, | |
1176 | 'false': 18, | |
1177 | } | |
d800609c | 1178 | return RATING_TABLE.get(family_friendly.lower()) |
69319969 | 1179 | |
0c708f11 JMF |
1180 | def _twitter_search_player(self, html): |
1181 | return self._html_search_meta('twitter:player', html, | |
9e1a5b84 | 1182 | 'twitter card player') |
0c708f11 | 1183 | |
95b31e26 | 1184 | def _search_json_ld(self, html, video_id, expected_type=None, **kwargs): |
4ca2a3cf | 1185 | json_ld = self._search_regex( |
0685d972 | 1186 | JSON_LD_RE, html, 'JSON-LD', group='json_ld', **kwargs) |
321b5e08 | 1187 | default = kwargs.get('default', NO_DEFAULT) |
4ca2a3cf | 1188 | if not json_ld: |
321b5e08 S |
1189 | return default if default is not NO_DEFAULT else {} |
1190 | # JSON-LD may be malformed and thus `fatal` should be respected. | |
1191 | # At the same time `default` may be passed that assumes `fatal=False` | |
1192 | # for _search_regex. Let's simulate the same behavior here as well. | |
1193 | fatal = kwargs.get('fatal', True) if default == NO_DEFAULT else False | |
1194 | return self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type) | |
4ca2a3cf | 1195 | |
95b31e26 | 1196 | def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): |
4ca2a3cf S |
1197 | if isinstance(json_ld, compat_str): |
1198 | json_ld = self._parse_json(json_ld, video_id, fatal=fatal) | |
1199 | if not json_ld: | |
1200 | return {} | |
1201 | info = {} | |
46933a15 S |
1202 | if not isinstance(json_ld, (list, tuple, dict)): |
1203 | return info | |
1204 | if isinstance(json_ld, dict): | |
1205 | json_ld = [json_ld] | |
bae14048 | 1206 | |
e7e4a6e0 S |
1207 | INTERACTION_TYPE_MAP = { |
1208 | 'CommentAction': 'comment', | |
1209 | 'AgreeAction': 'like', | |
1210 | 'DisagreeAction': 'dislike', | |
1211 | 'LikeAction': 'like', | |
1212 | 'DislikeAction': 'dislike', | |
1213 | 'ListenAction': 'view', | |
1214 | 'WatchAction': 'view', | |
1215 | 'ViewAction': 'view', | |
1216 | } | |
1217 | ||
1218 | def extract_interaction_statistic(e): | |
1219 | interaction_statistic = e.get('interactionStatistic') | |
1220 | if not isinstance(interaction_statistic, list): | |
1221 | return | |
1222 | for is_e in interaction_statistic: | |
1223 | if not isinstance(is_e, dict): | |
1224 | continue | |
1225 | if is_e.get('@type') != 'InteractionCounter': | |
1226 | continue | |
1227 | interaction_type = is_e.get('interactionType') | |
1228 | if not isinstance(interaction_type, compat_str): | |
1229 | continue | |
1230 | interaction_count = int_or_none(is_e.get('userInteractionCount')) | |
1231 | if interaction_count is None: | |
1232 | continue | |
1233 | count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1]) | |
1234 | if not count_kind: | |
1235 | continue | |
1236 | count_key = '%s_count' % count_kind | |
1237 | if info.get(count_key) is not None: | |
1238 | continue | |
1239 | info[count_key] = interaction_count | |
1240 | ||
bae14048 S |
1241 | def extract_video_object(e): |
1242 | assert e['@type'] == 'VideoObject' | |
1243 | info.update({ | |
bebef109 | 1244 | 'url': url_or_none(e.get('contentUrl')), |
bae14048 S |
1245 | 'title': unescapeHTML(e.get('name')), |
1246 | 'description': unescapeHTML(e.get('description')), | |
bebef109 | 1247 | 'thumbnail': url_or_none(e.get('thumbnailUrl') or e.get('thumbnailURL')), |
bae14048 S |
1248 | 'duration': parse_duration(e.get('duration')), |
1249 | 'timestamp': unified_timestamp(e.get('uploadDate')), | |
1250 | 'filesize': float_or_none(e.get('contentSize')), | |
1251 | 'tbr': int_or_none(e.get('bitrate')), | |
1252 | 'width': int_or_none(e.get('width')), | |
1253 | 'height': int_or_none(e.get('height')), | |
33a81c2c | 1254 | 'view_count': int_or_none(e.get('interactionCount')), |
bae14048 | 1255 | }) |
e7e4a6e0 | 1256 | extract_interaction_statistic(e) |
bae14048 | 1257 | |
46933a15 | 1258 | for e in json_ld: |
66b68672 | 1259 | if isinstance(e.get('@context'), compat_str) and re.match(r'^https?://schema.org/?$', e.get('@context')): |
46933a15 S |
1260 | item_type = e.get('@type') |
1261 | if expected_type is not None and expected_type != item_type: | |
1262 | return info | |
c69701c6 | 1263 | if item_type in ('TVEpisode', 'Episode'): |
440863ad | 1264 | episode_name = unescapeHTML(e.get('name')) |
46933a15 | 1265 | info.update({ |
440863ad | 1266 | 'episode': episode_name, |
46933a15 S |
1267 | 'episode_number': int_or_none(e.get('episodeNumber')), |
1268 | 'description': unescapeHTML(e.get('description')), | |
1269 | }) | |
440863ad S |
1270 | if not info.get('title') and episode_name: |
1271 | info['title'] = episode_name | |
46933a15 | 1272 | part_of_season = e.get('partOfSeason') |
c69701c6 | 1273 | if isinstance(part_of_season, dict) and part_of_season.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'): |
458fd30f S |
1274 | info.update({ |
1275 | 'season': unescapeHTML(part_of_season.get('name')), | |
1276 | 'season_number': int_or_none(part_of_season.get('seasonNumber')), | |
1277 | }) | |
d16b3c66 | 1278 | part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries') |
c69701c6 | 1279 | if isinstance(part_of_series, dict) and part_of_series.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'): |
46933a15 | 1280 | info['series'] = unescapeHTML(part_of_series.get('name')) |
391256dc S |
1281 | elif item_type == 'Movie': |
1282 | info.update({ | |
1283 | 'title': unescapeHTML(e.get('name')), | |
1284 | 'description': unescapeHTML(e.get('description')), | |
1285 | 'duration': parse_duration(e.get('duration')), | |
1286 | 'timestamp': unified_timestamp(e.get('dateCreated')), | |
1287 | }) | |
3931b845 | 1288 | elif item_type in ('Article', 'NewsArticle'): |
46933a15 S |
1289 | info.update({ |
1290 | 'timestamp': parse_iso8601(e.get('datePublished')), | |
1291 | 'title': unescapeHTML(e.get('headline')), | |
1292 | 'description': unescapeHTML(e.get('articleBody')), | |
1293 | }) | |
1294 | elif item_type == 'VideoObject': | |
bae14048 | 1295 | extract_video_object(e) |
c69701c6 S |
1296 | continue |
1297 | video = e.get('video') | |
1298 | if isinstance(video, dict) and video.get('@type') == 'VideoObject': | |
1299 | extract_video_object(video) | |
46933a15 | 1300 | break |
4ca2a3cf S |
1301 | return dict((k, v) for k, v in info.items() if v is not None) |
1302 | ||
27713812 | 1303 | @staticmethod |
f8da79f8 | 1304 | def _hidden_inputs(html): |
586f1cc5 | 1305 | html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html) |
201ea3ee | 1306 | hidden_inputs = {} |
c8498368 S |
1307 | for input in re.findall(r'(?i)(<input[^>]+>)', html): |
1308 | attrs = extract_attributes(input) | |
1309 | if not input: | |
201ea3ee | 1310 | continue |
c8498368 | 1311 | if attrs.get('type') not in ('hidden', 'submit'): |
201ea3ee | 1312 | continue |
c8498368 S |
1313 | name = attrs.get('name') or attrs.get('id') |
1314 | value = attrs.get('value') | |
1315 | if name and value is not None: | |
1316 | hidden_inputs[name] = value | |
201ea3ee | 1317 | return hidden_inputs |
27713812 | 1318 | |
cf61d96d S |
1319 | def _form_hidden_inputs(self, form_id, html): |
1320 | form = self._search_regex( | |
73eb13df | 1321 | r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id, |
cf61d96d S |
1322 | html, '%s form' % form_id, group='form') |
1323 | return self._hidden_inputs(form) | |
1324 | ||
3ded7bac | 1325 | def _sort_formats(self, formats, field_preference=None): |
7e8caf30 | 1326 | if not formats: |
f1a9d64e | 1327 | raise ExtractorError('No video formats found') |
7e8caf30 | 1328 | |
b0d21ded S |
1329 | for f in formats: |
1330 | # Automatically determine tbr when missing based on abr and vbr (improves | |
1331 | # formats sorting in some cases) | |
350cf045 | 1332 | if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None: |
b0d21ded S |
1333 | f['tbr'] = f['abr'] + f['vbr'] |
1334 | ||
4bcc7bd1 | 1335 | def _formats_key(f): |
e6812ac9 PH |
1336 | # TODO remove the following workaround |
1337 | from ..utils import determine_ext | |
1338 | if not f.get('ext') and 'url' in f: | |
1339 | f['ext'] = determine_ext(f['url']) | |
1340 | ||
3ded7bac | 1341 | if isinstance(field_preference, (list, tuple)): |
bf8dd790 S |
1342 | return tuple( |
1343 | f.get(field) | |
1344 | if f.get(field) is not None | |
1345 | else ('' if field == 'format_id' else -1) | |
1346 | for field in field_preference) | |
3ded7bac | 1347 | |
4bcc7bd1 PH |
1348 | preference = f.get('preference') |
1349 | if preference is None: | |
d497a201 | 1350 | preference = 0 |
4bcc7bd1 PH |
1351 | if f.get('ext') in ['f4f', 'f4m']: # Not yet supported |
1352 | preference -= 0.5 | |
1353 | ||
8b408545 RA |
1354 | protocol = f.get('protocol') or determine_protocol(f) |
1355 | proto_preference = 0 if protocol in ['http', 'https'] else (-0.5 if protocol == 'rtsp' else -0.1) | |
d497a201 | 1356 | |
4bcc7bd1 | 1357 | if f.get('vcodec') == 'none': # audio only |
dd867805 | 1358 | preference -= 50 |
4bcc7bd1 | 1359 | if self._downloader.params.get('prefer_free_formats'): |
f1a9d64e | 1360 | ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus'] |
4bcc7bd1 | 1361 | else: |
f1a9d64e | 1362 | ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a'] |
4bcc7bd1 PH |
1363 | ext_preference = 0 |
1364 | try: | |
1365 | audio_ext_preference = ORDER.index(f['ext']) | |
1366 | except ValueError: | |
1367 | audio_ext_preference = -1 | |
1368 | else: | |
dd867805 | 1369 | if f.get('acodec') == 'none': # video only |
1370 | preference -= 40 | |
4bcc7bd1 | 1371 | if self._downloader.params.get('prefer_free_formats'): |
f1a9d64e | 1372 | ORDER = ['flv', 'mp4', 'webm'] |
4bcc7bd1 | 1373 | else: |
f1a9d64e | 1374 | ORDER = ['webm', 'flv', 'mp4'] |
4bcc7bd1 PH |
1375 | try: |
1376 | ext_preference = ORDER.index(f['ext']) | |
1377 | except ValueError: | |
1378 | ext_preference = -1 | |
1379 | audio_ext_preference = 0 | |
1380 | ||
1381 | return ( | |
1382 | preference, | |
aff2f4f4 | 1383 | f.get('language_preference') if f.get('language_preference') is not None else -1, |
5d73273f | 1384 | f.get('quality') if f.get('quality') is not None else -1, |
9933b574 | 1385 | f.get('tbr') if f.get('tbr') is not None else -1, |
03cd72b0 | 1386 | f.get('filesize') if f.get('filesize') is not None else -1, |
4bcc7bd1 | 1387 | f.get('vbr') if f.get('vbr') is not None else -1, |
1a6373ef PH |
1388 | f.get('height') if f.get('height') is not None else -1, |
1389 | f.get('width') if f.get('width') is not None else -1, | |
d497a201 | 1390 | proto_preference, |
1e1896f2 | 1391 | ext_preference, |
4bcc7bd1 PH |
1392 | f.get('abr') if f.get('abr') is not None else -1, |
1393 | audio_ext_preference, | |
