]>
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, |
e9c0cdd3 | 20 | compat_etree_fromstring, |
e64b7569 | 21 | compat_getpass, |
d6983cb4 | 22 | compat_http_client, |
e9c0cdd3 YCH |
23 | compat_os_name, |
24 | compat_str, | |
d6983cb4 | 25 | compat_urllib_error, |
98763ee3 | 26 | compat_urllib_parse_unquote, |
15707c7e | 27 | compat_urllib_parse_urlencode, |
41d06b04 | 28 | compat_urllib_request, |
f0b5d6af | 29 | compat_urlparse, |
e01c3d2e | 30 | compat_xml_parse_error, |
8c25f81b | 31 | ) |
b22ca762 | 32 | from ..downloader.f4m import remove_encrypted_media |
8c25f81b | 33 | from ..utils import ( |
c342041f | 34 | NO_DEFAULT, |
05900629 | 35 | age_restricted, |
02dc0a36 | 36 | base_url, |
08f2a92c | 37 | bug_reports_message, |
d6983cb4 PH |
38 | clean_html, |
39 | compiled_regex_type, | |
70f0f5a8 | 40 | determine_ext, |
46b18f23 | 41 | determine_protocol, |
9b9c5355 | 42 | error_to_compat_str, |
d6983cb4 | 43 | ExtractorError, |
46b18f23 | 44 | extract_attributes, |
97f4aecf | 45 | fix_xml_ampersands, |
b14f3a4c | 46 | float_or_none, |
773f291d S |
47 | GeoRestrictedError, |
48 | GeoUtils, | |
31bb8d3f | 49 | int_or_none, |
a4a554a7 | 50 | js_to_json, |
46b18f23 JH |
51 | mimetype2ext, |
52 | orderedSet, | |
53 | parse_codecs, | |
54 | parse_duration, | |
4ca2a3cf | 55 | parse_iso8601, |
46b18f23 | 56 | parse_m3u8_attributes, |
55b3e45b | 57 | RegexNotFoundError, |
5c2266df | 58 | sanitized_Request, |
46b18f23 | 59 | sanitize_filename, |
f38de77f | 60 | unescapeHTML, |
647eab45 | 61 | unified_strdate, |
6b3a3098 | 62 | unified_timestamp, |
46b18f23 JH |
63 | update_Request, |
64 | update_url_query, | |
65 | urljoin, | |
a107193e | 66 | url_basename, |
a6571f10 | 67 | xpath_element, |
8d6765cf S |
68 | xpath_text, |
69 | xpath_with_ns, | |
d6983cb4 | 70 | ) |
c342041f | 71 | |
d6983cb4 PH |
72 | |
73 | class InfoExtractor(object): | |
74 | """Information Extractor class. | |
75 | ||
76 | Information extractors are the classes that, given a URL, extract | |
77 | information about the video (or videos) the URL refers to. This | |
78 | information includes the real video URL, the video title, author and | |
79 | others. The information is stored in a dictionary which is then | |
5d380852 | 80 | passed to the YoutubeDL. The YoutubeDL processes this |
d6983cb4 PH |
81 | information possibly downloading the video to the file system, among |
82 | other possible outcomes. | |
83 | ||
cf0649f8 | 84 | The type field determines the type of the result. |
fed5d032 PH |
85 | By far the most common value (and the default if _type is missing) is |
86 | "video", which indicates a single video. | |
87 | ||
88 | For a video, the dictionaries must include the following fields: | |
d6983cb4 PH |
89 | |
90 | id: Video identifier. | |
d6983cb4 | 91 | title: Video title, unescaped. |
d67b0b15 | 92 | |
f49d89ee | 93 | Additionally, it must contain either a formats entry or a url one: |
d67b0b15 | 94 | |
f49d89ee PH |
95 | formats: A list of dictionaries for each format available, ordered |
96 | from worst to best quality. | |
97 | ||
98 | Potential fields: | |
86f4d14f S |
99 | * url Mandatory. The URL of the video file |
100 | * manifest_url | |
101 | The URL of the manifest file in case of | |
102 | fragmented media (DASH, hls, hds) | |
10952eb2 | 103 | * ext Will be calculated from URL if missing |
d67b0b15 PH |
104 | * format A human-readable description of the format |
105 | ("mp4 container with h264/opus"). | |
106 | Calculated from the format_id, width, height. | |
107 | and format_note fields if missing. | |
108 | * format_id A short description of the format | |
5d4f3985 PH |
109 | ("mp4_h264_opus" or "19"). |
110 | Technically optional, but strongly recommended. | |
d67b0b15 PH |
111 | * format_note Additional info about the format |
112 | ("3D" or "DASH video") | |
113 | * width Width of the video, if known | |
114 | * height Height of the video, if known | |
f49d89ee | 115 | * resolution Textual description of width and height |
7217e148 | 116 | * tbr Average bitrate of audio and video in KBit/s |
d67b0b15 PH |
117 | * abr Average audio bitrate in KBit/s |
118 | * acodec Name of the audio codec in use | |
dd27fd17 | 119 | * asr Audio sampling rate in Hertz |
d67b0b15 | 120 | * vbr Average video bitrate in KBit/s |
fbb21cf5 | 121 | * fps Frame rate |
d67b0b15 | 122 | * vcodec Name of the video codec in use |
1394ce65 | 123 | * container Name of the container format |
d67b0b15 | 124 | * filesize The number of bytes, if known in advance |
9732d77e | 125 | * filesize_approx An estimate for the number of bytes |
d67b0b15 | 126 | * player_url SWF Player URL (used for rtmpdump). |
c7deaa4c PH |
127 | * protocol The protocol that will be used for the actual |
128 | download, lower-case. | |
b04b8852 | 129 | "http", "https", "rtsp", "rtmp", "rtmpe", |
af7d5a63 | 130 | "m3u8", "m3u8_native" or "http_dash_segments". |
c58c2d63 S |
131 | * fragment_base_url |
132 | Base URL for fragments. Each fragment's path | |
133 | value (if present) will be relative to | |
134 | this URL. | |
135 | * fragments A list of fragments of a fragmented media. | |
136 | Each fragment entry must contain either an url | |
137 | or a path. If an url is present it should be | |
138 | considered by a client. Otherwise both path and | |
139 | fragment_base_url must be present. Here is | |
140 | the list of all potential fields: | |
141 | * "url" - fragment's URL | |
142 | * "path" - fragment's path relative to | |
143 | fragment_base_url | |
a0d5077c S |
144 | * "duration" (optional, int or float) |
145 | * "filesize" (optional, int) | |
f49d89ee | 146 | * preference Order number of this format. If this field is |
08d13955 | 147 | present and not None, the formats get sorted |
38d63d84 | 148 | by this field, regardless of all other values. |
f49d89ee PH |
149 | -1 for default (order by other properties), |
150 | -2 or smaller for less than default. | |
e65566a9 PH |
151 | < -1000 to hide the format (if there is |
152 | another one which is strictly better) | |
32f90364 PH |
153 | * language Language code, e.g. "de" or "en-US". |
154 | * language_preference Is this in the language mentioned in | |
155 | the URL? | |
aff2f4f4 PH |
156 | 10 if it's what the URL is about, |
157 | -1 for default (don't know), | |
158 | -10 otherwise, other values reserved for now. | |
5d73273f PH |
159 | * quality Order number of the video quality of this |
160 | format, irrespective of the file format. | |
161 | -1 for default (order by other properties), | |
162 | -2 or smaller for less than default. | |
c64ed2a3 PH |
163 | * source_preference Order number for this video source |
164 | (quality takes higher priority) | |
165 | -1 for default (order by other properties), | |
166 | -2 or smaller for less than default. | |
d769be6c PH |
167 | * http_headers A dictionary of additional HTTP headers |
168 | to add to the request. | |
6271f1ca | 169 | * stretched_ratio If given and not 1, indicates that the |
3dee7826 PH |
170 | video's pixels are not square. |
171 | width : height ratio as float. | |
172 | * no_resume The server does not support resuming the | |
173 | (HTTP or RTMP) download. Boolean. | |
174 | ||
c0ba0f48 | 175 | url: Final video URL. |
d6983cb4 | 176 | ext: Video filename extension. |
d67b0b15 PH |
177 | format: The video format, defaults to ext (used for --get-format) |
178 | player_url: SWF Player URL (used for rtmpdump). | |
2f5865cc | 179 | |
d6983cb4 PH |
180 | The following fields are optional: |
181 | ||
f5e43bc6 | 182 | alt_title: A secondary title of the video. |
0afef30b PH |
183 | display_id An alternative identifier for the video, not necessarily |
184 | unique, but available before title. Typically, id is | |
185 | something like "4234987", title "Dancing naked mole rats", | |
186 | and display_id "dancing-naked-mole-rats" | |
d5519808 | 187 | thumbnails: A list of dictionaries, with the following entries: |
cfb56d1a | 188 | * "id" (optional, string) - Thumbnail format ID |
d5519808 | 189 | * "url" |
cfb56d1a | 190 | * "preference" (optional, int) - quality of the image |
d5519808 PH |
191 | * "width" (optional, int) |
192 | * "height" (optional, int) | |
193 | * "resolution" (optional, string "{width}x{height"}, | |
194 | deprecated) | |
2de624fd | 195 | * "filesize" (optional, int) |
d6983cb4 | 196 | thumbnail: Full URL to a video thumbnail image. |
f5e43bc6 | 197 | description: Full video description. |
d6983cb4 | 198 | uploader: Full name of the video uploader. |
2bc0c46f | 199 | license: License name the video is licensed under. |
8a92e51c | 200 | creator: The creator of the video. |
8aab976b | 201 | release_date: The date (YYYYMMDD) when the video was released. |
955c4514 | 202 | timestamp: UNIX timestamp of the moment the video became available. |
d6983cb4 | 203 | upload_date: Video upload date (YYYYMMDD). |
955c4514 | 204 | If not explicitly set, calculated from timestamp. |
d6983cb4 | 205 | uploader_id: Nickname or id of the video uploader. |
7bcd2830 | 206 | uploader_url: Full URL to a personal webpage of the video uploader. |
da9ec3b9 | 207 | location: Physical location where the video was filmed. |
a504ced0 | 208 | subtitles: The available subtitles as a dictionary in the format |
4606c34e YCH |
209 | {tag: subformats}. "tag" is usually a language code, and |
210 | "subformats" is a list sorted from lower to higher | |
211 | preference, each element is a dictionary with the "ext" | |
212 | entry and one of: | |
a504ced0 | 213 | * "data": The subtitles file contents |
10952eb2 | 214 | * "url": A URL pointing to the subtitles file |
4bba3716 | 215 | "ext" will be calculated from URL if missing |
360e1ca5 JMF |
216 | automatic_captions: Like 'subtitles', used by the YoutubeIE for |
217 | automatically generated captions | |
62d231c0 | 218 | duration: Length of the video in seconds, as an integer or float. |
f3d29461 | 219 | view_count: How many users have watched the video on the platform. |
19e3dfc9 PH |
220 | like_count: Number of positive ratings of the video |
221 | dislike_count: Number of negative ratings of the video | |
02835c6b | 222 | repost_count: Number of reposts of the video |
2d30521a | 223 | average_rating: Average rating give by users, the scale used depends on the webpage |
19e3dfc9 | 224 | comment_count: Number of comments on the video |
dd622d7c PH |
225 | comments: A list of comments, each with one or more of the following |
226 | properties (all but one of text or html optional): | |
227 | * "author" - human-readable name of the comment author | |
228 | * "author_id" - user ID of the comment author | |
229 | * "id" - Comment ID | |
230 | * "html" - Comment as HTML | |
231 | * "text" - Plain text of the comment | |
232 | * "timestamp" - UNIX timestamp of comment | |
233 | * "parent" - ID of the comment this one is replying to. | |