2c8e03d9 | 1394 | f.get('fps') if f.get('fps') is not None else -1, |
9732d77e | 1395 | f.get('filesize_approx') if f.get('filesize_approx') is not None else -1, |
c64ed2a3 | 1396 | f.get('source_preference') if f.get('source_preference') is not None else -1, |
74f72824 | 1397 | f.get('format_id') if f.get('format_id') is not None else '', |
4bcc7bd1 PH |
1398 | ) |
1399 | formats.sort(key=_formats_key) | |
59040888 | 1400 | |
96a53167 S |
1401 | def _check_formats(self, formats, video_id): |
1402 | if formats: | |
1403 | formats[:] = filter( | |
1404 | lambda f: self._is_valid_url( | |
1405 | f['url'], video_id, | |
1406 | item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'), | |
1407 | formats) | |
1408 | ||
f5bdb444 S |
1409 | @staticmethod |
1410 | def _remove_duplicate_formats(formats): | |
1411 | format_urls = set() | |
1412 | unique_formats = [] | |
1413 | for f in formats: | |
1414 | if f['url'] not in format_urls: | |
1415 | format_urls.add(f['url']) | |
1416 | unique_formats.append(f) | |
1417 | formats[:] = unique_formats | |
1418 | ||
45024183 | 1419 | def _is_valid_url(self, url, video_id, item='video', headers={}): |
2f0f6578 S |
1420 | url = self._proto_relative_url(url, scheme='http:') |
1421 | # For now assume non HTTP(S) URLs always valid | |
1422 | if not (url.startswith('http://') or url.startswith('https://')): | |
1423 | return True | |
96a53167 | 1424 | try: |
45024183 | 1425 | self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers) |
96a53167 | 1426 | return True |
25e911a9 S |
1427 | except ExtractorError: |
1428 | self.to_screen( | |
1429 | '%s: %s URL is invalid, skipping' % (video_id, item)) | |
1430 | return False | |
96a53167 | 1431 | |
20991253 | 1432 | def http_scheme(self): |
1ede5b24 | 1433 | """ Either "http:" or "https:", depending on the user's preferences """ |
20991253 PH |
1434 | return ( |
1435 | 'http:' | |
1436 | if self._downloader.params.get('prefer_insecure', False) | |
1437 | else 'https:') | |
1438 | ||
57c7411f PH |
1439 | def _proto_relative_url(self, url, scheme=None): |
1440 | if url is None: | |
1441 | return url | |
1442 | if url.startswith('//'): | |
1443 | if scheme is None: | |
1444 | scheme = self.http_scheme() | |
1445 | return scheme + url | |
1446 | else: | |
1447 | return url | |
1448 | ||
4094b6e3 PH |
1449 | def _sleep(self, timeout, video_id, msg_template=None): |
1450 | if msg_template is None: | |
f1a9d64e | 1451 | msg_template = '%(video_id)s: Waiting for %(timeout)s seconds' |
4094b6e3 PH |
1452 | msg = msg_template % {'video_id': video_id, 'timeout': timeout} |
1453 | self.to_screen(msg) | |
1454 | time.sleep(timeout) | |
1455 | ||
a38436e8 | 1456 | def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None, |
4de61310 | 1457 | transform_source=lambda s: fix_xml_ampersands(s).strip(), |
7360c06f | 1458 | fatal=True, m3u8_id=None, data=None, headers={}, query={}): |
f036a632 JMF |
1459 | manifest = self._download_xml( |
1460 | manifest_url, video_id, 'Downloading f4m manifest', | |
97f4aecf S |
1461 | 'Unable to download f4m manifest', |
1462 | # Some manifests may be malformed, e.g. prosiebensat1 generated manifests | |
067aa17e | 1463 | # (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244) |
4de61310 | 1464 | transform_source=transform_source, |
7360c06f | 1465 | fatal=fatal, data=data, headers=headers, query=query) |
4de61310 S |
1466 | |
1467 | if manifest is False: | |
8d29e47f | 1468 | return [] |
31bb8d3f | 1469 | |
0fdbb332 S |
1470 | return self._parse_f4m_formats( |
1471 | manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id, | |
448bb5f3 | 1472 | transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id) |
0fdbb332 S |
1473 | |
1474 | def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None, | |
1475 | transform_source=lambda s: fix_xml_ampersands(s).strip(), | |
448bb5f3 | 1476 | fatal=True, m3u8_id=None): |
ee0ba927 | 1477 | if not isinstance(manifest, compat_etree_Element) and not fatal: |
d9eb580a S |
1478 | return [] |
1479 | ||
fb72ec58 | 1480 | # currently youtube-dl cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy |
1481 | akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0') | |
1482 | if akamai_pv is not None and ';' in akamai_pv.text: | |
1483 | playerVerificationChallenge = akamai_pv.text.split(';')[0] | |
1484 | if playerVerificationChallenge.strip() != '': | |
1485 | return [] | |
1486 | ||
31bb8d3f | 1487 | formats = [] |
7a47d07c | 1488 | manifest_version = '1.0' |
b2527359 | 1489 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media') |
34e48bed | 1490 | if not media_nodes: |
7a47d07c | 1491 | manifest_version = '2.0' |
34e48bed | 1492 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') |
b22ca762 | 1493 | # Remove unsupported DRM protected media from final formats |
067aa17e | 1494 | # rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573). |
b22ca762 S |
1495 | media_nodes = remove_encrypted_media(media_nodes) |
1496 | if not media_nodes: | |
1497 | return formats | |
48107c19 S |
1498 | |
1499 | manifest_base_url = get_base_url(manifest) | |
0a5685b2 | 1500 | |
a6571f10 | 1501 | bootstrap_info = xpath_element( |
0a5685b2 YCH |
1502 | manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'], |
1503 | 'bootstrap info', default=None) | |
1504 | ||
edd6074c RA |
1505 | vcodec = None |
1506 | mime_type = xpath_text( | |
1507 | manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'], | |
1508 | 'base URL', default=None) | |
1509 | if mime_type and mime_type.startswith('audio/'): | |
1510 | vcodec = 'none' | |
1511 | ||
b2527359 | 1512 | for i, media_el in enumerate(media_nodes): |
77b8b4e6 S |
1513 | tbr = int_or_none(media_el.attrib.get('bitrate')) |
1514 | width = int_or_none(media_el.attrib.get('width')) | |
1515 | height = int_or_none(media_el.attrib.get('height')) | |
1516 | format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])) | |
448bb5f3 YCH |
1517 | # If <bootstrapInfo> is present, the specified f4m is a |
1518 | # stream-level manifest, and only set-level manifests may refer to | |
1519 | # external resources. See section 11.4 and section 4 of F4M spec | |
1520 | if bootstrap_info is None: | |
1521 | media_url = None | |
1522 | # @href is introduced in 2.0, see section 11.6 of F4M spec | |
1523 | if manifest_version == '2.0': | |
1524 | media_url = media_el.attrib.get('href') | |
1525 | if media_url is None: | |
1526 | media_url = media_el.attrib.get('url') | |
31c746e5 S |
1527 | if not media_url: |
1528 | continue | |
cc357c4d S |
1529 | manifest_url = ( |
1530 | media_url if media_url.startswith('http://') or media_url.startswith('https://') | |
48107c19 | 1531 | else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url)) |
70f0f5a8 S |
1532 | # If media_url is itself a f4m manifest do the recursive extraction |
1533 | # since bitrates in parent manifest (this one) and media_url manifest | |
1534 | # may differ leading to inability to resolve the format by requested | |
1535 | # bitrate in f4m downloader | |
240b6045 YCH |
1536 | ext = determine_ext(manifest_url) |
1537 | if ext == 'f4m': | |
77b8b4e6 | 1538 | f4m_formats = self._extract_f4m_formats( |
0fdbb332 | 1539 | manifest_url, video_id, preference=preference, f4m_id=f4m_id, |
77b8b4e6 S |
1540 | transform_source=transform_source, fatal=fatal) |
1541 | # Sometimes stream-level manifest contains single media entry that | |
1542 | # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player). | |
1543 | # At the same time parent's media entry in set-level manifest may | |
1544 | # contain it. We will copy it from parent in such cases. | |
1545 | if len(f4m_formats) == 1: | |
1546 | f = f4m_formats[0] | |
1547 | f.update({ | |
1548 | 'tbr': f.get('tbr') or tbr, | |
1549 | 'width': f.get('width') or width, | |
1550 | 'height': f.get('height') or height, | |
1551 | 'format_id': f.get('format_id') if not tbr else format_id, | |
edd6074c | 1552 | 'vcodec': vcodec, |
77b8b4e6 S |
1553 | }) |
1554 | formats.extend(f4m_formats) | |
70f0f5a8 | 1555 | continue |
240b6045 YCH |
1556 | elif ext == 'm3u8': |
1557 | formats.extend(self._extract_m3u8_formats( | |
1558 | manifest_url, video_id, 'mp4', preference=preference, | |
fac2af3c | 1559 | m3u8_id=m3u8_id, fatal=fatal)) |
240b6045 | 1560 | continue |
31bb8d3f | 1561 | formats.append({ |
77b8b4e6 | 1562 | 'format_id': format_id, |
31bb8d3f | 1563 | 'url': manifest_url, |
30d0b549 | 1564 | 'manifest_url': manifest_url, |
a6571f10 | 1565 | 'ext': 'flv' if bootstrap_info is not None else None, |
187ee66c | 1566 | 'protocol': 'f4m', |
b2527359 | 1567 | 'tbr': tbr, |
77b8b4e6 S |
1568 | 'width': width, |
1569 | 'height': height, | |
edd6074c | 1570 | 'vcodec': vcodec, |
60ca389c | 1571 | 'preference': preference, |
31bb8d3f | 1572 | }) |
31bb8d3f JMF |
1573 | return formats |
1574 | ||
16da9bbc YCH |
1575 | def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, m3u8_id=None): |
1576 | return { | |
f207019c | 1577 | 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])), |
704df56d PH |
1578 | 'url': m3u8_url, |
1579 | 'ext': ext, | |
1580 | 'protocol': 'm3u8', | |
37768f92 | 1581 | 'preference': preference - 100 if preference else -100, |
704df56d PH |
1582 | 'resolution': 'multiple', |
1583 | 'format_note': 'Quality selection URL', | |
16da9bbc YCH |
1584 | } |
1585 | ||
1586 | def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None, | |
1587 | entry_protocol='m3u8', preference=None, | |
1588 | m3u8_id=None, note=None, errnote=None, | |
7360c06f S |
1589 | fatal=True, live=False, data=None, headers={}, |
1590 | query={}): | |
dbd82a1d | 1591 | res = self._download_webpage_handle( |
81515ad9 | 1592 | m3u8_url, video_id, |
621ed9f5 | 1593 | note=note or 'Downloading m3u8 information', |
13af92fd | 1594 | errnote=errnote or 'Failed to download m3u8 information', |
7360c06f | 1595 | fatal=fatal, data=data, headers=headers, query=query) |
cb252080 | 1596 | |
dbd82a1d | 1597 | if res is False: |
8d29e47f | 1598 | return [] |
cb252080 | 1599 | |
dbd82a1d | 1600 | m3u8_doc, urlh = res |
37113045 | 1601 | m3u8_url = urlh.geturl() |
9cdffeeb | 1602 | |
cb252080 S |
1603 | return self._parse_m3u8_formats( |
1604 | m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol, | |
1605 | preference=preference, m3u8_id=m3u8_id, live=live) | |
1606 | ||
1607 | def _parse_m3u8_formats(self, m3u8_doc, m3u8_url, ext=None, | |
1608 | entry_protocol='m3u8', preference=None, | |
1609 | m3u8_id=None, live=False): | |
08a00eef RA |
1610 | if '#EXT-X-FAXS-CM:' in m3u8_doc: # Adobe Flash Access |
1611 | return [] | |
1612 | ||
ea229584 RA |
1613 | if re.search(r'#EXT-X-SESSION-KEY:.*?URI="skd://', m3u8_doc): # Apple FairPlay |
1614 | return [] | |
1615 | ||
ff99fe52 | 1616 | formats = [] |
0def7587 RA |
1617 | |
1618 | format_url = lambda u: ( | |
1619 | u | |
1620 | if re.match(r'^https?://', u) | |
1621 | else compat_urlparse.urljoin(m3u8_url, u)) | |
1622 | ||
cb252080 S |
1623 | # References: |
1624 | # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21 | |
067aa17e S |
1625 | # 2. https://github.com/ytdl-org/youtube-dl/issues/12211 |
1626 | # 3. https://github.com/ytdl-org/youtube-dl/issues/18923 | |
cb252080 S |
1627 | |
1628 | # We should try extracting formats only from master playlists [1, 4.3.4], | |
1629 | # i.e. playlists that describe available qualities. On the other hand | |
1630 | # media playlists [1, 4.3.3] should be returned as is since they contain | |
1631 | # just the media without qualities renditions. | |
9cdffeeb | 1632 | # Fortunately, master playlist can be easily distinguished from media |
cb252080 S |
1633 | # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4] |
1634 | # master playlist tags MUST NOT appear in a media playist and vice versa. | |
1635 | # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every | |
1636 | # media playlist and MUST NOT appear in master playlist thus we can | |
1637 | # clearly detect media playlist with this criterion. | |
1638 | ||
9cdffeeb | 1639 | if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is |
7f32e5dc | 1640 | return [{ |
1641 | 'url': m3u8_url, | |
1642 | 'format_id': m3u8_id, | |
1643 | 'ext': ext, | |
1644 | 'protocol': entry_protocol, | |
1645 | 'preference': preference, | |
1646 | }] | |
cb252080 S |
1647 | |
1648 | groups = {} | |
1649 | last_stream_inf = {} | |
1650 | ||
1651 | def extract_media(x_media_line): | |
1652 | media = parse_m3u8_attributes(x_media_line) | |
1653 | # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED | |
1654 | media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME') | |
1655 | if not (media_type and group_id and name): | |
1656 | return | |
1657 | groups.setdefault(group_id, []).append(media) | |
1658 | if media_type not in ('VIDEO', 'AUDIO'): | |
1659 | return | |
1660 | media_url = media.get('URI') | |
1661 | if media_url: | |
1662 | format_id = [] | |
9211e331 | 1663 | for v in (m3u8_id, group_id, name): |
cb252080 S |
1664 | if v: |
1665 | format_id.append(v) | |
1666 | f = { | |
1667 | 'format_id': '-'.join(format_id), | |
1668 | 'url': format_url(media_url), | |
c89b49f7 | 1669 | 'manifest_url': m3u8_url, |
cb252080 S |
1670 | 'language': media.get('LANGUAGE'), |
1671 | 'ext': ext, | |
1672 | 'protocol': entry_protocol, | |
1673 | 'preference': preference, | |
1674 | } | |
1675 | if media_type == 'AUDIO': | |
1676 | f['vcodec'] = 'none' | |
1677 | formats.append(f) | |
1678 | ||
1679 | def build_stream_name(): | |
1680 | # Despite specification does not mention NAME attribute for | |
3019cb0c S |
1681 | # EXT-X-STREAM-INF tag it still sometimes may be present (see [1] |
1682 | # or vidio test in TestInfoExtractor.test_parse_m3u8_formats) | |
ddd258f9 | 1683 | # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015 |
cb252080 S |
1684 | stream_name = last_stream_inf.get('NAME') |
1685 | if stream_name: | |
1686 | return stream_name | |
1687 | # If there is no NAME in EXT-X-STREAM-INF it will be obtained | |
1688 | # from corresponding rendition group | |
1689 | stream_group_id = last_stream_inf.get('VIDEO') | |
1690 | if not stream_group_id: | |
1691 | return | |
1692 | stream_group = groups.get(stream_group_id) | |
1693 | if not stream_group: | |
1694 | return stream_group_id | |
1695 | rendition = stream_group[0] | |
1696 | return rendition.get('NAME') or stream_group_id | |
1697 | ||
379306ef | 1698 | # parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the |
2bfc1d9d RA |
1699 | # chance to detect video only formats when EXT-X-STREAM-INF tags |
1700 | # precede EXT-X-MEDIA tags in HLS manifest such as [3]. | |
1701 | for line in m3u8_doc.splitlines(): | |
1702 | if line.startswith('#EXT-X-MEDIA:'): | |
1703 | extract_media(line) | |
1704 | ||
704df56d PH |
1705 | for line in m3u8_doc.splitlines(): |
1706 | if line.startswith('#EXT-X-STREAM-INF:'): | |
cb252080 | 1707 | last_stream_inf = parse_m3u8_attributes(line) |
704df56d PH |
1708 | elif line.startswith('#') or not line.strip(): |
1709 | continue | |
1710 | else: | |
9c99bef7 | 1711 | tbr = float_or_none( |
3089bc74 S |
1712 | last_stream_inf.get('AVERAGE-BANDWIDTH') |
1713 | or last_stream_inf.get('BANDWIDTH'), scale=1000) | |
8dc9d361 S |
1714 | format_id = [] |
1715 | if m3u8_id: | |
1716 | format_id.append(m3u8_id) | |
cb252080 | 1717 | stream_name = build_stream_name() |
b24d6336 KH |
1718 | # Bandwidth of live streams may differ over time thus making |
1719 | # format_id unpredictable. So it's better to keep provided | |
1720 | # format_id intact. | |
e9c6cdf4 | 1721 | if not live: |
ed56f260 | 1722 | format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats))) |
30d0b549 | 1723 | manifest_url = format_url(line.strip()) |
704df56d | 1724 | f = { |
8dc9d361 | 1725 | 'format_id': '-'.join(format_id), |
30d0b549 | 1726 | 'url': manifest_url, |
ff99fe52 | 1727 | 'manifest_url': m3u8_url, |
704df56d PH |
1728 | 'tbr': tbr, |
1729 | 'ext': ext, | |
cb252080 | 1730 | 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')), |
f0b5d6af PH |
1731 | 'protocol': entry_protocol, |
1732 | 'preference': preference, | |
704df56d | 1733 | } |
cb252080 | 1734 | resolution = last_stream_inf.get('RESOLUTION') |
704df56d | 1735 | if resolution: |
c4c9b844 S |
1736 | mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution) |
1737 | if mobj: | |
1738 | f['width'] = int(mobj.group('width')) | |
1739 | f['height'] = int(mobj.group('height')) | |
00f4764c RA |
1740 | # Unified Streaming Platform |
1741 | mobj = re.search( | |
1742 | r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url']) | |
1743 | if mobj: | |
1744 | abr, vbr = mobj.groups() | |
1745 | abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000) | |
fbb6edd2 | 1746 | f.update({ |
00f4764c RA |
1747 | 'vbr': vbr, |
1748 | 'abr': abr, | |
fbb6edd2 | 1749 | }) |
cb252080 S |
1750 | codecs = parse_codecs(last_stream_inf.get('CODECS')) |
1751 | f.update(codecs) | |
1752 | audio_group_id = last_stream_inf.get('AUDIO') | |
1753 | # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which | |
1754 | # references a rendition group MUST have a CODECS attribute. | |
1755 | # However, this is not always respected, for example, [2] | |
1756 | # contains EXT-X-STREAM-INF tag which references AUDIO | |
1757 | # rendition group but does not have CODECS and despite | |
0e7b8d3e S |
1758 | # referencing an audio group it represents a complete |
1759 | # (with audio and video) format. So, for such cases we will | |
1760 | # ignore references to rendition groups and treat them | |
cb252080 S |
1761 | # as complete formats. |
1762 | if audio_group_id and codecs and f.get('vcodec') != 'none': | |
1763 | audio_group = groups.get(audio_group_id) | |
1764 | if audio_group and audio_group[0].get('URI'): | |
1765 | # TODO: update acodec for audio only formats with | |
1766 | # the same GROUP-ID | |
1767 | f['acodec'] = 'none' | |
704df56d | 1768 | formats.append(f) |
5ef62fc4 RA |
1769 | |
1770 | # for DailyMotion | |
1771 | progressive_uri = last_stream_inf.get('PROGRESSIVE-URI') | |
1772 | if progressive_uri: | |
1773 | http_f = f.copy() | |
1774 | del http_f['manifest_url'] | |
1775 | http_f.update({ | |
1776 | 'format_id': f['format_id'].replace('hls-', 'http-'), | |
1777 | 'protocol': 'http', | |
1778 | 'url': progressive_uri, | |
1779 | }) | |
1780 | formats.append(http_f) | |
1781 | ||
cb252080 | 1782 | last_stream_inf = {} |
704df56d PH |
1783 | return formats |
1784 | ||
a107193e S |
1785 | @staticmethod |
1786 | def _xpath_ns(path, namespace=None): | |
1787 | if not namespace: | |
1788 | return path | |
1789 | out = [] | |
1790 | for c in path.split('/'): | |
1791 | if not c or c == '.': | |
1792 | out.append(c) | |
1793 | else: | |
1794 | out.append('{%s}%s' % (namespace, c)) | |
1795 | return '/'.join(out) | |
1796 | ||
09f572fb | 1797 | def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None): |
1798 | smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source) | |
a107193e | 1799 | |
995029a1 PH |
1800 | if smil is False: |
1801 | assert not fatal | |
1802 | return [] | |
e89a2aab | 1803 | |
17712eeb | 1804 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
1805 | |
1806 | return self._parse_smil_formats( | |
1807 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
1808 | ||
1809 | def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None): | |
1810 | smil = self._download_smil(smil_url, video_id, fatal=fatal) | |
1811 | if smil is False: | |
1812 | return {} | |
1813 | return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params) | |
1814 | ||
09f572fb | 1815 | def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None): |
a107193e S |
1816 | return self._download_xml( |
1817 | smil_url, video_id, 'Downloading SMIL file', | |
09f572fb | 1818 | 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source) |
a107193e S |
1819 | |
1820 | def _parse_smil(self, smil, smil_url, video_id, f4m_params=None): | |
17712eeb | 1821 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
1822 | |
1823 | formats = self._parse_smil_formats( | |
1824 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
1825 | subtitles = self._parse_smil_subtitles(smil, namespace=namespace) | |
1826 | ||
1827 | video_id = os.path.splitext(url_basename(smil_url))[0] | |
1828 | title = None | |
1829 | description = None | |
647eab45 | 1830 | upload_date = None |
a107193e S |
1831 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): |
1832 | name = meta.attrib.get('name') | |
1833 | content = meta.attrib.get('content') | |
1834 | if not name or not content: | |
1835 | continue | |
1836 | if not title and name == 'title': | |
1837 | title = content | |
1838 | elif not description and name in ('description', 'abstract'): | |
1839 | description = content | |
647eab45 S |
1840 | elif not upload_date and name == 'date': |
1841 | upload_date = unified_strdate(content) | |
a107193e | 1842 | |
1e5bcdec S |
1843 | thumbnails = [{ |
1844 | 'id': image.get('type'), | |
1845 | 'url': image.get('src'), | |
1846 | 'width': int_or_none(image.get('width')), | |
1847 | 'height': int_or_none(image.get('height')), | |
1848 | } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')] | |
1849 | ||
a107193e S |
1850 | return { |
1851 | 'id': video_id, | |
1852 | 'title': title or video_id, | |