234 | Set to "root" to indicate that this is a | |
235 | comment to the original video. | |
8dbe9899 | 236 | age_limit: Age restriction for the video, as an integer (years) |
10952eb2 | 237 | webpage_url: The URL to the video webpage, if given to youtube-dl it |
9103bbc5 JMF |
238 | should allow to get the same result again. (It will be set |
239 | by YoutubeDL if it's missing) | |
ad3bc6ac PH |
240 | categories: A list of categories that the video falls in, for example |
241 | ["Sports", "Berlin"] | |
864f24bd | 242 | tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"] |
7267bd53 PH |
243 | is_live: True, False, or None (=unknown). Whether this video is a |
244 | live stream that goes on instead of a fixed-length video. | |
7c80519c | 245 | start_time: Time in seconds where the reproduction should start, as |
10952eb2 | 246 | specified in the URL. |
297a564b | 247 | end_time: Time in seconds where the reproduction should end, as |
10952eb2 | 248 | specified in the URL. |
55949fed | 249 | chapters: A list of dictionaries, with the following entries: |
250 | * "start_time" - The start time of the chapter in seconds | |
251 | * "end_time" - The end time of the chapter in seconds | |
252 | * "title" (optional, string) | |
d6983cb4 | 253 | |
7109903e S |
254 | The following fields should only be used when the video belongs to some logical |
255 | chapter or section: | |
256 | ||
257 | chapter: Name or title of the chapter the video belongs to. | |
27bfd4e5 S |
258 | chapter_number: Number of the chapter the video belongs to, as an integer. |
259 | chapter_id: Id of the chapter the video belongs to, as a unicode string. | |
7109903e S |
260 | |
261 | The following fields should only be used when the video is an episode of some | |
8d76bdf1 | 262 | series, programme or podcast: |
7109903e S |
263 | |
264 | series: Title of the series or programme the video episode belongs to. | |
265 | season: Title of the season the video episode belongs to. | |
27bfd4e5 S |
266 | season_number: Number of the season the video episode belongs to, as an integer. |
267 | season_id: Id of the season the video episode belongs to, as a unicode string. | |
7109903e S |
268 | episode: Title of the video episode. Unlike mandatory video title field, |
269 | this field should denote the exact title of the video episode | |
270 | without any kind of decoration. | |
27bfd4e5 S |
271 | episode_number: Number of the video episode within a season, as an integer. |
272 | episode_id: Id of the video episode, as a unicode string. | |
7109903e | 273 | |
7a93ab5f S |
274 | The following fields should only be used when the media is a track or a part of |
275 | a music album: | |
276 | ||
277 | track: Title of the track. | |
278 | track_number: Number of the track within an album or a disc, as an integer. | |
279 | track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii), | |
280 | as a unicode string. | |
281 | artist: Artist(s) of the track. | |
282 | genre: Genre(s) of the track. | |
283 | album: Title of the album the track belongs to. | |
284 | album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc). | |
285 | album_artist: List of all artists appeared on the album (e.g. | |
286 | "Ash Borer / Fell Voices" or "Various Artists", useful for splits | |
287 | and compilations). | |
288 | disc_number: Number of the disc or other physical medium the track belongs to, | |
289 | as an integer. | |
290 | release_year: Year (YYYY) when the album was released. | |
291 | ||
deefc05b | 292 | Unless mentioned otherwise, the fields should be Unicode strings. |
d6983cb4 | 293 | |
d838b1bd PH |
294 | Unless mentioned otherwise, None is equivalent to absence of information. |
295 | ||
fed5d032 PH |
296 | |
297 | _type "playlist" indicates multiple videos. | |
b82f815f PH |
298 | There must be a key "entries", which is a list, an iterable, or a PagedList |
299 | object, each element of which is a valid dictionary by this specification. | |
fed5d032 | 300 | |
e0b9d78f S |
301 | Additionally, playlists can have "title", "description" and "id" attributes |
302 | with the same semantics as videos (see above). | |
fed5d032 PH |
303 | |
304 | ||
305 | _type "multi_video" indicates that there are multiple videos that | |
306 | form a single show, for examples multiple acts of an opera or TV episode. | |
307 | It must have an entries key like a playlist and contain all the keys | |
308 | required for a video at the same time. | |
309 | ||
310 | ||
311 | _type "url" indicates that the video must be extracted from another | |
312 | location, possibly by a different extractor. Its only required key is: | |
313 | "url" - the next URL to extract. | |
f58766ce PH |
314 | The key "ie_key" can be set to the class name (minus the trailing "IE", |
315 | e.g. "Youtube") if the extractor class is known in advance. | |
316 | Additionally, the dictionary may have any properties of the resolved entity | |
317 | known in advance, for example "title" if the title of the referred video is | |
fed5d032 PH |
318 | known ahead of time. |
319 | ||
320 | ||
321 | _type "url_transparent" entities have the same specification as "url", but | |
322 | indicate that the given additional information is more precise than the one | |
323 | associated with the resolved URL. | |
324 | This is useful when a site employs a video service that hosts the video and | |
325 | its technical metadata, but that video service does not embed a useful | |
326 | title, description etc. | |
327 | ||
328 | ||
d6983cb4 PH |
329 | Subclasses of this one should re-define the _real_initialize() and |
330 | _real_extract() methods and define a _VALID_URL regexp. | |
331 | Probably, they should also be added to the list of extractors. | |
332 | ||
4248dad9 | 333 | _GEO_BYPASS attribute may be set to False in order to disable |
773f291d S |
334 | geo restriction bypass mechanisms for a particular extractor. |
335 | Though it won't disable explicit geo restriction bypass based on | |
4248dad9 S |
336 | country code provided with geo_bypass_country. (experimental) |
337 | ||
338 | _GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted | |
339 | countries for this extractor. One of these countries will be used by | |
340 | geo restriction bypass mechanism right away in order to bypass | |
341 | geo restriction, of course, if the mechanism is not disabled. (experimental) | |
773f291d | 342 | |
3ccdde8c S |
343 | NB: both these geo attributes are experimental and may change in future |
344 | or be completely removed. | |
345 | ||
d6983cb4 PH |
346 | Finally, the _WORKING attribute should be set to False for broken IEs |
347 | in order to warn the users and skip the tests. | |
348 | """ | |
349 | ||
350 | _ready = False | |
351 | _downloader = None | |
773f291d | 352 | _x_forwarded_for_ip = None |
4248dad9 S |
353 | _GEO_BYPASS = True |
354 | _GEO_COUNTRIES = None | |
d6983cb4 PH |
355 | _WORKING = True |
356 | ||
357 | def __init__(self, downloader=None): | |
358 | """Constructor. Receives an optional downloader.""" | |
359 | self._ready = False | |
773f291d | 360 | self._x_forwarded_for_ip = None |
d6983cb4 PH |
361 | self.set_downloader(downloader) |
362 | ||
363 | @classmethod | |
364 | def suitable(cls, url): | |
365 | """Receives a URL and returns True if suitable for this IE.""" | |
79cb2577 PH |
366 | |
367 | # This does not use has/getattr intentionally - we want to know whether | |
368 | # we have cached the regexp for *this* class, whereas getattr would also | |
369 | # match the superclass | |
370 | if '_VALID_URL_RE' not in cls.__dict__: | |
371 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) | |
372 | return cls._VALID_URL_RE.match(url) is not None | |
d6983cb4 | 373 | |
ed9266db PH |
374 | @classmethod |
375 | def _match_id(cls, url): | |
376 | if '_VALID_URL_RE' not in cls.__dict__: | |
377 | cls._VALID_URL_RE = re.compile(cls._VALID_URL) | |
378 | m = cls._VALID_URL_RE.match(url) | |
379 | assert m | |
1afd0b0d | 380 | return compat_str(m.group('id')) |
ed9266db | 381 | |
d6983cb4 PH |
382 | @classmethod |
383 | def working(cls): | |
384 | """Getter method for _WORKING.""" | |
385 | return cls._WORKING | |
386 | ||
387 | def initialize(self): | |
388 | """Initializes an instance (authentication, etc).""" | |
e39b5d4a | 389 | self._initialize_geo_bypass(self._GEO_COUNTRIES) |
4248dad9 S |
390 | if not self._ready: |
391 | self._real_initialize() | |
392 | self._ready = True | |
393 | ||
e39b5d4a S |
394 | def _initialize_geo_bypass(self, countries): |
395 | """ | |
396 | Initialize geo restriction bypass mechanism. | |
397 | ||
398 | This method is used to initialize geo bypass mechanism based on faking | |
399 | X-Forwarded-For HTTP header. A random country from provided country list | |
dc0a869e | 400 | is selected and a random IP belonging to this country is generated. This |
e39b5d4a S |
401 | IP will be passed as X-Forwarded-For HTTP header in all subsequent |
402 | HTTP requests. | |
e39b5d4a S |
403 | |
404 | This method will be used for initial geo bypass mechanism initialization | |
405 | during the instance initialization with _GEO_COUNTRIES. | |
406 | ||
407 | You may also manually call it from extractor's code if geo countries | |
408 | information is not available beforehand (e.g. obtained during | |
409 | extraction) or due to some another reason. | |
410 | """ | |
773f291d | 411 | if not self._x_forwarded_for_ip: |
0a840f58 | 412 | country_code = self._downloader.params.get('geo_bypass_country', None) |
4248dad9 S |
413 | # If there is no explicit country for geo bypass specified and |
414 | # the extractor is known to be geo restricted let's fake IP | |
415 | # as X-Forwarded-For right away. | |
416 | if (not country_code and | |
417 | self._GEO_BYPASS and | |
418 | self._downloader.params.get('geo_bypass', True) and | |
e39b5d4a S |
419 | countries): |
420 | country_code = random.choice(countries) | |
773f291d S |
421 | if country_code: |
422 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) | |
4248dad9 | 423 | if self._downloader.params.get('verbose', False): |
6a9cb295 | 424 | self._downloader.to_screen( |
eea0716c S |
425 | '[debug] Using fake IP %s (%s) as X-Forwarded-For.' |
426 | % (self._x_forwarded_for_ip, country_code.upper())) | |
d6983cb4 PH |
427 | |
428 | def extract(self, url): | |
429 | """Extracts URL information and returns it in list of dicts.""" | |
3a5bcd03 | 430 | try: |
773f291d S |
431 | for _ in range(2): |
432 | try: | |
433 | self.initialize() | |
0016b84e S |
434 | ie_result = self._real_extract(url) |
435 | if self._x_forwarded_for_ip: | |
436 | ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip | |
437 | return ie_result | |
773f291d | 438 | except GeoRestrictedError as e: |
4248dad9 S |
439 | if self.__maybe_fake_ip_and_retry(e.countries): |
440 | continue | |