1853 | 'description': description, | |
647eab45 | 1854 | 'upload_date': upload_date, |
1e5bcdec | 1855 | 'thumbnails': thumbnails, |
a107193e S |
1856 | 'formats': formats, |
1857 | 'subtitles': subtitles, | |
1858 | } | |
1859 | ||
17712eeb S |
1860 | def _parse_smil_namespace(self, smil): |
1861 | return self._search_regex( | |
1862 | r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None) | |
1863 | ||
f877c6ae | 1864 | def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None): |
a107193e S |
1865 | base = smil_url |
1866 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): | |
1867 | b = meta.get('base') or meta.get('httpBase') | |
1868 | if b: | |
1869 | base = b | |
1870 | break | |
e89a2aab S |
1871 | |
1872 | formats = [] | |
1873 | rtmp_count = 0 | |
a107193e | 1874 | http_count = 0 |
7f32e5dc | 1875 | m3u8_count = 0 |
a107193e | 1876 | |
81e1c4e2 | 1877 | srcs = [] |
ad96b4c8 YCH |
1878 | media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace)) |
1879 | for medium in media: | |
1880 | src = medium.get('src') | |
81e1c4e2 | 1881 | if not src or src in srcs: |
a107193e | 1882 | continue |
81e1c4e2 | 1883 | srcs.append(src) |
a107193e | 1884 | |
ad96b4c8 YCH |
1885 | bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000) |
1886 | filesize = int_or_none(medium.get('size') or medium.get('fileSize')) | |
1887 | width = int_or_none(medium.get('width')) | |
1888 | height = int_or_none(medium.get('height')) | |
1889 | proto = medium.get('proto') | |
1890 | ext = medium.get('ext') | |
a107193e | 1891 | src_ext = determine_ext(src) |
ad96b4c8 | 1892 | streamer = medium.get('streamer') or base |
a107193e S |
1893 | |
1894 | if proto == 'rtmp' or streamer.startswith('rtmp'): | |
1895 | rtmp_count += 1 | |
1896 | formats.append({ | |
1897 | 'url': streamer, | |
1898 | 'play_path': src, | |
1899 | 'ext': 'flv', | |
1900 | 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), | |
1901 | 'tbr': bitrate, | |
1902 | 'filesize': filesize, | |
1903 | 'width': width, | |
1904 | 'height': height, | |
1905 | }) | |
f877c6ae YCH |
1906 | if transform_rtmp_url: |
1907 | streamer, src = transform_rtmp_url(streamer, src) | |
1908 | formats[-1].update({ | |
1909 | 'url': streamer, | |
1910 | 'play_path': src, | |
1911 | }) | |
a107193e S |
1912 | continue |
1913 | ||
1914 | src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src) | |
c349456e | 1915 | src_url = src_url.strip() |
a107193e S |
1916 | |
1917 | if proto == 'm3u8' or src_ext == 'm3u8': | |
7f32e5dc | 1918 | m3u8_formats = self._extract_m3u8_formats( |
1919 | src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False) | |
1920 | if len(m3u8_formats) == 1: | |
1921 | m3u8_count += 1 | |
1922 | m3u8_formats[0].update({ | |
1923 | 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate), | |
1924 | 'tbr': bitrate, | |
1925 | 'width': width, | |
1926 | 'height': height, | |
1927 | }) | |
1928 | formats.extend(m3u8_formats) | |
bd21ead2 | 1929 | elif src_ext == 'f4m': |
a107193e S |
1930 | f4m_url = src_url |
1931 | if not f4m_params: | |
1932 | f4m_params = { | |
1933 | 'hdcore': '3.2.0', | |
1934 | 'plugin': 'flowplayer-3.2.0.1', | |
1935 | } | |
1936 | f4m_url += '&' if '?' in f4m_url else '?' | |
15707c7e | 1937 | f4m_url += compat_urllib_parse_urlencode(f4m_params) |
7e5edcfd | 1938 | formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False)) |
bd21ead2 RA |
1939 | elif src_ext == 'mpd': |
1940 | formats.extend(self._extract_mpd_formats( | |
1941 | src_url, video_id, mpd_id='dash', fatal=False)) | |
1942 | elif re.search(r'\.ism/[Mm]anifest', src_url): | |
1943 | formats.extend(self._extract_ism_formats( | |
1944 | src_url, video_id, ism_id='mss', fatal=False)) | |
1945 | elif src_url.startswith('http') and self._is_valid_url(src, video_id): | |
a107193e S |
1946 | http_count += 1 |
1947 | formats.append({ | |
1948 | 'url': src_url, | |
1949 | 'ext': ext or src_ext or 'flv', | |
1950 | 'format_id': 'http-%d' % (bitrate or http_count), | |
1951 | 'tbr': bitrate, | |
1952 | 'filesize': filesize, | |
1953 | 'width': width, | |
1954 | 'height': height, | |
1955 | }) | |
63757032 | 1956 | |
e89a2aab S |
1957 | return formats |
1958 | ||
ce00af87 | 1959 | def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'): |
d413095f | 1960 | urls = [] |
a107193e S |
1961 | subtitles = {} |
1962 | for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))): | |
1963 | src = textstream.get('src') | |
d413095f | 1964 | if not src or src in urls: |
a107193e | 1965 | continue |
d413095f | 1966 | urls.append(src) |
df634be2 | 1967 | ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src) |
03bc7237 | 1968 | lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang |
a107193e S |
1969 | subtitles.setdefault(lang, []).append({ |
1970 | 'url': src, | |
1971 | 'ext': ext, | |
1972 | }) | |
1973 | return subtitles | |
63757032 | 1974 | |
47a5cb77 | 1975 | def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True): |
942acef5 | 1976 | xspf = self._download_xml( |
47a5cb77 | 1977 | xspf_url, playlist_id, 'Downloading xpsf playlist', |
942acef5 S |
1978 | 'Unable to download xspf manifest', fatal=fatal) |
1979 | if xspf is False: | |
1980 | return [] | |
47a5cb77 S |
1981 | return self._parse_xspf( |
1982 | xspf, playlist_id, xspf_url=xspf_url, | |
1983 | xspf_base_url=base_url(xspf_url)) | |
8d6765cf | 1984 | |
47a5cb77 | 1985 | def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None): |
8d6765cf S |
1986 | NS_MAP = { |
1987 | 'xspf': 'http://xspf.org/ns/0/', | |
1988 | 's1': 'http://static.streamone.nl/player/ns/0', | |
1989 | } | |
1990 | ||
1991 | entries = [] | |
47a5cb77 | 1992 | for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)): |
8d6765cf | 1993 | title = xpath_text( |
98044462 | 1994 | track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id) |
8d6765cf S |
1995 | description = xpath_text( |
1996 | track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description') | |
1997 | thumbnail = xpath_text( | |
1998 | track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail') | |
1999 | duration = float_or_none( | |
2000 | xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000) | |
2001 | ||
47a5cb77 S |
2002 | formats = [] |
2003 | for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)): | |
2004 | format_url = urljoin(xspf_base_url, location.text) | |
2005 | if not format_url: | |
2006 | continue | |
2007 | formats.append({ | |
2008 | 'url': format_url, | |
2009 | 'manifest_url': xspf_url, | |
2010 | 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)), | |
2011 | 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))), | |
2012 | 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))), | |
2013 | }) | |
8d6765cf S |
2014 | self._sort_formats(formats) |
2015 | ||
2016 | entries.append({ | |
2017 | 'id': playlist_id, | |
2018 | 'title': title, | |
2019 | 'description': description, | |
2020 | 'thumbnail': thumbnail, | |
2021 | 'duration': duration, | |
2022 | 'formats': formats, | |
2023 | }) | |
2024 | return entries | |
2025 | ||
7360c06f | 2026 | 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 | 2027 | res = self._download_xml_handle( |
1bac3455 | 2028 | mpd_url, video_id, |
2029 | note=note or 'Downloading MPD manifest', | |
2030 | errnote=errnote or 'Failed to download MPD manifest', | |
7360c06f | 2031 | fatal=fatal, data=data, headers=headers, query=query) |
1bac3455 | 2032 | if res is False: |
2d2fa82d | 2033 | return [] |
47a5cb77 | 2034 | mpd_doc, urlh = res |
c25720ef RA |
2035 | if mpd_doc is None: |
2036 | return [] | |
02dc0a36 | 2037 | mpd_base_url = base_url(urlh.geturl()) |
1bac3455 | 2038 | |
91cb6b50 | 2039 | return self._parse_mpd_formats( |
47a5cb77 | 2040 | mpd_doc, mpd_id=mpd_id, mpd_base_url=mpd_base_url, |
86f4d14f | 2041 | formats_dict=formats_dict, mpd_url=mpd_url) |
2d2fa82d | 2042 | |
86f4d14f | 2043 | def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}, mpd_url=None): |
f0948348 S |
2044 | """ |
2045 | Parse formats from MPD manifest. | |
2046 | References: | |
2047 | 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E), | |
2048 | http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip | |
2049 | 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP | |
2050 | """ | |
1bac3455 | 2051 | if mpd_doc.get('type') == 'dynamic': |
2052 | return [] | |
2d2fa82d | 2053 | |
91cb6b50 | 2054 | namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None) |
f14be228 | 2055 | |
2056 | def _add_ns(path): | |
2057 | return self._xpath_ns(path, namespace) | |
2058 | ||
675d0016 | 2059 | def is_drm_protected(element): |
2060 | return element.find(_add_ns('ContentProtection')) is not None | |
2061 | ||
1bac3455 | 2062 | def extract_multisegment_info(element, ms_parent_info): |
2063 | ms_info = ms_parent_info.copy() | |
b4c1d6e8 S |
2064 | |
2065 | # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some | |
2066 | # common attributes and elements. We will only extract relevant | |
2067 | # for us. | |
2068 | def extract_common(source): | |
2069 | segment_timeline = source.find(_add_ns('SegmentTimeline')) | |
2070 | if segment_timeline is not None: | |
2071 | s_e = segment_timeline.findall(_add_ns('S')) | |
2072 | if s_e: | |
2073 | ms_info['total_number'] = 0 | |
2074 | ms_info['s'] = [] | |
2075 | for s in s_e: | |
2076 | r = int(s.get('r', 0)) | |
2077 | ms_info['total_number'] += 1 + r | |
2078 | ms_info['s'].append({ | |
2079 | 't': int(s.get('t', 0)), | |
2080 | # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60]) | |
2081 | 'd': int(s.attrib['d']), | |
2082 | 'r': r, | |
2083 | }) | |
2084 | start_number = source.get('startNumber') | |
2085 | if start_number: | |
2086 | ms_info['start_number'] = int(start_number) | |
2087 | timescale = source.get('timescale') | |
2088 | if timescale: | |
2089 | ms_info['timescale'] = int(timescale) | |
2090 | segment_duration = source.get('duration') | |
2091 | if segment_duration: | |
48504785 | 2092 | ms_info['segment_duration'] = float(segment_duration) |
b4c1d6e8 S |
2093 | |
2094 | def extract_Initialization(source): | |
2095 | initialization = source.find(_add_ns('Initialization')) | |
2096 | if initialization is not None: | |
2097 | ms_info['initialization_url'] = initialization.attrib['sourceURL'] | |
2098 | ||
f14be228 | 2099 | segment_list = element.find(_add_ns('SegmentList')) |
1bac3455 | 2100 | if segment_list is not None: |
b4c1d6e8 S |
2101 | extract_common(segment_list) |
2102 | extract_Initialization(segment_list) | |
f14be228 | 2103 | segment_urls_e = segment_list.findall(_add_ns('SegmentURL')) |
1bac3455 | 2104 | if segment_urls_e: |
2105 | ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e] | |
1bac3455 | 2106 | else: |
f14be228 | 2107 | segment_template = element.find(_add_ns('SegmentTemplate')) |
1bac3455 | 2108 | if segment_template is not None: |
b4c1d6e8 | 2109 | extract_common(segment_template) |
e228616c S |
2110 | media = segment_template.get('media') |
2111 | if media: | |
2112 | ms_info['media'] = media | |
1bac3455 | 2113 | initialization = segment_template.get('initialization') |
2114 | if initialization: | |
e228616c | 2115 | ms_info['initialization'] = initialization |
1bac3455 | 2116 | else: |
b4c1d6e8 | 2117 | extract_Initialization(segment_template) |
1bac3455 | 2118 | return ms_info |
b323e170 | 2119 | |
1bac3455 | 2120 | mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration')) |
17b598d3 | 2121 | formats = [] |
f14be228 | 2122 | for period in mpd_doc.findall(_add_ns('Period')): |
1bac3455 | 2123 | period_duration = parse_duration(period.get('duration')) or mpd_duration |
2124 | period_ms_info = extract_multisegment_info(period, { | |
2125 | 'start_number': 1, | |
2126 | 'timescale': 1, | |
2127 | }) | |
f14be228 | 2128 | for adaptation_set in period.findall(_add_ns('AdaptationSet')): |
675d0016 | 2129 | if is_drm_protected(adaptation_set): |
2130 | continue | |
1bac3455 | 2131 | adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info) |
f14be228 | 2132 | for representation in adaptation_set.findall(_add_ns('Representation')): |
675d0016 | 2133 | if is_drm_protected(representation): |
2134 | continue | |
1bac3455 | 2135 | representation_attrib = adaptation_set.attrib.copy() |
2136 | representation_attrib.update(representation.attrib) | |
f0948348 | 2137 | # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory |
a6c8b759 YCH |
2138 | mime_type = representation_attrib['mimeType'] |
2139 | content_type = mime_type.split('/')[0] | |
1bac3455 | 2140 | if content_type == 'text': |
2141 | # TODO implement WebVTT downloading | |
2142 | pass | |
40fcba5e | 2143 | elif content_type in ('video', 'audio'): |
1bac3455 | 2144 | base_url = '' |
2145 | for element in (representation, adaptation_set, period, mpd_doc): | |
f14be228 | 2146 | base_url_e = element.find(_add_ns('BaseURL')) |
1bac3455 | 2147 | if base_url_e is not None: |
2148 | base_url = base_url_e.text + base_url | |
2149 | if re.match(r'^https?://', base_url): | |
2150 | break | |
bb20526b S |
2151 | if mpd_base_url and not re.match(r'^https?://', base_url): |
2152 | if not mpd_base_url.endswith('/') and not base_url.startswith('/'): | |
2153 | mpd_base_url += '/' | |
1bac3455 | 2154 | base_url = mpd_base_url + base_url |
2155 | representation_id = representation_attrib.get('id') | |
d577c796 | 2156 | lang = representation_attrib.get('lang') |
51e9094f | 2157 | url_el = representation.find(_add_ns('BaseURL')) |
2158 | 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 | 2159 | bandwidth = int_or_none(representation_attrib.get('bandwidth')) |
1bac3455 | 2160 | f = { |
154c209e | 2161 | 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id, |
86f4d14f | 2162 | 'manifest_url': mpd_url, |
a6c8b759 | 2163 | 'ext': mimetype2ext(mime_type), |
1bac3455 | 2164 | 'width': int_or_none(representation_attrib.get('width')), |
2165 | 'height': int_or_none(representation_attrib.get('height')), | |
9c99bef7 | 2166 | 'tbr': float_or_none(bandwidth, 1000), |
1bac3455 | 2167 | 'asr': int_or_none(representation_attrib.get('audioSamplingRate')), |
2168 | 'fps': int_or_none(representation_attrib.get('frameRate')), | |
d577c796 | 2169 | 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None, |
1bac3455 | 2170 | 'format_note': 'DASH %s' % content_type, |
51e9094f | 2171 | 'filesize': filesize, |
126f225b | 2172 | 'container': mimetype2ext(mime_type) + '_dash', |
1bac3455 | 2173 | } |
7fe15920 | 2174 | f.update(parse_codecs(representation_attrib.get('codecs'))) |
1bac3455 | 2175 | representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info) |
b4c1d6e8 | 2176 | |
e228616c | 2177 | def prepare_template(template_name, identifiers): |
eca1f0d1 S |
2178 | tmpl = representation_ms_info[template_name] |
2179 | # First of, % characters outside $...$ templates | |
2180 | # must be escaped by doubling for proper processing | |
2181 | # by % operator string formatting used further (see | |
067aa17e | 2182 | # https://github.com/ytdl-org/youtube-dl/issues/16867). |
eca1f0d1 S |
2183 | t = '' |
2184 | in_template = False | |
2185 | for c in tmpl: | |
2186 | t += c | |
2187 | if c == '$': | |
2188 | in_template = not in_template | |
2189 | elif c == '%' and not in_template: | |
2190 | t += c | |
2191 | # Next, $...$ templates are translated to their | |
2192 | # %(...) counterparts to be used with % operator | |
e228616c S |
2193 | t = t.replace('$RepresentationID$', representation_id) |
2194 | t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t) | |
2195 | t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t) | |
2196 | t.replace('$$', '$') | |
2197 | return t | |
2198 | ||
2199 | # @initialization is a regular template like @media one | |
2200 | # so it should be handled just the same way (see | |
067aa17e | 2201 | # https://github.com/ytdl-org/youtube-dl/issues/11605) |
e228616c S |
2202 | if 'initialization' in representation_ms_info: |
2203 | initialization_template = prepare_template( | |
2204 | 'initialization', | |
2205 | # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and | |
2206 | # $Time$ shall not be included for @initialization thus | |
2207 | # only $Bandwidth$ remains | |
2208 | ('Bandwidth', )) | |
2209 | representation_ms_info['initialization_url'] = initialization_template % { | |
2210 | 'Bandwidth': bandwidth, | |
2211 | } | |
2212 | ||
1141e910 S |
2213 | def location_key(location): |
2214 | return 'url' if re.match(r'^https?://', location) else 'path' | |
2215 | ||
e228616c S |
2216 | if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info: |
2217 | ||
2218 | media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time')) | |
1141e910 | 2219 | media_location_key = location_key(media_template) |
f0948348 S |
2220 | |
2221 | # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$ | |
2222 | # can't be used at the same time | |
b4c1d6e8 S |
2223 | if '%(Number' in media_template and 's' not in representation_ms_info: |
2224 | segment_duration = None | |
c110944f | 2225 | if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info: |
b4c1d6e8 S |
2226 | segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale']) |
2227 | representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration)) | |
b4c1d6e8 | 2228 | representation_ms_info['fragments'] = [{ |
1141e910 | 2229 | media_location_key: media_template % { |
b4c1d6e8 | 2230 | 'Number': segment_number, |
e228616c | 2231 | 'Bandwidth': bandwidth, |
b4c1d6e8 S |
2232 | }, |
2233 | 'duration': segment_duration, | |
2234 | } for segment_number in range( | |
2235 | representation_ms_info['start_number'], | |
2236 | representation_ms_info['total_number'] + representation_ms_info['start_number'])] | |
f0948348 | 2237 | else: |
b4c1d6e8 S |
2238 | # $Number*$ or $Time$ in media template with S list available |
2239 | # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg | |
2240 | # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411 | |
b4c1d6e8 | 2241 | representation_ms_info['fragments'] = [] |
f0948348 | 2242 | segment_time = 0 |
b4c1d6e8 S |
2243 | segment_d = None |
2244 | segment_number = representation_ms_info['start_number'] | |
f0948348 S |
2245 | |
2246 | def add_segment_url(): | |
b4c1d6e8 S |
2247 | segment_url = media_template % { |
2248 | 'Time': segment_time, | |
e228616c | 2249 | 'Bandwidth': bandwidth, |
b4c1d6e8 S |
2250 | 'Number': segment_number, |
2251 | } | |
b4c1d6e8 | 2252 | representation_ms_info['fragments'].append({ |
1141e910 | 2253 | media_location_key: segment_url, |
b4c1d6e8 S |
2254 | 'duration': float_or_none(segment_d, representation_ms_info['timescale']), |
2255 | }) | |
f0948348 S |
2256 | |
2257 | for num, s in enumerate(representation_ms_info['s']): | |
2258 | segment_time = s.get('t') or segment_time | |
b4c1d6e8 | 2259 | segment_d = s['d'] |
f0948348 | 2260 | add_segment_url() |
b4c1d6e8 | 2261 | segment_number += 1 |
f0948348 | 2262 | for r in range(s.get('r', 0)): |
b4c1d6e8 | 2263 | segment_time += segment_d |
f0948348 | 2264 | add_segment_url() |
b4c1d6e8 S |
2265 | segment_number += 1 |
2266 | segment_time += segment_d | |
2267 | elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info: | |
2268 | # No media template | |
2269 | # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI | |
2270 | # or any YouTube dashsegments video | |
2271 | fragments = [] | |
d04621da S |
2272 | segment_index = 0 |
2273 | timescale = representation_ms_info['timescale'] | |
2274 | for s in representation_ms_info['s']: | |
2275 | duration = float_or_none(s['d'], timescale) | |
b4c1d6e8 | 2276 | for r in range(s.get('r', 0) + 1): |
1141e910 | 2277 | segment_uri = representation_ms_info['segment_urls'][segment_index] |
b4c1d6e8 | 2278 | fragments.append({ |
1141e910 | 2279 | location_key(segment_uri): segment_uri, |
d04621da | 2280 | 'duration': duration, |
b4c1d6e8 | 2281 | }) |
d04621da | 2282 | segment_index += 1 |
b4c1d6e8 | 2283 | representation_ms_info['fragments'] = fragments |
41bf647e PN |
2284 | elif 'segment_urls' in representation_ms_info: |
2285 | # Segment URLs with no SegmentTimeline | |
2286 | # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091 | |
067aa17e | 2287 | # https://github.com/ytdl-org/youtube-dl/pull/14844 |
41bf647e | 2288 | fragments = [] |
603fc4e0 S |
2289 | segment_duration = float_or_none( |
2290 | representation_ms_info['segment_duration'], | |
2291 | representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None | |
41bf647e | 2292 | for segment_url in representation_ms_info['segment_urls']: |
603fc4e0 | 2293 | fragment = { |
41bf647e | 2294 | location_key(segment_url): segment_url, |
603fc4e0 S |
2295 | } |
2296 | if segment_duration: | |
2297 | fragment['duration'] = segment_duration | |
2298 | fragments.append(fragment) | |
41bf647e | 2299 | representation_ms_info['fragments'] = fragments |
79d2077e S |
2300 | # If there is a fragments key available then we correctly recognized fragmented media. |
2301 | # Otherwise we will assume unfragmented media with direct access. Technically, such | |
2302 | # assumption is not necessarily correct since we may simply have no support for | |
2303 | # some forms of fragmented media renditions yet, but for now we'll use this fallback. | |