773f291d | 441 | raise |
3a5bcd03 PH |
442 | except ExtractorError: |
443 | raise | |
444 | except compat_http_client.IncompleteRead as e: | |
dfb1b146 | 445 | raise ExtractorError('A network error has occurred.', cause=e, expected=True) |
9650885b | 446 | except (KeyError, StopIteration) as e: |
dfb1b146 | 447 | raise ExtractorError('An extractor error has occurred.', cause=e) |
d6983cb4 | 448 | |
4248dad9 S |
449 | def __maybe_fake_ip_and_retry(self, countries): |
450 | if (not self._downloader.params.get('geo_bypass_country', None) and | |
451 | self._GEO_BYPASS and | |
452 | self._downloader.params.get('geo_bypass', True) and | |
453 | not self._x_forwarded_for_ip and | |
454 | countries): | |
eea0716c S |
455 | country_code = random.choice(countries) |
456 | self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) | |
4248dad9 S |
457 | if self._x_forwarded_for_ip: |
458 | self.report_warning( | |
eea0716c S |
459 | 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.' |
460 | % (self._x_forwarded_for_ip, country_code.upper())) | |
4248dad9 S |
461 | return True |
462 | return False | |
463 | ||
d6983cb4 PH |
464 | def set_downloader(self, downloader): |
465 | """Sets the downloader for this IE.""" | |
466 | self._downloader = downloader | |
467 | ||
468 | def _real_initialize(self): | |
469 | """Real initialization process. Redefine in subclasses.""" | |
470 | pass | |
471 | ||
472 | def _real_extract(self, url): | |
473 | """Real extraction process. Redefine in subclasses.""" | |
474 | pass | |
475 | ||
56c73665 JMF |
476 | @classmethod |
477 | def ie_key(cls): | |
478 | """A string for getting the InfoExtractor with get_info_extractor""" | |
dc519b54 | 479 | return compat_str(cls.__name__[:-2]) |
56c73665 | 480 | |
d6983cb4 PH |
481 | @property |
482 | def IE_NAME(self): | |
dc519b54 | 483 | return compat_str(type(self).__name__[:-2]) |
d6983cb4 | 484 | |
41d06b04 | 485 | def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers={}, query={}): |
d6983cb4 PH |
486 | """ Returns the response handle """ |
487 | if note is None: | |
488 | self.report_download_webpage(video_id) | |
489 | elif note is not False: | |
7cc3570e | 490 | if video_id is None: |
f1a9d64e | 491 | self.to_screen('%s' % (note,)) |
7cc3570e | 492 | else: |
f1a9d64e | 493 | self.to_screen('%s: %s' % (video_id, note)) |
41d06b04 S |
494 | if isinstance(url_or_request, compat_urllib_request.Request): |
495 | url_or_request = update_Request( | |
496 | url_or_request, data=data, headers=headers, query=query) | |
497 | else: | |
cdfee168 | 498 | if query: |
499 | url_or_request = update_url_query(url_or_request, query) | |
2c0d9c62 | 500 | if data is not None or headers: |
41d06b04 | 501 | url_or_request = sanitized_Request(url_or_request, data, headers) |
d6983cb4 | 502 | try: |
dca08720 | 503 | return self._downloader.urlopen(url_or_request) |
d6983cb4 | 504 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: |
aa94a6d3 PH |
505 | if errnote is False: |
506 | return False | |
d6983cb4 | 507 | if errnote is None: |
f1a9d64e | 508 | errnote = 'Unable to download webpage' |
7f8b2714 | 509 | |
9b9c5355 | 510 | errmsg = '%s: %s' % (errnote, error_to_compat_str(err)) |
7cc3570e PH |
511 | if fatal: |
512 | raise ExtractorError(errmsg, sys.exc_info()[2], cause=err) | |
513 | else: | |
514 | self._downloader.report_warning(errmsg) | |
515 | return False | |
d6983cb4 | 516 | |
41d06b04 | 517 | def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}): |
d6983cb4 | 518 | """ Returns a tuple (page content as string, URL handle) """ |
b9d3e163 PH |
519 | # Strip hashes from the URL (#1038) |
520 | if isinstance(url_or_request, (compat_str, str)): | |
521 | url_or_request = url_or_request.partition('#')[0] | |
522 | ||
773f291d S |
523 | # Some sites check X-Forwarded-For HTTP header in order to figure out |
524 | # the origin of the client behind proxy. This allows bypassing geo | |
525 | # restriction by faking this header's value to IP that belongs to some | |
526 | # geo unrestricted country. We will do so once we encounter any | |
527 | # geo restriction error. | |
528 | if self._x_forwarded_for_ip: | |
529 | if 'X-Forwarded-For' not in headers: | |
530 | headers['X-Forwarded-For'] = self._x_forwarded_for_ip | |
531 | ||
cdfee168 | 532 | urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query) |
7cc3570e PH |
533 | if urlh is False: |
534 | assert not fatal | |
535 | return False | |
c9a77969 | 536 | content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding) |
23be51d8 PH |
537 | return (content, urlh) |
538 | ||
c9a77969 YCH |
539 | @staticmethod |
540 | def _guess_encoding_from_content(content_type, webpage_bytes): | |
d6983cb4 PH |
541 | m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type) |
542 | if m: | |
543 | encoding = m.group(1) | |
544 | else: | |
0d75ae2c | 545 | m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]', |
f143d86a PH |
546 | webpage_bytes[:1024]) |
547 | if m: | |
548 | encoding = m.group(1).decode('ascii') | |
b60016e8 PH |
549 | elif webpage_bytes.startswith(b'\xff\xfe'): |
550 | encoding = 'utf-16' | |
f143d86a PH |
551 | else: |
552 | encoding = 'utf-8' | |
c9a77969 YCH |
553 | |
554 | return encoding | |
555 | ||
4457823d S |
556 | def __check_blocked(self, content): |
557 | first_block = content[:512] | |
558 | if ('<title>Access to this site is blocked</title>' in content and | |
559 | 'Websense' in first_block): | |
560 | msg = 'Access to this webpage has been blocked by Websense filtering software in your network.' | |
561 | blocked_iframe = self._html_search_regex( | |
562 | r'<iframe src="([^"]+)"', content, | |
563 | 'Websense information URL', default=None) | |
564 | if blocked_iframe: | |
565 | msg += ' Visit %s for more details' % blocked_iframe | |
566 | raise ExtractorError(msg, expected=True) | |
567 | if '<title>The URL you requested has been blocked</title>' in first_block: | |
568 | msg = ( | |
569 | 'Access to this webpage has been blocked by Indian censorship. ' | |
570 | 'Use a VPN or proxy server (with --proxy) to route around it.') | |
571 | block_msg = self._html_search_regex( | |
572 | r'</h1><p>(.*?)</p>', | |
573 | content, 'block message', default=None) | |
574 | if block_msg: | |
575 | msg += ' (Message: "%s")' % block_msg.replace('\n', ' ') | |
576 | raise ExtractorError(msg, expected=True) | |
577 | if ('<title>TTK :: Доступ к ресурсу ограничен</title>' in content and | |
578 | 'blocklist.rkn.gov.ru' in content): | |
579 | raise ExtractorError( | |
580 | 'Access to this webpage has been blocked by decision of the Russian government. ' | |
581 | 'Visit http://blocklist.rkn.gov.ru/ for a block reason.', | |
582 | expected=True) | |
583 | ||
c9a77969 YCH |
584 | def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None): |
585 | content_type = urlh.headers.get('Content-Type', '') | |
586 | webpage_bytes = urlh.read() | |
587 | if prefix is not None: | |
588 | webpage_bytes = prefix + webpage_bytes | |
589 | if not encoding: | |
590 | encoding = self._guess_encoding_from_content(content_type, webpage_bytes) | |
d6983cb4 PH |
591 | if self._downloader.params.get('dump_intermediate_pages', False): |
592 | try: | |
593 | url = url_or_request.get_full_url() | |
594 | except AttributeError: | |
595 | url = url_or_request | |
f1a9d64e | 596 | self.to_screen('Dumping request to ' + url) |
d6983cb4 PH |
597 | dump = base64.b64encode(webpage_bytes).decode('ascii') |
598 | self._downloader.to_screen(dump) | |
d41e6efc PH |
599 | if self._downloader.params.get('write_pages', False): |
600 | try: | |
601 | url = url_or_request.get_full_url() | |
602 | except AttributeError: | |
603 | url = url_or_request | |
5afa7f8b | 604 | basen = '%s_%s' % (video_id, url) |
c1bce22f | 605 | if len(basen) > 240: |
f1a9d64e | 606 | h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest() |
c1bce22f PH |
607 | basen = basen[:240 - len(h)] + h |
608 | raw_filename = basen + '.dump' | |
d41e6efc | 609 | filename = sanitize_filename(raw_filename, restricted=True) |
f1a9d64e | 610 | self.to_screen('Saving request to ' + filename) |
5f58165d S |
611 | # Working around MAX_PATH limitation on Windows (see |
612 | # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx) | |
e9c0cdd3 | 613 | if compat_os_name == 'nt': |
5f58165d S |
614 | absfilepath = os.path.abspath(filename) |
615 | if len(absfilepath) > 259: | |
616 | filename = '\\\\?\\' + absfilepath | |
d41e6efc PH |
617 | with open(filename, 'wb') as outf: |
618 | outf.write(webpage_bytes) | |
619 | ||
ec0fafbb AA |
620 | try: |
621 | content = webpage_bytes.decode(encoding, 'replace') | |
622 | except LookupError: | |
623 | content = webpage_bytes.decode('utf-8', 'replace') | |
2410c43d | 624 | |
4457823d | 625 | self.__check_blocked(content) |
2410c43d | 626 | |
23be51d8 | 627 | return content |
d6983cb4 | 628 | |
41d06b04 | 629 | def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None, data=None, headers={}, query={}): |
d6983cb4 | 630 | """ Returns the data of the page as a string """ |
995ad69c TF |
631 | success = False |
632 | try_count = 0 | |
633 | while success is False: | |
634 | try: | |
cdfee168 | 635 | res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding, data=data, headers=headers, query=query) |
995ad69c TF |
636 | success = True |
637 | except compat_http_client.IncompleteRead as e: | |
638 | try_count += 1 | |
639 | if try_count >= tries: | |
640 | raise e | |
641 | self._sleep(timeout, video_id) | |
7cc3570e PH |
642 | if res is False: |
643 | return res | |
644 | else: | |
645 | content, _ = res | |
646 | return content | |
d6983cb4 | 647 | |
2a275ab0 | 648 | def _download_xml(self, url_or_request, video_id, |
f1a9d64e | 649 | note='Downloading XML', errnote='Unable to download XML', |
e01c3d2e S |
650 | transform_source=None, fatal=True, encoding=None, |
651 | data=None, headers={}, query={}): | |
267ed0c5 | 652 | """Return the xml as an xml.etree.ElementTree.Element""" |
28746fbd | 653 | xml_string = self._download_webpage( |
e01c3d2e S |
654 | url_or_request, video_id, note, errnote, fatal=fatal, |
655 | encoding=encoding, data=data, headers=headers, query=query) | |
28746fbd PH |
656 | if xml_string is False: |
657 | return xml_string | |
e01c3d2e S |
658 | return self._parse_xml( |
659 | xml_string, video_id, transform_source=transform_source, | |
660 | fatal=fatal) | |
661 | ||
662 | def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True): | |
e2b38da9 PH |
663 | if transform_source: |
664 | xml_string = transform_source(xml_string) | |
e01c3d2e S |
665 | try: |
666 | return compat_etree_fromstring(xml_string.encode('utf-8')) | |
667 | except compat_xml_parse_error as ve: | |
668 | errmsg = '%s: Failed to parse XML ' % video_id | |
669 | if fatal: | |
670 | raise ExtractorError(errmsg, cause=ve) | |