86f4d14f | 2304 | if 'fragments' in representation_ms_info: |
1bac3455 | 2305 | f.update({ |
79d2077e S |
2306 | # NB: mpd_url may be empty when MPD manifest is parsed from a string |
2307 | 'url': mpd_url or base_url, | |
1141e910 | 2308 | 'fragment_base_url': base_url, |
b4c1d6e8 | 2309 | 'fragments': [], |
1bac3455 | 2310 | 'protocol': 'http_dash_segments', |
df374b52 | 2311 | }) |
1bac3455 | 2312 | if 'initialization_url' in representation_ms_info: |
e228616c | 2313 | initialization_url = representation_ms_info['initialization_url'] |
1bac3455 | 2314 | if not f.get('url'): |
2315 | f['url'] = initialization_url | |
1141e910 | 2316 | f['fragments'].append({location_key(initialization_url): initialization_url}) |
b4c1d6e8 | 2317 | f['fragments'].extend(representation_ms_info['fragments']) |
79d2077e S |
2318 | else: |
2319 | # Assuming direct URL to unfragmented media. | |
2320 | f['url'] = base_url | |
2321 | ||
9d6ac71c S |
2322 | # According to [1, 5.3.5.2, Table 7, page 35] @id of Representation |
2323 | # is not necessarily unique within a Period thus formats with | |
2324 | # the same `format_id` are quite possible. There are numerous examples | |
067aa17e S |
2325 | # of such manifests (see https://github.com/ytdl-org/youtube-dl/issues/15111, |
2326 | # https://github.com/ytdl-org/youtube-dl/issues/13919) | |
9d6ac71c S |
2327 | full_info = formats_dict.get(representation_id, {}).copy() |
2328 | full_info.update(f) | |
2329 | formats.append(full_info) | |
17b598d3 | 2330 | else: |
1bac3455 | 2331 | self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) |
17b598d3 YCH |
2332 | return formats |
2333 | ||
7360c06f | 2334 | def _extract_ism_formats(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}): |
47a5cb77 | 2335 | res = self._download_xml_handle( |
b2758123 RA |
2336 | ism_url, video_id, |
2337 | note=note or 'Downloading ISM manifest', | |
2338 | errnote=errnote or 'Failed to download ISM manifest', | |
7360c06f | 2339 | fatal=fatal, data=data, headers=headers, query=query) |
b2758123 RA |
2340 | if res is False: |
2341 | return [] | |
47a5cb77 | 2342 | ism_doc, urlh = res |
b2758123 | 2343 | |
7947a1f7 | 2344 | return self._parse_ism_formats(ism_doc, urlh.geturl(), ism_id) |
b2758123 RA |
2345 | |
2346 | def _parse_ism_formats(self, ism_doc, ism_url, ism_id=None): | |
76d5a363 S |
2347 | """ |
2348 | Parse formats from ISM manifest. | |
2349 | References: | |
2350 | 1. [MS-SSTR]: Smooth Streaming Protocol, | |
2351 | https://msdn.microsoft.com/en-us/library/ff469518.aspx | |
2352 | """ | |
b2758123 RA |
2353 | if ism_doc.get('IsLive') == 'TRUE' or ism_doc.find('Protection') is not None: |
2354 | return [] | |
2355 | ||
b2758123 RA |
2356 | duration = int(ism_doc.attrib['Duration']) |
2357 | timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000 | |
2358 | ||
2359 | formats = [] | |
2360 | for stream in ism_doc.findall('StreamIndex'): | |
2361 | stream_type = stream.get('Type') | |
2362 | if stream_type not in ('video', 'audio'): | |
2363 | continue | |
2364 | url_pattern = stream.attrib['Url'] | |
2365 | stream_timescale = int_or_none(stream.get('TimeScale')) or timescale | |
2366 | stream_name = stream.get('Name') | |
2367 | for track in stream.findall('QualityLevel'): | |
2501d41e | 2368 | fourcc = track.get('FourCC', 'AACL' if track.get('AudioTag') == '255' else None) |
b2758123 RA |
2369 | # TODO: add support for WVC1 and WMAP |
2370 | if fourcc not in ('H264', 'AVC1', 'AACL'): | |
2371 | self.report_warning('%s is not a supported codec' % fourcc) | |
2372 | continue | |
2373 | tbr = int(track.attrib['Bitrate']) // 1000 | |
76d5a363 S |
2374 | # [1] does not mention Width and Height attributes. However, |
2375 | # they're often present while MaxWidth and MaxHeight are | |
2376 | # missing, so should be used as fallbacks | |
2377 | width = int_or_none(track.get('MaxWidth') or track.get('Width')) | |
2378 | height = int_or_none(track.get('MaxHeight') or track.get('Height')) | |
b2758123 RA |
2379 | sampling_rate = int_or_none(track.get('SamplingRate')) |
2380 | ||
2381 | track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern) | |
2382 | track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern) | |
2383 | ||
2384 | fragments = [] | |
2385 | fragment_ctx = { | |
2386 | 'time': 0, | |
2387 | } | |
2388 | stream_fragments = stream.findall('c') | |
2389 | for stream_fragment_index, stream_fragment in enumerate(stream_fragments): | |
2390 | fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time'] | |
2391 | fragment_repeat = int_or_none(stream_fragment.get('r')) or 1 | |
2392 | fragment_ctx['duration'] = int_or_none(stream_fragment.get('d')) | |
2393 | if not fragment_ctx['duration']: | |
2394 | try: | |
2395 | next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t']) | |
2396 | except IndexError: | |
2397 | next_fragment_time = duration | |
1616f9b4 | 2398 | fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat |
b2758123 RA |
2399 | for _ in range(fragment_repeat): |
2400 | fragments.append({ | |
1616f9b4 | 2401 | 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern), |
b2758123 RA |
2402 | 'duration': fragment_ctx['duration'] / stream_timescale, |
2403 | }) | |
2404 | fragment_ctx['time'] += fragment_ctx['duration'] | |
2405 | ||
2406 | format_id = [] | |
2407 | if ism_id: | |
2408 | format_id.append(ism_id) | |
2409 | if stream_name: | |
2410 | format_id.append(stream_name) | |
2411 | format_id.append(compat_str(tbr)) | |
2412 | ||
2413 | formats.append({ | |
2414 | 'format_id': '-'.join(format_id), | |
2415 | 'url': ism_url, | |
2416 | 'manifest_url': ism_url, | |
2417 | 'ext': 'ismv' if stream_type == 'video' else 'isma', | |
2418 | 'width': width, | |
2419 | 'height': height, | |
2420 | 'tbr': tbr, | |
2421 | 'asr': sampling_rate, | |
2422 | 'vcodec': 'none' if stream_type == 'audio' else fourcc, | |
2423 | 'acodec': 'none' if stream_type == 'video' else fourcc, | |
2424 | 'protocol': 'ism', | |
2425 | 'fragments': fragments, | |
2426 | '_download_params': { | |
2427 | 'duration': duration, | |
2428 | 'timescale': stream_timescale, | |
2429 | 'width': width or 0, | |
2430 | 'height': height or 0, | |
2431 | 'fourcc': fourcc, | |
2432 | 'codec_private_data': track.get('CodecPrivateData'), | |
2433 | 'sampling_rate': sampling_rate, | |
2434 | 'channels': int_or_none(track.get('Channels', 2)), | |
2435 | 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)), | |
2436 | 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)), | |
2437 | }, | |
2438 | }) | |
2439 | return formats | |
2440 | ||
eeb0a956 | 2441 | 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 |
2442 | def absolute_url(item_url): |
2443 | return urljoin(base_url, item_url) | |
59bbe491 | 2444 | |
2445 | def parse_content_type(content_type): | |
2446 | if not content_type: | |
2447 | return {} | |
2448 | ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type) | |
2449 | if ctr: | |
2450 | mimetype, codecs = ctr.groups() | |
2451 | f = parse_codecs(codecs) | |
2452 | f['ext'] = mimetype2ext(mimetype) | |
2453 | return f | |
2454 | return {} | |
2455 | ||
868f79db | 2456 | def _media_formats(src, cur_media_type, type_info={}): |
520251c0 | 2457 | full_url = absolute_url(src) |
82889d4a | 2458 | ext = type_info.get('ext') or determine_ext(full_url) |
87a449c1 | 2459 | if ext == 'm3u8': |
520251c0 YCH |
2460 | is_plain_url = False |
2461 | formats = self._extract_m3u8_formats( | |
ad120ae1 | 2462 | full_url, video_id, ext='mp4', |
eeb0a956 | 2463 | entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id, |
b359e977 | 2464 | preference=preference, fatal=False) |
87a449c1 S |
2465 | elif ext == 'mpd': |
2466 | is_plain_url = False | |
2467 | formats = self._extract_mpd_formats( | |
b359e977 | 2468 | full_url, video_id, mpd_id=mpd_id, fatal=False) |
520251c0 YCH |
2469 | else: |
2470 | is_plain_url = True | |
2471 | formats = [{ | |
2472 | 'url': full_url, | |
2473 | 'vcodec': 'none' if cur_media_type == 'audio' else None, | |
2474 | }] | |
2475 | return is_plain_url, formats | |
2476 | ||
59bbe491 | 2477 | entries = [] |
4328ddf8 S |
2478 | # amp-video and amp-audio are very similar to their HTML5 counterparts |
2479 | # so we wll include them right here (see | |
2480 | # https://www.ampproject.org/docs/reference/components/amp-video) | |
cea364f7 YCH |
2481 | media_tags = [(media_tag, media_type, '') |
2482 | for media_tag, media_type | |
4328ddf8 | 2483 | in re.findall(r'(?s)(<(?:amp-)?(video|audio)[^>]*/>)', webpage)] |
2aec7256 S |
2484 | media_tags.extend(re.findall( |
2485 | # We only allow video|audio followed by a whitespace or '>'. | |
2486 | # Allowing more characters may end up in significant slow down (see | |
067aa17e | 2487 | # https://github.com/ytdl-org/youtube-dl/issues/11979, example URL: |
2aec7256 | 2488 | # http://www.porntrex.com/maps/videositemap.xml). |
4328ddf8 | 2489 | r'(?s)(<(?P<tag>(?:amp-)?(?:video|audio))(?:\s+[^>]*)?>)(.*?)</(?P=tag)>', webpage)) |
cea364f7 | 2490 | for media_tag, media_type, media_content in media_tags: |
59bbe491 | 2491 | media_info = { |
2492 | 'formats': [], | |
2493 | 'subtitles': {}, | |
2494 | } | |
2495 | media_attributes = extract_attributes(media_tag) | |
f856816b | 2496 | src = strip_or_none(media_attributes.get('src')) |
59bbe491 | 2497 | if src: |
dedb1770 | 2498 | _, formats = _media_formats(src, media_type) |
520251c0 | 2499 | media_info['formats'].extend(formats) |
6780154e | 2500 | media_info['thumbnail'] = absolute_url(media_attributes.get('poster')) |
59bbe491 | 2501 | if media_content: |
2502 | for source_tag in re.findall(r'<source[^>]+>', media_content): | |
d493f15c S |
2503 | s_attr = extract_attributes(source_tag) |
2504 | # data-video-src and data-src are non standard but seen | |
2505 | # several times in the wild | |
f856816b | 2506 | src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src'))) |
59bbe491 | 2507 | if not src: |
2508 | continue | |
d493f15c | 2509 | f = parse_content_type(s_attr.get('type')) |
868f79db | 2510 | is_plain_url, formats = _media_formats(src, media_type, f) |
520251c0 | 2511 | if is_plain_url: |
d493f15c S |
2512 | # width, height, res, label and title attributes are |
2513 | # all not standard but seen several times in the wild | |
2514 | labels = [ | |
2515 | s_attr.get(lbl) | |
2516 | for lbl in ('label', 'title') | |
2517 | if str_or_none(s_attr.get(lbl)) | |
2518 | ] | |
2519 | width = int_or_none(s_attr.get('width')) | |
3089bc74 S |
2520 | height = (int_or_none(s_attr.get('height')) |
2521 | or int_or_none(s_attr.get('res'))) | |
d493f15c S |
2522 | if not width or not height: |
2523 | for lbl in labels: | |
2524 | resolution = parse_resolution(lbl) | |
2525 | if not resolution: | |
2526 | continue | |
2527 | width = width or resolution.get('width') | |
2528 | height = height or resolution.get('height') | |
2529 | for lbl in labels: | |
2530 | tbr = parse_bitrate(lbl) | |
2531 | if tbr: | |
2532 | break | |
2533 | else: | |
2534 | tbr = None | |
1ed45499 | 2535 | f.update({ |
d493f15c S |
2536 | 'width': width, |
2537 | 'height': height, | |
2538 | 'tbr': tbr, | |
2539 | 'format_id': s_attr.get('label') or s_attr.get('title'), | |
1ed45499 | 2540 | }) |
520251c0 YCH |
2541 | f.update(formats[0]) |
2542 | media_info['formats'].append(f) | |
2543 | else: | |
2544 | media_info['formats'].extend(formats) | |
59bbe491 | 2545 | for track_tag in re.findall(r'<track[^>]+>', media_content): |
2546 | track_attributes = extract_attributes(track_tag) | |
2547 | kind = track_attributes.get('kind') | |
5968d7d2 | 2548 | if not kind or kind in ('subtitles', 'captions'): |
f856816b | 2549 | src = strip_or_none(track_attributes.get('src')) |
59bbe491 | 2550 | if not src: |
2551 | continue | |
2552 | lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label') | |
2553 | media_info['subtitles'].setdefault(lang, []).append({ | |
2554 | 'url': absolute_url(src), | |
2555 | }) | |
5e8e2fa5 S |
2556 | for f in media_info['formats']: |
2557 | f.setdefault('http_headers', {})['Referer'] = base_url | |
5968d7d2 | 2558 | if media_info['formats'] or media_info['subtitles']: |
59bbe491 | 2559 | entries.append(media_info) |
2560 | return entries | |
2561 | ||
c4251b9a | 2562 | def _extract_akamai_formats(self, manifest_url, video_id, hosts={}): |
c7c43a93 | 2563 | formats = [] |
e71a4509 | 2564 | hdcore_sign = 'hdcore=3.7.0' |
ff6f9a67 | 2565 | f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') |
c4251b9a RA |
2566 | hds_host = hosts.get('hds') |
2567 | if hds_host: | |
2568 | f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url) | |
e71a4509 RA |
2569 | if 'hdcore=' not in f4m_url: |
2570 | f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign | |
2571 | f4m_formats = self._extract_f4m_formats( | |
2572 | f4m_url, video_id, f4m_id='hds', fatal=False) | |
2573 | for entry in f4m_formats: | |
2574 | entry.update({'extra_param_to_segment_url': hdcore_sign}) | |
2575 | formats.extend(f4m_formats) | |
c4251b9a RA |
2576 | m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8') |
2577 | hls_host = hosts.get('hls') | |
2578 | if hls_host: | |
2579 | m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url) | |
c7c43a93 RA |
2580 | formats.extend(self._extract_m3u8_formats( |
2581 | m3u8_url, video_id, 'mp4', 'm3u8_native', | |
2582 | m3u8_id='hls', fatal=False)) | |
2583 | return formats | |
2584 | ||
6ad02195 | 2585 | def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]): |
044eeb14 | 2586 | query = compat_urlparse.urlparse(url).query |
6ad02195 | 2587 | url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url) |
240f2622 S |
2588 | mobj = re.search( |
2589 | r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url) | |
2590 | url_base = mobj.group('url') | |
2591 | http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base) | |
6ad02195 | 2592 | formats = [] |
044eeb14 S |
2593 | |
2594 | def manifest_url(manifest): | |
2595 | m_url = '%s/%s' % (http_base_url, manifest) | |
2596 | if query: | |
2597 | m_url += '?%s' % query | |
2598 | return m_url | |
2599 | ||
6ad02195 RA |
2600 | if 'm3u8' not in skip_protocols: |
2601 | formats.extend(self._extract_m3u8_formats( | |
044eeb14 | 2602 | manifest_url('playlist.m3u8'), video_id, 'mp4', |
6ad02195 RA |
2603 | m3u8_entry_protocol, m3u8_id='hls', fatal=False)) |
2604 | if 'f4m' not in skip_protocols: | |
2605 | formats.extend(self._extract_f4m_formats( | |
044eeb14 | 2606 | manifest_url('manifest.f4m'), |
6ad02195 | 2607 | video_id, f4m_id='hds', fatal=False)) |
0384932e RA |
2608 | if 'dash' not in skip_protocols: |
2609 | formats.extend(self._extract_mpd_formats( | |
044eeb14 | 2610 | manifest_url('manifest.mpd'), |
0384932e | 2611 | video_id, mpd_id='dash', fatal=False)) |
6ad02195 | 2612 | if re.search(r'(?:/smil:|\.smil)', url_base): |
6ad02195 RA |
2613 | if 'smil' not in skip_protocols: |
2614 | rtmp_formats = self._extract_smil_formats( | |
044eeb14 | 2615 | manifest_url('jwplayer.smil'), |
6ad02195 RA |
2616 | video_id, fatal=False) |
2617 | for rtmp_format in rtmp_formats: | |
2618 | rtsp_format = rtmp_format.copy() | |
2619 | rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path']) | |
2620 | del rtsp_format['play_path'] | |
2621 | del rtsp_format['ext'] | |
2622 | rtsp_format.update({ | |
2623 | 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'), | |
2624 | 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'), | |
2625 | 'protocol': 'rtsp', | |
2626 | }) | |
2627 | formats.extend([rtmp_format, rtsp_format]) | |
2628 | else: | |
2629 | for protocol in ('rtmp', 'rtsp'): | |
2630 | if protocol not in skip_protocols: | |
2631 | formats.append({ | |
f2e2f0c7 | 2632 | 'url': '%s:%s' % (protocol, url_base), |
6ad02195 RA |
2633 | 'format_id': protocol, |
2634 | 'protocol': protocol, | |
2635 | }) | |
2636 | return formats | |
2637 | ||
c73e330e | 2638 | def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json): |
a4a554a7 | 2639 | mobj = re.search( |
ac9c69ac | 2640 | r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)', |
a4a554a7 YCH |
2641 | webpage) |
2642 | if mobj: | |
c73e330e RU |
2643 | try: |
2644 | jwplayer_data = self._parse_json(mobj.group('options'), | |
2645 | video_id=video_id, | |
2646 | transform_source=transform_source) | |
2647 | except ExtractorError: | |
2648 | pass | |
2649 | else: | |
2650 | if isinstance(jwplayer_data, dict): | |
2651 | return jwplayer_data | |
a4a554a7 YCH |
2652 | |
2653 | def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs): | |
c73e330e RU |
2654 | jwplayer_data = self._find_jwplayer_data( |
2655 | webpage, video_id, transform_source=js_to_json) | |
a4a554a7 YCH |
2656 | return self._parse_jwplayer_data( |
2657 | jwplayer_data, video_id, *args, **kwargs) | |
2658 | ||
2659 | def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True, | |
2660 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
2661 | # JWPlayer backward compatibility: flattened playlists | |
2662 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96 | |
2663 | if 'playlist' not in jwplayer_data: | |
2664 | jwplayer_data = {'playlist': [jwplayer_data]} | |
2665 | ||
2666 | entries = [] | |
2667 | ||
2668 | # JWPlayer backward compatibility: single playlist item | |
2669 | # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10 | |
2670 | if not isinstance(jwplayer_data['playlist'], list): | |
2671 | jwplayer_data['playlist'] = [jwplayer_data['playlist']] | |
2672 | ||
2673 | for video_data in jwplayer_data['playlist']: | |
2674 | # JWPlayer backward compatibility: flattened sources | |
2675 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35 | |
2676 | if 'sources' not in video_data: | |
2677 | video_data['sources'] = [video_data] | |
2678 | ||
2679 | this_video_id = video_id or video_data['mediaid'] | |
2680 | ||
1a2192cb S |
2681 | formats = self._parse_jwplayer_formats( |
2682 | video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id, | |
2683 | mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url) | |
a4a554a7 YCH |
2684 | |
2685 | subtitles = {} | |
2686 | tracks = video_data.get('tracks') | |
2687 | if tracks and isinstance(tracks, list): | |
2688 | for track in tracks: | |
96a2daa1 S |
2689 | if not isinstance(track, dict): |
2690 | continue | |
f4b74272 S |
2691 | track_kind = track.get('kind') |
2692 | if not track_kind or not isinstance(track_kind, compat_str): | |
2693 | continue | |
2694 | if track_kind.lower() not in ('captions', 'subtitles'): | |
a4a554a7 YCH |
2695 | continue |
2696 | track_url = urljoin(base_url, track.get('file')) | |
2697 | if not track_url: | |
2698 | continue | |
2699 | subtitles.setdefault(track.get('label') or 'en', []).append({ | |
2700 | 'url': self._proto_relative_url(track_url) | |
2701 | }) | |
2702 | ||
50d808f5 | 2703 | entry = { |
a4a554a7 | 2704 | 'id': this_video_id, |
50d808f5 | 2705 | 'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')), |
f81dd65b | 2706 | 'description': clean_html(video_data.get('description')), |
6945b9e7 | 2707 | 'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))), |
a4a554a7 YCH |
2708 | 'timestamp': int_or_none(video_data.get('pubdate')), |
2709 | 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')), | |
2710 | 'subtitles': subtitles, | |
50d808f5 RA |
2711 | } |
2712 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32 | |
2713 | if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']): | |
2714 | entry.update({ | |
2715 | '_type': 'url_transparent', | |
2716 | 'url': formats[0]['url'], | |
2717 | }) | |
2718 | else: | |
2719 | self._sort_formats(formats) | |
2720 | entry['formats'] = formats | |
2721 | entries.append(entry) | |
a4a554a7 YCH |
2722 | if len(entries) == 1: |
2723 | return entries[0] | |
2724 | else: | |
2725 | return self.playlist_result(entries) | |
2726 | ||
ed0cf9b3 S |
2727 | def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None, |
2728 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
bf1b87cd | 2729 | urls = [] |
ed0cf9b3 | 2730 | formats = [] |
1a2192cb | 2731 | for source in jwplayer_sources_data: |
0a268c6e S |
2732 | if not isinstance(source, dict): |
2733 | continue | |
6945b9e7 RA |
2734 | source_url = urljoin( |
2735 | base_url, self._proto_relative_url(source.get('file'))) | |
2736 | if not source_url or source_url in urls: | |
bf1b87cd RA |
2737 | continue |
2738 | urls.append(source_url) | |
ed0cf9b3 S |
2739 | source_type = source.get('type') or '' |
2740 | ext = mimetype2ext(source_type) or determine_ext(source_url) | |
2741 | if source_type == 'hls' or ext == 'm3u8': | |
2742 | formats.extend(self._extract_m3u8_formats( | |
0236cd0d S |
2743 | source_url, video_id, 'mp4', entry_protocol='m3u8_native', |
2744 | m3u8_id=m3u8_id, fatal=False)) | |
0d9c48de | 2745 | elif source_type == 'dash' or ext == 'mpd': |
ed0cf9b3 S |
2746 | formats.extend(self._extract_mpd_formats( |
2747 | source_url, video_id, mpd_id=mpd_id, fatal=False)) | |
b51dc9db S |
2748 | elif ext == 'smil': |
2749 | formats.extend(self._extract_smil_formats( | |
2750 | source_url, video_id, fatal=False)) | |
ed0cf9b3 | 2751 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67 |
0236cd0d S |
2752 | elif source_type.startswith('audio') or ext in ( |