671 | else: | |
672 | self.report_warning(errmsg + str(ve)) | |
267ed0c5 | 673 | |
3d3538e4 | 674 | def _download_json(self, url_or_request, video_id, |
f1a9d64e PH |
675 | note='Downloading JSON metadata', |
676 | errnote='Unable to download JSON metadata', | |
b090af59 | 677 | transform_source=None, |
41d06b04 | 678 | fatal=True, encoding=None, data=None, headers={}, query={}): |
b090af59 | 679 | json_string = self._download_webpage( |
c9a77969 | 680 | url_or_request, video_id, note, errnote, fatal=fatal, |
cdfee168 | 681 | encoding=encoding, data=data, headers=headers, query=query) |
b090af59 PH |
682 | if (not fatal) and json_string is False: |
683 | return None | |
ebb64199 TF |
684 | return self._parse_json( |
685 | json_string, video_id, transform_source=transform_source, fatal=fatal) | |
686 | ||
687 | def _parse_json(self, json_string, video_id, transform_source=None, fatal=True): | |
81c2f20b PH |
688 | if transform_source: |
689 | json_string = transform_source(json_string) | |
3d3538e4 PH |
690 | try: |
691 | return json.loads(json_string) | |
692 | except ValueError as ve: | |
e7b6d122 PH |
693 | errmsg = '%s: Failed to parse JSON ' % video_id |
694 | if fatal: | |
695 | raise ExtractorError(errmsg, cause=ve) | |
696 | else: | |
697 | self.report_warning(errmsg + str(ve)) | |
3d3538e4 | 698 | |
f45f96f8 | 699 | def report_warning(self, msg, video_id=None): |
f1a9d64e | 700 | idstr = '' if video_id is None else '%s: ' % video_id |
f45f96f8 | 701 | self._downloader.report_warning( |
f1a9d64e | 702 | '[%s] %s%s' % (self.IE_NAME, idstr, msg)) |
f45f96f8 | 703 | |
d6983cb4 PH |
704 | def to_screen(self, msg): |
705 | """Print msg to screen, prefixing it with '[ie_name]'""" | |
f1a9d64e | 706 | self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg)) |
d6983cb4 PH |
707 | |
708 | def report_extraction(self, id_or_name): | |
709 | """Report information extraction.""" | |
f1a9d64e | 710 | self.to_screen('%s: Extracting information' % id_or_name) |
d6983cb4 PH |
711 | |
712 | def report_download_webpage(self, video_id): | |
713 | """Report webpage download.""" | |
f1a9d64e | 714 | self.to_screen('%s: Downloading webpage' % video_id) |
d6983cb4 PH |
715 | |
716 | def report_age_confirmation(self): | |
717 | """Report attempt to confirm age.""" | |
f1a9d64e | 718 | self.to_screen('Confirming age') |
d6983cb4 | 719 | |
fc79158d JMF |
720 | def report_login(self): |
721 | """Report attempt to log in.""" | |
f1a9d64e | 722 | self.to_screen('Logging in') |
fc79158d | 723 | |
43e7d3c9 S |
724 | @staticmethod |
725 | def raise_login_required(msg='This video is only available for registered users'): | |
726 | raise ExtractorError( | |
727 | '%s. Use --username and --password or --netrc to provide account credentials.' % msg, | |
728 | expected=True) | |
729 | ||
c430802e | 730 | @staticmethod |
773f291d S |
731 | def raise_geo_restricted(msg='This video is not available from your location due to geo restriction', countries=None): |
732 | raise GeoRestrictedError(msg, countries=countries) | |
c430802e | 733 | |
5f6a1245 | 734 | # Methods for following #608 |
c0d0b01f | 735 | @staticmethod |
830d53bf | 736 | def url_result(url, ie=None, video_id=None, video_title=None): |
10952eb2 | 737 | """Returns a URL that points to a page that should be processed""" |
5f6a1245 | 738 | # TODO: ie should be the class used for getting the info |
d6983cb4 PH |
739 | video_info = {'_type': 'url', |
740 | 'url': url, | |
741 | 'ie_key': ie} | |
7012b23c PH |
742 | if video_id is not None: |
743 | video_info['id'] = video_id | |
830d53bf S |
744 | if video_title is not None: |
745 | video_info['title'] = video_title | |
d6983cb4 | 746 | return video_info |
5f6a1245 | 747 | |
749ca5ec S |
748 | def playlist_from_matches(self, matches, playlist_id=None, playlist_title=None, getter=None, ie=None): |
749 | urls = orderedSet( | |
46b18f23 JH |
750 | self.url_result(self._proto_relative_url(getter(m) if getter else m), ie) |
751 | for m in matches) | |
752 | return self.playlist_result( | |
749ca5ec | 753 | urls, playlist_id=playlist_id, playlist_title=playlist_title) |
46b18f23 | 754 | |
c0d0b01f | 755 | @staticmethod |
acf5cbfe | 756 | def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None): |
d6983cb4 PH |
757 | """Returns a playlist""" |
758 | video_info = {'_type': 'playlist', | |
759 | 'entries': entries} | |
760 | if playlist_id: | |
761 | video_info['id'] = playlist_id | |
762 | if playlist_title: | |
763 | video_info['title'] = playlist_title | |
acf5cbfe S |
764 | if playlist_description: |
765 | video_info['description'] = playlist_description | |
d6983cb4 PH |
766 | return video_info |
767 | ||
c342041f | 768 | def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
769 | """ |
770 | Perform a regex search on the given string, using a single or a list of | |
771 | patterns returning the first matching group. | |
772 | In case of failure return a default value or raise a WARNING or a | |
55b3e45b | 773 | RegexNotFoundError, depending on fatal, specifying the field name. |
d6983cb4 PH |
774 | """ |
775 | if isinstance(pattern, (str, compat_str, compiled_regex_type)): | |
776 | mobj = re.search(pattern, string, flags) | |
777 | else: | |
778 | for p in pattern: | |
779 | mobj = re.search(p, string, flags) | |
c3415d1b PH |
780 | if mobj: |
781 | break | |
d6983cb4 | 782 | |
e9c0cdd3 | 783 | if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty(): |
f1a9d64e | 784 | _name = '\033[0;34m%s\033[0m' % name |
d6983cb4 PH |
785 | else: |
786 | _name = name | |
787 | ||
788 | if mobj: | |
711ede6e PH |
789 | if group is None: |
790 | # return the first matching group | |
791 | return next(g for g in mobj.groups() if g is not None) | |
792 | else: | |
793 | return mobj.group(group) | |
c342041f | 794 | elif default is not NO_DEFAULT: |
d6983cb4 PH |
795 | return default |
796 | elif fatal: | |
f1a9d64e | 797 | raise RegexNotFoundError('Unable to extract %s' % _name) |
d6983cb4 | 798 | else: |
08f2a92c | 799 | self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) |
d6983cb4 PH |
800 | return None |
801 | ||
c342041f | 802 | def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): |
d6983cb4 PH |
803 | """ |
804 | Like _search_regex, but strips HTML tags and unescapes entities. | |
805 | """ | |
711ede6e | 806 | res = self._search_regex(pattern, string, name, default, fatal, flags, group) |
d6983cb4 PH |
807 | if res: |
808 | return clean_html(res).strip() | |
809 | else: | |
810 | return res | |
811 | ||
2118fdd1 RA |
812 | def _get_netrc_login_info(self, netrc_machine=None): |
813 | username = None | |
814 | password = None | |
815 | netrc_machine = netrc_machine or self._NETRC_MACHINE | |
816 | ||
817 | if self._downloader.params.get('usenetrc', False): | |
818 | try: | |
819 | info = netrc.netrc().authenticators(netrc_machine) | |
820 | if info is not None: | |
821 | username = info[0] | |
822 | password = info[2] | |
823 | else: | |
dcce092e S |
824 | raise netrc.NetrcParseError( |
825 | 'No authenticators for %s' % netrc_machine) | |
2118fdd1 | 826 | except (IOError, netrc.NetrcParseError) as err: |
dcce092e S |
827 | self._downloader.report_warning( |
828 | 'parsing .netrc: %s' % error_to_compat_str(err)) | |
2118fdd1 | 829 | |
dcce092e | 830 | return username, password |
2118fdd1 | 831 | |
1b6712ab | 832 | def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None): |
fc79158d | 833 | """ |
cf0649f8 | 834 | Get the login info as (username, password) |
32443dd3 S |
835 | First look for the manually specified credentials using username_option |
836 | and password_option as keys in params dictionary. If no such credentials | |
837 | available look in the netrc file using the netrc_machine or _NETRC_MACHINE | |
838 | value. | |
fc79158d JMF |
839 | If there's no info available, return (None, None) |
840 | """ | |
841 | if self._downloader is None: | |
842 | return (None, None) | |
843 | ||
fc79158d JMF |
844 | downloader_params = self._downloader.params |
845 | ||
846 | # Attempt to use provided username and password or .netrc data | |
1b6712ab RA |
847 | if downloader_params.get(username_option) is not None: |
848 | username = downloader_params[username_option] | |
849 | password = downloader_params[password_option] | |
2118fdd1 | 850 | else: |
1b6712ab | 851 | username, password = self._get_netrc_login_info(netrc_machine) |
5f6a1245 | 852 | |
2133565c | 853 | return username, password |
fc79158d | 854 | |
e64b7569 | 855 | def _get_tfa_info(self, note='two-factor verification code'): |
83317f69 | 856 | """ |
857 | Get the two-factor authentication info | |
858 | TODO - asking the user will be required for sms/phone verify | |
859 | currently just uses the command line option | |
860 | If there's no info available, return None | |
861 | """ | |
862 | if self._downloader is None: | |
83317f69 | 863 | return None |
864 | downloader_params = self._downloader.params | |
865 | ||
d800609c | 866 | if downloader_params.get('twofactor') is not None: |
83317f69 | 867 | return downloader_params['twofactor'] |
868 | ||
e64b7569 | 869 | return compat_getpass('Type %s and press [Return]: ' % note) |
83317f69 | 870 | |
46720279 JMF |
871 | # Helper functions for extracting OpenGraph info |
872 | @staticmethod | |
ab2d5247 | 873 | def _og_regexes(prop): |
448ef1f3 | 874 | content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))' |
7a6d76a6 S |
875 | property_re = (r'(?:name|property)=(?:\'og:%(prop)s\'|"og:%(prop)s"|\s*og:%(prop)s\b)' |
876 | % {'prop': re.escape(prop)}) | |
78fb87b2 | 877 | template = r'<meta[^>]+?%s[^>]+?%s' |
ab2d5247 | 878 | return [ |
78fb87b2 JMF |
879 | template % (property_re, content_re), |
880 | template % (content_re, property_re), | |
ab2d5247 | 881 | ] |
46720279 | 882 | |
864f24bd S |
883 | @staticmethod |
884 | def _meta_regex(prop): | |
885 | return r'''(?isx)<meta | |
8b9848ac | 886 | (?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1) |
864f24bd S |
887 | [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop) |
888 | ||
3c4e6d83 | 889 | def _og_search_property(self, prop, html, name=None, **kargs): |
b070564e S |
890 | if not isinstance(prop, (list, tuple)): |
891 | prop = [prop] | |
46720279 | 892 | if name is None: |
b070564e S |
893 | name = 'OpenGraph %s' % prop[0] |
894 | og_regexes = [] | |
895 | for p in prop: | |
896 | og_regexes.extend(self._og_regexes(p)) | |
897 | escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs) | |
eb0a8398 PH |
898 | if escaped is None: |
899 | return None | |
900 | return unescapeHTML(escaped) | |
46720279 JMF |
901 | |
902 | def _og_search_thumbnail(self, html, **kargs): | |
10952eb2 | 903 | return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs) |
46720279 JMF |
904 | |
905 | def _og_search_description(self, html, **kargs): | |