2753 | 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'): | |
ed0cf9b3 S |
2754 | formats.append({ |
2755 | 'url': source_url, | |
2756 | 'vcodec': 'none', | |
2757 | 'ext': ext, | |
2758 | }) | |
2759 | else: | |
2760 | height = int_or_none(source.get('height')) | |
2761 | if height is None: | |
2762 | # Often no height is provided but there is a label in | |
0236cd0d | 2763 | # format like "1080p", "720p SD", or 1080. |
ed0cf9b3 | 2764 | height = int_or_none(self._search_regex( |
0236cd0d | 2765 | r'^(\d{3,4})[pP]?(?:\b|$)', compat_str(source.get('label') or ''), |
ed0cf9b3 S |
2766 | 'height', default=None)) |
2767 | a_format = { | |
2768 | 'url': source_url, | |
2769 | 'width': int_or_none(source.get('width')), | |
2770 | 'height': height, | |
0236cd0d | 2771 | 'tbr': int_or_none(source.get('bitrate')), |
ed0cf9b3 S |
2772 | 'ext': ext, |
2773 | } | |
2774 | if source_url.startswith('rtmp'): | |
2775 | a_format['ext'] = 'flv' | |
ed0cf9b3 S |
2776 | # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as |
2777 | # of jwplayer.flash.swf | |
2778 | rtmp_url_parts = re.split( | |
2779 | r'((?:mp4|mp3|flv):)', source_url, 1) | |
2780 | if len(rtmp_url_parts) == 3: | |
2781 | rtmp_url, prefix, play_path = rtmp_url_parts | |
2782 | a_format.update({ | |
2783 | 'url': rtmp_url, | |
2784 | 'play_path': prefix + play_path, | |
2785 | }) | |
2786 | if rtmp_params: | |
2787 | a_format.update(rtmp_params) | |
2788 | formats.append(a_format) | |
2789 | return formats | |
2790 | ||
f4b1c7ad PH |
2791 | def _live_title(self, name): |
2792 | """ Generate the title for a live video """ | |
2793 | now = datetime.datetime.now() | |
611c1dd9 | 2794 | now_str = now.strftime('%Y-%m-%d %H:%M') |
f4b1c7ad PH |
2795 | return name + ' ' + now_str |
2796 | ||
b14f3a4c PH |
2797 | def _int(self, v, name, fatal=False, **kwargs): |
2798 | res = int_or_none(v, **kwargs) | |
2799 | if 'get_attr' in kwargs: | |
2800 | print(getattr(v, kwargs['get_attr'])) | |
2801 | if res is None: | |
2802 | msg = 'Failed to extract %s: Could not parse value %r' % (name, v) | |
2803 | if fatal: | |
2804 | raise ExtractorError(msg) | |
2805 | else: | |
2806 | self._downloader.report_warning(msg) | |
2807 | return res | |
2808 | ||
2809 | def _float(self, v, name, fatal=False, **kwargs): | |
2810 | res = float_or_none(v, **kwargs) | |
2811 | if res is None: | |
2812 | msg = 'Failed to extract %s: Could not parse value %r' % (name, v) | |
2813 | if fatal: | |
2814 | raise ExtractorError(msg) | |
2815 | else: | |
2816 | self._downloader.report_warning(msg) | |
2817 | return res | |
2818 | ||
40e41780 TF |
2819 | def _set_cookie(self, domain, name, value, expire_time=None, port=None, |
2820 | path='/', secure=False, discard=False, rest={}, **kwargs): | |
810fb84d | 2821 | cookie = compat_cookiejar.Cookie( |
4ed2d7b7 | 2822 | 0, name, value, port, port is not None, domain, True, |
40e41780 TF |
2823 | domain.startswith('.'), path, True, secure, expire_time, |
2824 | discard, None, None, rest) | |
42939b61 JMF |
2825 | self._downloader.cookiejar.set_cookie(cookie) |
2826 | ||
799207e8 | 2827 | def _get_cookies(self, url): |
2828 | """ Return a compat_cookies.SimpleCookie with the cookies for the url """ | |
5c2266df | 2829 | req = sanitized_Request(url) |
799207e8 | 2830 | self._downloader.cookiejar.add_cookie_header(req) |
2831 | return compat_cookies.SimpleCookie(req.get_header('Cookie')) | |
2832 | ||
e3c1266f | 2833 | def _apply_first_set_cookie_header(self, url_handle, cookie): |
ce2fe4c0 S |
2834 | """ |
2835 | Apply first Set-Cookie header instead of the last. Experimental. | |
2836 | ||
2837 | Some sites (e.g. [1-3]) may serve two cookies under the same name | |
2838 | in Set-Cookie header and expect the first (old) one to be set rather | |
2839 | than second (new). However, as of RFC6265 the newer one cookie | |
2840 | should be set into cookie store what actually happens. | |
2841 | We will workaround this issue by resetting the cookie to | |
2842 | the first one manually. | |
2843 | 1. https://new.vk.com/ | |
2844 | 2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201 | |
2845 | 3. https://learning.oreilly.com/ | |
2846 | """ | |
e3c1266f S |
2847 | for header, cookies in url_handle.headers.items(): |
2848 | if header.lower() != 'set-cookie': | |
2849 | continue | |
2850 | if sys.version_info[0] >= 3: | |
2851 | cookies = cookies.encode('iso-8859-1') | |
2852 | cookies = cookies.decode('utf-8') | |
2853 | cookie_value = re.search( | |
2854 | r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies) | |
2855 | if cookie_value: | |
2856 | value, domain = cookie_value.groups() | |
2857 | self._set_cookie(domain, cookie, value) | |
2858 | break | |
2859 | ||
05900629 PH |
2860 | def get_testcases(self, include_onlymatching=False): |
2861 | t = getattr(self, '_TEST', None) | |
2862 | if t: | |
2863 | assert not hasattr(self, '_TESTS'), \ | |
2864 | '%s has _TEST and _TESTS' % type(self).__name__ | |
2865 | tests = [t] | |
2866 | else: | |
2867 | tests = getattr(self, '_TESTS', []) | |
2868 | for t in tests: | |
2869 | if not include_onlymatching and t.get('only_matching', False): | |
2870 | continue | |
2871 | t['name'] = type(self).__name__[:-len('IE')] | |
2872 | yield t | |
2873 | ||
2874 | def is_suitable(self, age_limit): | |
2875 | """ Test whether the extractor is generally suitable for the given | |
2876 | age limit (i.e. pornographic sites are not, all others usually are) """ | |
2877 | ||
2878 | any_restricted = False | |
2879 | for tc in self.get_testcases(include_onlymatching=False): | |
40090e8d | 2880 | if tc.get('playlist', []): |
05900629 PH |
2881 | tc = tc['playlist'][0] |
2882 | is_restricted = age_restricted( | |
2883 | tc.get('info_dict', {}).get('age_limit'), age_limit) | |
2884 | if not is_restricted: | |
2885 | return True | |
2886 | any_restricted = any_restricted or is_restricted | |
2887 | return not any_restricted | |
2888 | ||
a504ced0 | 2889 | def extract_subtitles(self, *args, **kwargs): |
3089bc74 S |
2890 | if (self._downloader.params.get('writesubtitles', False) |
2891 | or self._downloader.params.get('listsubtitles')): | |
9868ea49 JMF |
2892 | return self._get_subtitles(*args, **kwargs) |
2893 | return {} | |
a504ced0 JMF |
2894 | |
2895 | def _get_subtitles(self, *args, **kwargs): | |
611c1dd9 | 2896 | raise NotImplementedError('This method must be implemented by subclasses') |
a504ced0 | 2897 | |
912e0b7e YCH |
2898 | @staticmethod |
2899 | def _merge_subtitle_items(subtitle_list1, subtitle_list2): | |
2900 | """ Merge subtitle items for one language. Items with duplicated URLs | |
2901 | will be dropped. """ | |
2902 | list1_urls = set([item['url'] for item in subtitle_list1]) | |
2903 | ret = list(subtitle_list1) | |
2904 | ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls]) | |
2905 | return ret | |
2906 | ||
2907 | @classmethod | |
8c97f819 | 2908 | def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2): |
912e0b7e | 2909 | """ Merge two subtitle dictionaries, language by language. """ |
912e0b7e YCH |
2910 | ret = dict(subtitle_dict1) |
2911 | for lang in subtitle_dict2: | |
8c97f819 | 2912 | ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang]) |
912e0b7e YCH |
2913 | return ret |
2914 | ||
360e1ca5 | 2915 | def extract_automatic_captions(self, *args, **kwargs): |
3089bc74 S |
2916 | if (self._downloader.params.get('writeautomaticsub', False) |
2917 | or self._downloader.params.get('listsubtitles')): | |
9868ea49 JMF |
2918 | return self._get_automatic_captions(*args, **kwargs) |
2919 | return {} | |
360e1ca5 JMF |
2920 | |
2921 | def _get_automatic_captions(self, *args, **kwargs): | |
611c1dd9 | 2922 | raise NotImplementedError('This method must be implemented by subclasses') |
360e1ca5 | 2923 | |
d77ab8e2 | 2924 | def mark_watched(self, *args, **kwargs): |
3089bc74 S |
2925 | if (self._downloader.params.get('mark_watched', False) |
2926 | and (self._get_login_info()[0] is not None | |
2927 | or self._downloader.params.get('cookiefile') is not None)): | |
d77ab8e2 S |
2928 | self._mark_watched(*args, **kwargs) |
2929 | ||
2930 | def _mark_watched(self, *args, **kwargs): | |
2931 | raise NotImplementedError('This method must be implemented by subclasses') | |
2932 | ||
38cce791 YCH |
2933 | def geo_verification_headers(self): |
2934 | headers = {} | |
2935 | geo_verification_proxy = self._downloader.params.get('geo_verification_proxy') | |
2936 | if geo_verification_proxy: | |
2937 | headers['Ytdl-request-proxy'] = geo_verification_proxy | |
2938 | return headers | |
2939 | ||
98763ee3 YCH |
2940 | def _generic_id(self, url): |
2941 | return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0]) | |
2942 | ||
2943 | def _generic_title(self, url): | |
2944 | return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0]) | |
2945 | ||
8dbe9899 | 2946 | |
d6983cb4 PH |
2947 | class SearchInfoExtractor(InfoExtractor): |
2948 | """ | |
2949 | Base class for paged search queries extractors. | |
10952eb2 | 2950 | They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query} |
d6983cb4 PH |
2951 | Instances should define _SEARCH_KEY and _MAX_RESULTS. |
2952 | """ | |
2953 | ||
2954 | @classmethod | |
2955 | def _make_valid_url(cls): | |
2956 | return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY | |
2957 | ||
2958 | @classmethod | |
2959 | def suitable(cls, url): | |
2960 | return re.match(cls._make_valid_url(), url) is not None | |
2961 | ||
2962 | def _real_extract(self, query): | |
2963 | mobj = re.match(self._make_valid_url(), query) | |
2964 | if mobj is None: | |
f1a9d64e | 2965 | raise ExtractorError('Invalid search query "%s"' % query) |
d6983cb4 PH |
2966 | |
2967 | prefix = mobj.group('prefix') | |
2968 | query = mobj.group('query') | |
2969 | if prefix == '': | |
2970 | return self._get_n_results(query, 1) | |
2971 | elif prefix == 'all': | |
2972 | return self._get_n_results(query, self._MAX_RESULTS) | |
2973 | else: | |
2974 | n = int(prefix) | |
2975 | if n <= 0: | |
f1a9d64e | 2976 | raise ExtractorError('invalid download number %s for query "%s"' % (n, query)) |
d6983cb4 | 2977 | elif n > self._MAX_RESULTS: |
f1a9d64e | 2978 | self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n)) |
d6983cb4 PH |
2979 | n = self._MAX_RESULTS |
2980 | return self._get_n_results(query, n) | |
2981 | ||
2982 | def _get_n_results(self, query, n): | |
2983 | """Get a specified number of results for a query""" | |
611c1dd9 | 2984 | raise NotImplementedError('This method must be implemented by subclasses') |
0f818663 PH |
2985 | |
2986 | @property | |
2987 | def SEARCH_KEY(self): | |
2988 | return self._SEARCH_KEY |