906 | return self._og_search_property('description', html, fatal=False, **kargs) | |
907 | ||
908 | def _og_search_title(self, html, **kargs): | |
909 | return self._og_search_property('title', html, **kargs) | |
910 | ||
8ffa13e0 | 911 | def _og_search_video_url(self, html, name='video url', secure=True, **kargs): |
a3681973 PH |
912 | regexes = self._og_regexes('video') + self._og_regexes('video:url') |
913 | if secure: | |
914 | regexes = self._og_regexes('video:secure_url') + regexes | |
8ffa13e0 | 915 | return self._html_search_regex(regexes, html, name, **kargs) |
46720279 | 916 | |
78338f71 JMF |
917 | def _og_search_url(self, html, **kargs): |
918 | return self._og_search_property('url', html, **kargs) | |
919 | ||
40c696e5 | 920 | def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs): |
88d9f6c0 S |
921 | if not isinstance(name, (list, tuple)): |
922 | name = [name] | |
59040888 | 923 | if display_name is None: |
88d9f6c0 | 924 | display_name = name[0] |
59040888 | 925 | return self._html_search_regex( |
88d9f6c0 | 926 | [self._meta_regex(n) for n in name], |
711ede6e | 927 | html, display_name, fatal=fatal, group='content', **kwargs) |
59040888 PH |
928 | |
929 | def _dc_search_uploader(self, html): | |
930 | return self._html_search_meta('dc.creator', html, 'uploader') | |
931 | ||
8dbe9899 PH |
932 | def _rta_search(self, html): |
933 | # See http://www.rtalabel.org/index.php?content=howtofaq#single | |
934 | if re.search(r'(?ix)<meta\s+name="rating"\s+' | |
935 | r' content="RTA-5042-1996-1400-1577-RTA"', | |
936 | html): | |
937 | return 18 | |
938 | return 0 | |
939 | ||
59040888 PH |
940 | def _media_rating_search(self, html): |
941 | # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/ | |
942 | rating = self._html_search_meta('rating', html) | |
943 | ||
944 | if not rating: | |
945 | return None | |
946 | ||
947 | RATING_TABLE = { | |
948 | 'safe for kids': 0, | |
949 | 'general': 8, | |
950 | '14 years': 14, | |
951 | 'mature': 17, | |
952 | 'restricted': 19, | |
953 | } | |
d800609c | 954 | return RATING_TABLE.get(rating.lower()) |
59040888 | 955 | |
69319969 | 956 | def _family_friendly_search(self, html): |
6ca7732d | 957 | # See http://schema.org/VideoObject |
ac8491fc S |
958 | family_friendly = self._html_search_meta( |
959 | 'isFamilyFriendly', html, default=None) | |
69319969 NJ |
960 | |
961 | if not family_friendly: | |
962 | return None | |
963 | ||
964 | RATING_TABLE = { | |
965 | '1': 0, | |
966 | 'true': 0, | |
967 | '0': 18, | |
968 | 'false': 18, | |
969 | } | |
d800609c | 970 | return RATING_TABLE.get(family_friendly.lower()) |
69319969 | 971 | |
0c708f11 JMF |
972 | def _twitter_search_player(self, html): |
973 | return self._html_search_meta('twitter:player', html, | |
9e1a5b84 | 974 | 'twitter card player') |
0c708f11 | 975 | |
95b31e26 | 976 | def _search_json_ld(self, html, video_id, expected_type=None, **kwargs): |
4ca2a3cf S |
977 | json_ld = self._search_regex( |
978 | r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>', | |
0b26ba3f | 979 | html, 'JSON-LD', group='json_ld', **kwargs) |
321b5e08 | 980 | default = kwargs.get('default', NO_DEFAULT) |
4ca2a3cf | 981 | if not json_ld: |
321b5e08 S |
982 | return default if default is not NO_DEFAULT else {} |
983 | # JSON-LD may be malformed and thus `fatal` should be respected. | |
984 | # At the same time `default` may be passed that assumes `fatal=False` | |
985 | # for _search_regex. Let's simulate the same behavior here as well. | |
986 | fatal = kwargs.get('fatal', True) if default == NO_DEFAULT else False | |
987 | return self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type) | |
4ca2a3cf | 988 | |
95b31e26 | 989 | def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): |
4ca2a3cf S |
990 | if isinstance(json_ld, compat_str): |
991 | json_ld = self._parse_json(json_ld, video_id, fatal=fatal) | |
992 | if not json_ld: | |
993 | return {} | |
994 | info = {} | |
46933a15 S |
995 | if not isinstance(json_ld, (list, tuple, dict)): |
996 | return info | |
997 | if isinstance(json_ld, dict): | |
998 | json_ld = [json_ld] | |
bae14048 S |
999 | |
1000 | def extract_video_object(e): | |
1001 | assert e['@type'] == 'VideoObject' | |
1002 | info.update({ | |
1003 | 'url': e.get('contentUrl'), | |
1004 | 'title': unescapeHTML(e.get('name')), | |
1005 | 'description': unescapeHTML(e.get('description')), | |
1006 | 'thumbnail': e.get('thumbnailUrl') or e.get('thumbnailURL'), | |
1007 | 'duration': parse_duration(e.get('duration')), | |
1008 | 'timestamp': unified_timestamp(e.get('uploadDate')), | |
1009 | 'filesize': float_or_none(e.get('contentSize')), | |
1010 | 'tbr': int_or_none(e.get('bitrate')), | |
1011 | 'width': int_or_none(e.get('width')), | |
1012 | 'height': int_or_none(e.get('height')), | |
33a81c2c | 1013 | 'view_count': int_or_none(e.get('interactionCount')), |
bae14048 S |
1014 | }) |
1015 | ||
46933a15 S |
1016 | for e in json_ld: |
1017 | if e.get('@context') == 'http://schema.org': | |
1018 | item_type = e.get('@type') | |
1019 | if expected_type is not None and expected_type != item_type: | |
1020 | return info | |
c69701c6 | 1021 | if item_type in ('TVEpisode', 'Episode'): |
46933a15 S |
1022 | info.update({ |
1023 | 'episode': unescapeHTML(e.get('name')), | |
1024 | 'episode_number': int_or_none(e.get('episodeNumber')), | |
1025 | 'description': unescapeHTML(e.get('description')), | |
1026 | }) | |
1027 | part_of_season = e.get('partOfSeason') | |
c69701c6 | 1028 | if isinstance(part_of_season, dict) and part_of_season.get('@type') in ('TVSeason', 'Season', 'CreativeWorkSeason'): |
46933a15 | 1029 | info['season_number'] = int_or_none(part_of_season.get('seasonNumber')) |
d16b3c66 | 1030 | part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries') |
c69701c6 | 1031 | if isinstance(part_of_series, dict) and part_of_series.get('@type') in ('TVSeries', 'Series', 'CreativeWorkSeries'): |
46933a15 S |
1032 | info['series'] = unescapeHTML(part_of_series.get('name')) |
1033 | elif item_type == 'Article': | |
1034 | info.update({ | |
1035 | 'timestamp': parse_iso8601(e.get('datePublished')), | |
1036 | 'title': unescapeHTML(e.get('headline')), | |
1037 | 'description': unescapeHTML(e.get('articleBody')), | |
1038 | }) | |
1039 | elif item_type == 'VideoObject': | |
bae14048 | 1040 | extract_video_object(e) |
c69701c6 S |
1041 | continue |
1042 | video = e.get('video') | |
1043 | if isinstance(video, dict) and video.get('@type') == 'VideoObject': | |
1044 | extract_video_object(video) | |
46933a15 | 1045 | break |
4ca2a3cf S |
1046 | return dict((k, v) for k, v in info.items() if v is not None) |
1047 | ||
27713812 | 1048 | @staticmethod |
f8da79f8 | 1049 | def _hidden_inputs(html): |
586f1cc5 | 1050 | html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html) |
201ea3ee | 1051 | hidden_inputs = {} |
c8498368 S |
1052 | for input in re.findall(r'(?i)(<input[^>]+>)', html): |
1053 | attrs = extract_attributes(input) | |
1054 | if not input: | |
201ea3ee | 1055 | continue |
c8498368 | 1056 | if attrs.get('type') not in ('hidden', 'submit'): |
201ea3ee | 1057 | continue |
c8498368 S |
1058 | name = attrs.get('name') or attrs.get('id') |
1059 | value = attrs.get('value') | |
1060 | if name and value is not None: | |
1061 | hidden_inputs[name] = value | |
201ea3ee | 1062 | return hidden_inputs |
27713812 | 1063 | |
cf61d96d S |
1064 | def _form_hidden_inputs(self, form_id, html): |
1065 | form = self._search_regex( | |
73eb13df | 1066 | r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id, |
cf61d96d S |
1067 | html, '%s form' % form_id, group='form') |
1068 | return self._hidden_inputs(form) | |
1069 | ||
3ded7bac | 1070 | def _sort_formats(self, formats, field_preference=None): |
7e8caf30 | 1071 | if not formats: |
f1a9d64e | 1072 | raise ExtractorError('No video formats found') |
7e8caf30 | 1073 | |
b0d21ded S |
1074 | for f in formats: |
1075 | # Automatically determine tbr when missing based on abr and vbr (improves | |
1076 | # formats sorting in some cases) | |
350cf045 | 1077 | if 'tbr' not in f and f.get('abr') is not None and f.get('vbr') is not None: |
b0d21ded S |
1078 | f['tbr'] = f['abr'] + f['vbr'] |
1079 | ||
4bcc7bd1 | 1080 | def _formats_key(f): |
e6812ac9 PH |
1081 | # TODO remove the following workaround |
1082 | from ..utils import determine_ext | |
1083 | if not f.get('ext') and 'url' in f: | |
1084 | f['ext'] = determine_ext(f['url']) | |
1085 | ||
3ded7bac | 1086 | if isinstance(field_preference, (list, tuple)): |
bf8dd790 S |
1087 | return tuple( |
1088 | f.get(field) | |
1089 | if f.get(field) is not None | |
1090 | else ('' if field == 'format_id' else -1) | |
1091 | for field in field_preference) | |
3ded7bac | 1092 | |
4bcc7bd1 PH |
1093 | preference = f.get('preference') |
1094 | if preference is None: | |
d497a201 | 1095 | preference = 0 |
4bcc7bd1 PH |
1096 | if f.get('ext') in ['f4f', 'f4m']: # Not yet supported |
1097 | preference -= 0.5 | |
1098 | ||
8b408545 RA |
1099 | protocol = f.get('protocol') or determine_protocol(f) |
1100 | proto_preference = 0 if protocol in ['http', 'https'] else (-0.5 if protocol == 'rtsp' else -0.1) | |
d497a201 | 1101 | |
4bcc7bd1 | 1102 | if f.get('vcodec') == 'none': # audio only |
dd867805 | 1103 | preference -= 50 |
4bcc7bd1 | 1104 | if self._downloader.params.get('prefer_free_formats'): |
f1a9d64e | 1105 | ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus'] |
4bcc7bd1 | 1106 | else: |
f1a9d64e | 1107 | ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a'] |
4bcc7bd1 PH |
1108 | ext_preference = 0 |
1109 | try: | |
1110 | audio_ext_preference = ORDER.index(f['ext']) | |
1111 | except ValueError: | |
1112 | audio_ext_preference = -1 | |
1113 | else: | |
dd867805 | 1114 | if f.get('acodec') == 'none': # video only |
1115 | preference -= 40 | |
4bcc7bd1 | 1116 | if self._downloader.params.get('prefer_free_formats'): |
f1a9d64e | 1117 | ORDER = ['flv', 'mp4', 'webm'] |
4bcc7bd1 | 1118 | else: |
f1a9d64e | 1119 | ORDER = ['webm', 'flv', 'mp4'] |
4bcc7bd1 PH |
1120 | try: |
1121 | ext_preference = ORDER.index(f['ext']) | |
1122 | except ValueError: | |
1123 | ext_preference = -1 | |
1124 | audio_ext_preference = 0 | |
1125 | ||
1126 | return ( | |
1127 | preference, | |
aff2f4f4 | 1128 | f.get('language_preference') if f.get('language_preference') is not None else -1, |
5d73273f | 1129 | f.get('quality') if f.get('quality') is not None else -1, |
9933b574 | 1130 | f.get('tbr') if f.get('tbr') is not None else -1, |
03cd72b0 | 1131 | f.get('filesize') if f.get('filesize') is not None else -1, |
4bcc7bd1 | 1132 | f.get('vbr') if f.get('vbr') is not None else -1, |
1a6373ef PH |
1133 | f.get('height') if f.get('height') is not None else -1, |
1134 | f.get('width') if f.get('width') is not None else -1, | |
d497a201 | 1135 | proto_preference, |
1e1896f2 | 1136 | ext_preference, |
4bcc7bd1 PH |
1137 | f.get('abr') if f.get('abr') is not None else -1, |
1138 | audio_ext_preference, | |
2c8e03d9 | 1139 | f.get('fps') if f.get('fps') is not None else -1, |
9732d77e | 1140 | f.get('filesize_approx') if f.get('filesize_approx') is not None else -1, |
c64ed2a3 | 1141 | f.get('source_preference') if f.get('source_preference') is not None else -1, |
74f72824 | 1142 | f.get('format_id') if f.get('format_id') is not None else '', |
4bcc7bd1 PH |
1143 | ) |
1144 | formats.sort(key=_formats_key) | |
59040888 | 1145 | |
96a53167 S |
1146 | def _check_formats(self, formats, video_id): |
1147 | if formats: | |
1148 | formats[:] = filter( | |
1149 | lambda f: self._is_valid_url( | |
1150 | f['url'], video_id, | |
1151 | item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'), | |
1152 | formats) | |
1153 | ||
f5bdb444 S |
1154 | @staticmethod |
1155 | def _remove_duplicate_formats(formats): | |
1156 | format_urls = set() | |
1157 | unique_formats = [] | |
1158 | for f in formats: | |
1159 | if f['url'] not in format_urls: | |
1160 | format_urls.add(f['url']) | |
1161 | unique_formats.append(f) | |
1162 | formats[:] = unique_formats | |
1163 | ||
45024183 | 1164 | def _is_valid_url(self, url, video_id, item='video', headers={}): |
2f0f6578 S |
1165 | url = self._proto_relative_url(url, scheme='http:') |
1166 | # For now assume non HTTP(S) URLs always valid | |
1167 | if not (url.startswith('http://') or url.startswith('https://')): | |
1168 | return True | |
96a53167 | 1169 | try: |
45024183 | 1170 | self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers) |
96a53167 S |
1171 | return True |
1172 | except ExtractorError as e: | |
943a1e24 | 1173 | if isinstance(e.cause, compat_urllib_error.URLError): |
baa43cba S |
1174 | self.to_screen( |
1175 | '%s: %s URL is invalid, skipping' % (video_id, item)) | |
96a53167 S |
1176 | return False |
1177 | raise | |
1178 | ||
20991253 | 1179 | def http_scheme(self): |
1ede5b24 | 1180 | """ Either "http:" or "https:", depending on the user's preferences """ |
20991253 PH |
1181 | return ( |
1182 | 'http:' | |
1183 | if self._downloader.params.get('prefer_insecure', False) | |
1184 | else 'https:') | |
1185 | ||
57c7411f PH |
1186 | def _proto_relative_url(self, url, scheme=None): |
1187 | if url is None: | |
1188 | return url | |
1189 | if url.startswith('//'): | |
1190 | if scheme is None: | |
1191 | scheme = self.http_scheme() | |
1192 | return scheme + url | |
1193 | else: | |
1194 | return url | |
1195 | ||
4094b6e3 PH |
1196 | def _sleep(self, timeout, video_id, msg_template=None): |
1197 | if msg_template is None: | |
f1a9d64e | 1198 | msg_template = '%(video_id)s: Waiting for %(timeout)s seconds' |
4094b6e3 PH |
1199 | msg = msg_template % {'video_id': video_id, 'timeout': timeout} |
1200 | self.to_screen(msg) | |
1201 | time.sleep(timeout) | |
1202 | ||
a38436e8 | 1203 | def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None, |
4de61310 | 1204 | transform_source=lambda s: fix_xml_ampersands(s).strip(), |
448bb5f3 | 1205 | fatal=True, m3u8_id=None): |
f036a632 JMF |
1206 | manifest = self._download_xml( |
1207 | manifest_url, video_id, 'Downloading f4m manifest', | |
97f4aecf S |
1208 | 'Unable to download f4m manifest', |
1209 | # Some manifests may be malformed, e.g. prosiebensat1 generated manifests | |
1210 | # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244) | |
4de61310 S |
1211 | transform_source=transform_source, |
1212 | fatal=fatal) | |
1213 | ||
1214 | if manifest is False: | |
8d29e47f | 1215 | return [] |
31bb8d3f | 1216 | |
0fdbb332 S |
1217 | return self._parse_f4m_formats( |
1218 | manifest, manifest_url, video_id, preference=preference, f4m_id=f4m_id, | |
448bb5f3 | 1219 | transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id) |
0fdbb332 S |
1220 | |
1221 | def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, f4m_id=None, | |
1222 | transform_source=lambda s: fix_xml_ampersands(s).strip(), | |
448bb5f3 | 1223 | fatal=True, m3u8_id=None): |
fb72ec58 | 1224 | # currently youtube-dl cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy |
1225 | akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0') | |
1226 | if akamai_pv is not None and ';' in akamai_pv.text: | |
1227 | playerVerificationChallenge = akamai_pv.text.split(';')[0] | |
1228 | if playerVerificationChallenge.strip() != '': | |
1229 | return [] | |
1230 | ||
31bb8d3f | 1231 | formats = [] |
7a47d07c | 1232 | manifest_version = '1.0' |
b2527359 | 1233 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media') |
34e48bed | 1234 | if not media_nodes: |
7a47d07c | 1235 | manifest_version = '2.0' |
34e48bed | 1236 | media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') |
b22ca762 S |
1237 | # Remove unsupported DRM protected media from final formats |
1238 | # rendition (see https://github.com/rg3/youtube-dl/issues/8573). | |
1239 | media_nodes = remove_encrypted_media(media_nodes) | |
1240 | if not media_nodes: | |
1241 | return formats | |
019839fa S |
1242 | base_url = xpath_text( |
1243 | manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'], | |
1244 | 'base URL', default=None) | |
1245 | if base_url: | |
1246 | base_url = base_url.strip() | |
0a5685b2 | 1247 | |
a6571f10 | 1248 | bootstrap_info = xpath_element( |
0a5685b2 YCH |
1249 | manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'], |
1250 | 'bootstrap info', default=None) | |
1251 | ||
edd6074c RA |
1252 | vcodec = None |
1253 | mime_type = xpath_text( | |
1254 | manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'], | |
1255 | 'base URL', default=None) | |
1256 | if mime_type and mime_type.startswith('audio/'): | |
1257 | vcodec = 'none' | |
1258 | ||
b2527359 | 1259 | for i, media_el in enumerate(media_nodes): |
77b8b4e6 S |
1260 | tbr = int_or_none(media_el.attrib.get('bitrate')) |
1261 | width = int_or_none(media_el.attrib.get('width')) | |
1262 | height = int_or_none(media_el.attrib.get('height')) | |
1263 | format_id = '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])) | |
448bb5f3 YCH |
1264 | # If <bootstrapInfo> is present, the specified f4m is a |
1265 | # stream-level manifest, and only set-level manifests may refer to | |
1266 | # external resources. See section 11.4 and section 4 of F4M spec | |
1267 | if bootstrap_info is None: | |
1268 | media_url = None | |
1269 | # @href is introduced in 2.0, see section 11.6 of F4M spec | |
1270 | if manifest_version == '2.0': | |
1271 | media_url = media_el.attrib.get('href') | |
1272 | if media_url is None: | |
1273 | media_url = media_el.attrib.get('url') | |
31c746e5 S |
1274 | if not media_url: |
1275 | continue | |
cc357c4d S |
1276 | manifest_url = ( |
1277 | media_url if media_url.startswith('http://') or media_url.startswith('https://') | |
019839fa | 1278 | else ((base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url)) |
70f0f5a8 S |
1279 | # If media_url is itself a f4m manifest do the recursive extraction |
1280 | # since bitrates in parent manifest (this one) and media_url manifest | |
1281 | # may differ leading to inability to resolve the format by requested | |
1282 | # bitrate in f4m downloader | |
240b6045 YCH |
1283 | ext = determine_ext(manifest_url) |
1284 | if ext == 'f4m': | |
77b8b4e6 | 1285 | f4m_formats = self._extract_f4m_formats( |
0fdbb332 | 1286 | manifest_url, video_id, preference=preference, f4m_id=f4m_id, |
77b8b4e6 S |
1287 | transform_source=transform_source, fatal=fatal) |
1288 | # Sometimes stream-level manifest contains single media entry that | |
1289 | # does not contain any quality metadata (e.g. http://matchtv.ru/#live-player). | |
1290 | # At the same time parent's media entry in set-level manifest may | |
1291 | # contain it. We will copy it from parent in such cases. | |
1292 | if len(f4m_formats) == 1: | |
1293 | f = f4m_formats[0] | |
1294 | f.update({ | |
1295 | 'tbr': f.get('tbr') or tbr, | |
1296 | 'width': f.get('width') or width, | |
1297 | 'height': f.get('height') or height, | |
1298 | 'format_id': f.get('format_id') if not tbr else format_id, | |
edd6074c | 1299 | 'vcodec': vcodec, |
77b8b4e6 S |
1300 | }) |
1301 | formats.extend(f4m_formats) | |
70f0f5a8 | 1302 | continue |
240b6045 YCH |
1303 | elif ext == 'm3u8': |
1304 | formats.extend(self._extract_m3u8_formats( | |
1305 | manifest_url, video_id, 'mp4', preference=preference, | |
fac2af3c | 1306 | m3u8_id=m3u8_id, fatal=fatal)) |
240b6045 | 1307 | continue |
31bb8d3f | 1308 | formats.append({ |
77b8b4e6 | 1309 | 'format_id': format_id, |
31bb8d3f | 1310 | 'url': manifest_url, |
30d0b549 | 1311 | 'manifest_url': manifest_url, |
a6571f10 | 1312 | 'ext': 'flv' if bootstrap_info is not None else None, |
b2527359 | 1313 | 'tbr': tbr, |
77b8b4e6 S |
1314 | 'width': width, |
1315 | 'height': height, | |
edd6074c | 1316 | 'vcodec': vcodec, |
60ca389c | 1317 | 'preference': preference, |
31bb8d3f | 1318 | }) |
31bb8d3f JMF |
1319 | return formats |
1320 | ||
16da9bbc YCH |
1321 | def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, m3u8_id=None): |
1322 | return { | |
f207019c | 1323 | 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])), |
704df56d PH |
1324 | 'url': m3u8_url, |
1325 | 'ext': ext, | |
1326 | 'protocol': 'm3u8', | |
37768f92 | 1327 | 'preference': preference - 100 if preference else -100, |
704df56d PH |
1328 | 'resolution': 'multiple', |
1329 | 'format_note': 'Quality selection URL', | |
16da9bbc YCH |
1330 | } |
1331 | ||
1332 | def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None, | |
1333 | entry_protocol='m3u8', preference=None, | |
1334 | m3u8_id=None, note=None, errnote=None, | |
1335 | fatal=True, live=False): | |
dbd82a1d | 1336 | res = self._download_webpage_handle( |
81515ad9 | 1337 | m3u8_url, video_id, |
621ed9f5 | 1338 | note=note or 'Downloading m3u8 information', |
13af92fd YCH |
1339 | errnote=errnote or 'Failed to download m3u8 information', |
1340 | fatal=fatal) | |
cb252080 | 1341 | |
dbd82a1d | 1342 | if res is False: |
8d29e47f | 1343 | return [] |
cb252080 | 1344 | |
dbd82a1d | 1345 | m3u8_doc, urlh = res |
37113045 | 1346 | m3u8_url = urlh.geturl() |
9cdffeeb | 1347 | |
cb252080 S |
1348 | return self._parse_m3u8_formats( |
1349 | m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol, | |
1350 | preference=preference, m3u8_id=m3u8_id, live=live) | |
1351 | ||
1352 | def _parse_m3u8_formats(self, m3u8_doc, m3u8_url, ext=None, | |
1353 | entry_protocol='m3u8', preference=None, | |
1354 | m3u8_id=None, live=False): | |
08a00eef RA |
1355 | if '#EXT-X-FAXS-CM:' in m3u8_doc: # Adobe Flash Access |
1356 | return [] | |
1357 | ||
ff99fe52 | 1358 | formats = [] |
0def7587 RA |
1359 | |
1360 | format_url = lambda u: ( | |
1361 | u | |
1362 | if re.match(r'^https?://', u) | |
1363 | else compat_urlparse.urljoin(m3u8_url, u)) | |
1364 | ||
cb252080 S |
1365 | # References: |
1366 | # 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21 | |
1367 | # 2. https://github.com/rg3/youtube-dl/issues/12211 | |
1368 | ||
1369 | # We should try extracting formats only from master playlists [1, 4.3.4], | |
1370 | # i.e. playlists that describe available qualities. On the other hand | |
1371 | # media playlists [1, 4.3.3] should be returned as is since they contain | |
1372 | # just the media without qualities renditions. | |
9cdffeeb | 1373 | # Fortunately, master playlist can be easily distinguished from media |
cb252080 S |
1374 | # playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4] |
1375 | # master playlist tags MUST NOT appear in a media playist and vice versa. | |
1376 | # As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every | |
1377 | # media playlist and MUST NOT appear in master playlist thus we can | |
1378 | # clearly detect media playlist with this criterion. | |
1379 | ||
9cdffeeb | 1380 | if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is |
7f32e5dc | 1381 | return [{ |
1382 | 'url': m3u8_url, | |
1383 | 'format_id': m3u8_id, | |
1384 | 'ext': ext, | |
1385 | 'protocol': entry_protocol, | |
1386 | 'preference': preference, | |
1387 | }] | |
cb252080 S |
1388 | |
1389 | groups = {} | |
1390 | last_stream_inf = {} | |
1391 | ||
1392 | def extract_media(x_media_line): | |
1393 | media = parse_m3u8_attributes(x_media_line) | |
1394 | # As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED | |
1395 | media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME') | |
1396 | if not (media_type and group_id and name): | |
1397 | return | |
1398 | groups.setdefault(group_id, []).append(media) | |
1399 | if media_type not in ('VIDEO', 'AUDIO'): | |
1400 | return | |
1401 | media_url = media.get('URI') | |
1402 | if media_url: | |
1403 | format_id = [] | |
1404 | for v in (group_id, name): | |
1405 | if v: | |
1406 | format_id.append(v) | |
1407 | f = { | |
1408 | 'format_id': '-'.join(format_id), | |
1409 | 'url': format_url(media_url), | |
c89b49f7 | 1410 | 'manifest_url': m3u8_url, |
cb252080 S |
1411 | 'language': media.get('LANGUAGE'), |
1412 | 'ext': ext, | |
1413 | 'protocol': entry_protocol, | |
1414 | 'preference': preference, | |
1415 | } | |
1416 | if media_type == 'AUDIO': | |
1417 | f['vcodec'] = 'none' | |
1418 | formats.append(f) | |
1419 | ||
1420 | def build_stream_name(): | |
1421 | # Despite specification does not mention NAME attribute for | |
3019cb0c S |
1422 | # EXT-X-STREAM-INF tag it still sometimes may be present (see [1] |
1423 | # or vidio test in TestInfoExtractor.test_parse_m3u8_formats) | |
ddd258f9 | 1424 | # 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015 |
cb252080 S |
1425 | stream_name = last_stream_inf.get('NAME') |
1426 | if stream_name: | |
1427 | return stream_name | |
1428 | # If there is no NAME in EXT-X-STREAM-INF it will be obtained | |
1429 | # from corresponding rendition group | |
1430 | stream_group_id = last_stream_inf.get('VIDEO') | |
1431 | if not stream_group_id: | |
1432 | return | |
1433 | stream_group = groups.get(stream_group_id) | |
1434 | if not stream_group: | |
1435 | return stream_group_id | |
1436 | rendition = stream_group[0] | |
1437 | return rendition.get('NAME') or stream_group_id | |
1438 | ||
704df56d PH |
1439 | for line in m3u8_doc.splitlines(): |
1440 | if line.startswith('#EXT-X-STREAM-INF:'): | |
cb252080 | 1441 | last_stream_inf = parse_m3u8_attributes(line) |
4cd95bcb | 1442 | elif line.startswith('#EXT-X-MEDIA:'): |
cb252080 | 1443 | extract_media(line) |
704df56d PH |
1444 | elif line.startswith('#') or not line.strip(): |
1445 | continue | |
1446 | else: | |
9c99bef7 S |
1447 | tbr = float_or_none( |
1448 | last_stream_inf.get('AVERAGE-BANDWIDTH') or | |
1449 | last_stream_inf.get('BANDWIDTH'), scale=1000) | |
8dc9d361 S |
1450 | format_id = [] |
1451 | if m3u8_id: | |
1452 | format_id.append(m3u8_id) | |
cb252080 | 1453 | stream_name = build_stream_name() |
b24d6336 KH |
1454 | # Bandwidth of live streams may differ over time thus making |
1455 | # format_id unpredictable. So it's better to keep provided | |
1456 | # format_id intact. | |
e9c6cdf4 | 1457 | if not live: |
ed56f260 | 1458 | format_id.append(stream_name if stream_name else '%d' % (tbr if tbr else len(formats))) |
30d0b549 | 1459 | manifest_url = format_url(line.strip()) |
704df56d | 1460 | f = { |
8dc9d361 | 1461 | 'format_id': '-'.join(format_id), |
30d0b549 | 1462 | 'url': manifest_url, |
ff99fe52 | 1463 | 'manifest_url': m3u8_url, |
704df56d PH |
1464 | 'tbr': tbr, |
1465 | 'ext': ext, | |
cb252080 | 1466 | 'fps': float_or_none(last_stream_inf.get('FRAME-RATE')), |
f0b5d6af PH |
1467 | 'protocol': entry_protocol, |
1468 | 'preference': preference, | |
704df56d | 1469 | } |
cb252080 | 1470 | resolution = last_stream_inf.get('RESOLUTION') |
704df56d | 1471 | if resolution: |
c4c9b844 S |
1472 | mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution) |
1473 | if mobj: | |
1474 | f['width'] = int(mobj.group('width')) | |
1475 | f['height'] = int(mobj.group('height')) | |
00f4764c RA |
1476 | # Unified Streaming Platform |
1477 | mobj = re.search( | |
1478 | r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url']) | |
1479 | if mobj: | |
1480 | abr, vbr = mobj.groups() | |
1481 | abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000) | |
fbb6edd2 | 1482 | f.update({ |
00f4764c RA |
1483 | 'vbr': vbr, |
1484 | 'abr': abr, | |
fbb6edd2 | 1485 | }) |
cb252080 S |
1486 | codecs = parse_codecs(last_stream_inf.get('CODECS')) |
1487 | f.update(codecs) | |
1488 | audio_group_id = last_stream_inf.get('AUDIO') | |
1489 | # As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which | |
1490 | # references a rendition group MUST have a CODECS attribute. | |
1491 | # However, this is not always respected, for example, [2] | |
1492 | # contains EXT-X-STREAM-INF tag which references AUDIO | |
1493 | # rendition group but does not have CODECS and despite | |
1494 | # referencing audio group an audio group, it represents | |
1495 | # a complete (with audio and video) format. So, for such cases | |
1496 | # we will ignore references to rendition groups and treat them | |
1497 | # as complete formats. | |
1498 | if audio_group_id and codecs and f.get('vcodec') != 'none': | |
1499 | audio_group = groups.get(audio_group_id) | |
1500 | if audio_group and audio_group[0].get('URI'): | |
1501 | # TODO: update acodec for audio only formats with | |
1502 | # the same GROUP-ID | |
1503 | f['acodec'] = 'none' | |
704df56d | 1504 | formats.append(f) |
cb252080 | 1505 | last_stream_inf = {} |
704df56d PH |
1506 | return formats |
1507 | ||
a107193e S |
1508 | @staticmethod |
1509 | def _xpath_ns(path, namespace=None): | |
1510 | if not namespace: | |
1511 | return path | |
1512 | out = [] | |
1513 | for c in path.split('/'): | |
1514 | if not c or c == '.': | |
1515 | out.append(c) | |
1516 | else: | |
1517 | out.append('{%s}%s' % (namespace, c)) | |
1518 | return '/'.join(out) | |
1519 | ||
09f572fb | 1520 | def _extract_smil_formats(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None): |
1521 | smil = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source) | |
a107193e | 1522 | |
995029a1 PH |
1523 | if smil is False: |
1524 | assert not fatal | |
1525 | return [] | |
e89a2aab | 1526 | |
17712eeb | 1527 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
1528 | |
1529 | return self._parse_smil_formats( | |
1530 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
1531 | ||
1532 | def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None): | |
1533 | smil = self._download_smil(smil_url, video_id, fatal=fatal) | |
1534 | if smil is False: | |
1535 | return {} | |
1536 | return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params) | |
1537 | ||
09f572fb | 1538 | def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None): |
a107193e S |
1539 | return self._download_xml( |
1540 | smil_url, video_id, 'Downloading SMIL file', | |
09f572fb | 1541 | 'Unable to download SMIL file', fatal=fatal, transform_source=transform_source) |
a107193e S |
1542 | |
1543 | def _parse_smil(self, smil, smil_url, video_id, f4m_params=None): | |
17712eeb | 1544 | namespace = self._parse_smil_namespace(smil) |
a107193e S |
1545 | |
1546 | formats = self._parse_smil_formats( | |
1547 | smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params) | |
1548 | subtitles = self._parse_smil_subtitles(smil, namespace=namespace) | |
1549 | ||
1550 | video_id = os.path.splitext(url_basename(smil_url))[0] | |
1551 | title = None | |
1552 | description = None | |
647eab45 | 1553 | upload_date = None |
a107193e S |
1554 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): |
1555 | name = meta.attrib.get('name') | |
1556 | content = meta.attrib.get('content') | |
1557 | if not name or not content: | |
1558 | continue | |
1559 | if not title and name == 'title': | |
1560 | title = content | |
1561 | elif not description and name in ('description', 'abstract'): | |
1562 | description = content | |
647eab45 S |
1563 | elif not upload_date and name == 'date': |
1564 | upload_date = unified_strdate(content) | |
a107193e | 1565 | |
1e5bcdec S |
1566 | thumbnails = [{ |
1567 | 'id': image.get('type'), | |
1568 | 'url': image.get('src'), | |
1569 | 'width': int_or_none(image.get('width')), | |
1570 | 'height': int_or_none(image.get('height')), | |
1571 | } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')] | |
1572 | ||
a107193e S |
1573 | return { |
1574 | 'id': video_id, | |
1575 | 'title': title or video_id, | |
1576 | 'description': description, | |
647eab45 | 1577 | 'upload_date': upload_date, |
1e5bcdec | 1578 | 'thumbnails': thumbnails, |
a107193e S |
1579 | 'formats': formats, |
1580 | 'subtitles': subtitles, | |
1581 | } | |
1582 | ||
17712eeb S |
1583 | def _parse_smil_namespace(self, smil): |
1584 | return self._search_regex( | |
1585 | r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None) | |
1586 | ||
f877c6ae | 1587 | def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None): |
a107193e S |
1588 | base = smil_url |
1589 | for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): | |
1590 | b = meta.get('base') or meta.get('httpBase') | |
1591 | if b: | |
1592 | base = b | |
1593 | break | |
e89a2aab S |
1594 | |
1595 | formats = [] | |
1596 | rtmp_count = 0 | |
a107193e | 1597 | http_count = 0 |
7f32e5dc | 1598 | m3u8_count = 0 |
a107193e | 1599 | |
81e1c4e2 | 1600 | srcs = [] |
ad96b4c8 YCH |
1601 | media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace)) |
1602 | for medium in media: | |
1603 | src = medium.get('src') | |
81e1c4e2 | 1604 | if not src or src in srcs: |
a107193e | 1605 | continue |
81e1c4e2 | 1606 | srcs.append(src) |
a107193e | 1607 | |
ad96b4c8 YCH |
1608 | bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000) |
1609 | filesize = int_or_none(medium.get('size') or medium.get('fileSize')) | |
1610 | width = int_or_none(medium.get('width')) | |
1611 | height = int_or_none(medium.get('height')) | |
1612 | proto = medium.get('proto') | |
1613 | ext = medium.get('ext') | |
a107193e | 1614 | src_ext = determine_ext(src) |
ad96b4c8 | 1615 | streamer = medium.get('streamer') or base |
a107193e S |
1616 | |
1617 | if proto == 'rtmp' or streamer.startswith('rtmp'): | |
1618 | rtmp_count += 1 | |
1619 | formats.append({ | |
1620 | 'url': streamer, | |
1621 | 'play_path': src, | |
1622 | 'ext': 'flv', | |
1623 | 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), | |
1624 | 'tbr': bitrate, | |
1625 | 'filesize': filesize, | |
1626 | 'width': width, | |
1627 | 'height': height, | |
1628 | }) | |
f877c6ae YCH |
1629 | if transform_rtmp_url: |
1630 | streamer, src = transform_rtmp_url(streamer, src) | |
1631 | formats[-1].update({ | |
1632 | 'url': streamer, | |
1633 | 'play_path': src, | |
1634 | }) | |
a107193e S |
1635 | continue |
1636 | ||
1637 | src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src) | |
c349456e | 1638 | src_url = src_url.strip() |
a107193e S |
1639 | |
1640 | if proto == 'm3u8' or src_ext == 'm3u8': | |
7f32e5dc | 1641 | m3u8_formats = self._extract_m3u8_formats( |
1642 | src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False) | |
1643 | if len(m3u8_formats) == 1: | |
1644 | m3u8_count += 1 | |
1645 | m3u8_formats[0].update({ | |
1646 | 'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate), | |
1647 | 'tbr': bitrate, | |
1648 | 'width': width, | |
1649 | 'height': height, | |
1650 | }) | |
1651 | formats.extend(m3u8_formats) | |
a107193e S |
1652 | continue |
1653 | ||
1654 | if src_ext == 'f4m': | |
1655 | f4m_url = src_url | |
1656 | if not f4m_params: | |
1657 | f4m_params = { | |
1658 | 'hdcore': '3.2.0', | |
1659 | 'plugin': 'flowplayer-3.2.0.1', | |
1660 | } | |
1661 | f4m_url += '&' if '?' in f4m_url else '?' | |
15707c7e | 1662 | f4m_url += compat_urllib_parse_urlencode(f4m_params) |
7e5edcfd | 1663 | formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False)) |
a107193e S |
1664 | continue |
1665 | ||
c78e4817 | 1666 | if src_url.startswith('http') and self._is_valid_url(src, video_id): |
a107193e S |
1667 | http_count += 1 |
1668 | formats.append({ | |
1669 | 'url': src_url, | |
1670 | 'ext': ext or src_ext or 'flv', | |
1671 | 'format_id': 'http-%d' % (bitrate or http_count), | |
1672 | 'tbr': bitrate, | |
1673 | 'filesize': filesize, | |
1674 | 'width': width, | |
1675 | 'height': height, | |
1676 | }) | |
1677 | continue | |
63757032 | 1678 | |
e89a2aab S |
1679 | return formats |
1680 | ||
ce00af87 | 1681 | def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'): |
d413095f | 1682 | urls = [] |
a107193e S |
1683 | subtitles = {} |
1684 | for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))): | |
1685 | src = textstream.get('src') | |
d413095f | 1686 | if not src or src in urls: |
a107193e | 1687 | continue |
d413095f | 1688 | urls.append(src) |
df634be2 | 1689 | ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src) |
03bc7237 | 1690 | lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang |
a107193e S |
1691 | subtitles.setdefault(lang, []).append({ |
1692 | 'url': src, | |
1693 | 'ext': ext, | |
1694 | }) | |
1695 | return subtitles | |
63757032 | 1696 | |
942acef5 S |
1697 | def _extract_xspf_playlist(self, playlist_url, playlist_id, fatal=True): |
1698 | xspf = self._download_xml( | |
8d6765cf | 1699 | playlist_url, playlist_id, 'Downloading xpsf playlist', |
942acef5 S |
1700 | 'Unable to download xspf manifest', fatal=fatal) |
1701 | if xspf is False: | |
1702 | return [] | |
1703 | return self._parse_xspf(xspf, playlist_id) | |
8d6765cf | 1704 | |
942acef5 | 1705 | def _parse_xspf(self, playlist, playlist_id): |
8d6765cf S |
1706 | NS_MAP = { |
1707 | 'xspf': 'http://xspf.org/ns/0/', | |
1708 | 's1': 'http://static.streamone.nl/player/ns/0', | |
1709 | } | |
1710 | ||
1711 | entries = [] | |
1712 | for track in playlist.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)): | |
1713 | title = xpath_text( | |
98044462 | 1714 | track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id) |
8d6765cf S |
1715 | description = xpath_text( |
1716 | track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description') | |
1717 | thumbnail = xpath_text( | |
1718 | track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail') | |
1719 | duration = float_or_none( | |
1720 | xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000) | |
1721 | ||
1722 | formats = [{ | |
1723 | 'url': location.text, | |
1724 | 'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)), | |
1725 | 'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))), | |
1726 | 'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))), | |
1727 | } for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP))] | |
1728 | self._sort_formats(formats) | |
1729 | ||
1730 | entries.append({ | |
1731 | 'id': playlist_id, | |
1732 | 'title': title, | |
1733 | 'description': description, | |
1734 | 'thumbnail': thumbnail, | |
1735 | 'duration': duration, | |
1736 | 'formats': formats, | |
1737 | }) | |
1738 | return entries | |
1739 | ||
1bac3455 | 1740 | def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, formats_dict={}): |
1741 | res = self._download_webpage_handle( | |
1742 | mpd_url, video_id, | |
1743 | note=note or 'Downloading MPD manifest', | |
1744 | errnote=errnote or 'Failed to download MPD manifest', | |
2d2fa82d | 1745 | fatal=fatal) |
1bac3455 | 1746 | if res is False: |
2d2fa82d | 1747 | return [] |
1bac3455 | 1748 | mpd, urlh = res |
02dc0a36 | 1749 | mpd_base_url = base_url(urlh.geturl()) |
1bac3455 | 1750 | |
91cb6b50 | 1751 | return self._parse_mpd_formats( |
86f4d14f S |
1752 | compat_etree_fromstring(mpd.encode('utf-8')), mpd_id, mpd_base_url, |
1753 | formats_dict=formats_dict, mpd_url=mpd_url) | |
2d2fa82d | 1754 | |
86f4d14f | 1755 | def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', formats_dict={}, mpd_url=None): |
f0948348 S |
1756 | """ |
1757 | Parse formats from MPD manifest. | |
1758 | References: | |
1759 | 1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E), | |
1760 | http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip | |
1761 | 2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP | |
1762 | """ | |
1bac3455 | 1763 | if mpd_doc.get('type') == 'dynamic': |
1764 | return [] | |
2d2fa82d | 1765 | |
91cb6b50 | 1766 | namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None) |
f14be228 | 1767 | |
1768 | def _add_ns(path): | |
1769 | return self._xpath_ns(path, namespace) | |
1770 | ||
675d0016 | 1771 | def is_drm_protected(element): |
1772 | return element.find(_add_ns('ContentProtection')) is not None | |
1773 | ||
1bac3455 | 1774 | def extract_multisegment_info(element, ms_parent_info): |
1775 | ms_info = ms_parent_info.copy() | |
b4c1d6e8 S |
1776 | |
1777 | # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some | |
1778 | # common attributes and elements. We will only extract relevant | |
1779 | # for us. | |
1780 | def extract_common(source): | |
1781 | segment_timeline = source.find(_add_ns('SegmentTimeline')) | |
1782 | if segment_timeline is not None: | |
1783 | s_e = segment_timeline.findall(_add_ns('S')) | |
1784 | if s_e: | |
1785 | ms_info['total_number'] = 0 | |
1786 | ms_info['s'] = [] | |
1787 | for s in s_e: | |
1788 | r = int(s.get('r', 0)) | |
1789 | ms_info['total_number'] += 1 + r | |
1790 | ms_info['s'].append({ | |
1791 | 't': int(s.get('t', 0)), | |
1792 | # @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60]) | |
1793 | 'd': int(s.attrib['d']), | |
1794 | 'r': r, | |
1795 | }) | |
1796 | start_number = source.get('startNumber') | |
1797 | if start_number: | |
1798 | ms_info['start_number'] = int(start_number) | |
1799 | timescale = source.get('timescale') | |
1800 | if timescale: | |
1801 | ms_info['timescale'] = int(timescale) | |
1802 | segment_duration = source.get('duration') | |
1803 | if segment_duration: | |
48504785 | 1804 | ms_info['segment_duration'] = float(segment_duration) |
b4c1d6e8 S |
1805 | |
1806 | def extract_Initialization(source): | |
1807 | initialization = source.find(_add_ns('Initialization')) | |
1808 | if initialization is not None: | |
1809 | ms_info['initialization_url'] = initialization.attrib['sourceURL'] | |
1810 | ||
f14be228 | 1811 | segment_list = element.find(_add_ns('SegmentList')) |
1bac3455 | 1812 | if segment_list is not None: |
b4c1d6e8 S |
1813 | extract_common(segment_list) |
1814 | extract_Initialization(segment_list) | |
f14be228 | 1815 | segment_urls_e = segment_list.findall(_add_ns('SegmentURL')) |
1bac3455 | 1816 | if segment_urls_e: |
1817 | ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e] | |
1bac3455 | 1818 | else: |
f14be228 | 1819 | segment_template = element.find(_add_ns('SegmentTemplate')) |
1bac3455 | 1820 | if segment_template is not None: |
b4c1d6e8 | 1821 | extract_common(segment_template) |
e228616c S |
1822 | media = segment_template.get('media') |
1823 | if media: | |
1824 | ms_info['media'] = media | |
1bac3455 | 1825 | initialization = segment_template.get('initialization') |
1826 | if initialization: | |
e228616c | 1827 | ms_info['initialization'] = initialization |
1bac3455 | 1828 | else: |
b4c1d6e8 | 1829 | extract_Initialization(segment_template) |
1bac3455 | 1830 | return ms_info |
b323e170 | 1831 | |
1bac3455 | 1832 | mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration')) |
17b598d3 | 1833 | formats = [] |
f14be228 | 1834 | for period in mpd_doc.findall(_add_ns('Period')): |
1bac3455 | 1835 | period_duration = parse_duration(period.get('duration')) or mpd_duration |
1836 | period_ms_info = extract_multisegment_info(period, { | |
1837 | 'start_number': 1, | |
1838 | 'timescale': 1, | |
1839 | }) | |
f14be228 | 1840 | for adaptation_set in period.findall(_add_ns('AdaptationSet')): |
675d0016 | 1841 | if is_drm_protected(adaptation_set): |
1842 | continue | |
1bac3455 | 1843 | adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info) |
f14be228 | 1844 | for representation in adaptation_set.findall(_add_ns('Representation')): |
675d0016 | 1845 | if is_drm_protected(representation): |
1846 | continue | |
1bac3455 | 1847 | representation_attrib = adaptation_set.attrib.copy() |
1848 | representation_attrib.update(representation.attrib) | |
f0948348 | 1849 | # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory |
a6c8b759 YCH |
1850 | mime_type = representation_attrib['mimeType'] |
1851 | content_type = mime_type.split('/')[0] | |
1bac3455 | 1852 | if content_type == 'text': |
1853 | # TODO implement WebVTT downloading | |
1854 | pass | |
40fcba5e | 1855 | elif content_type in ('video', 'audio'): |
1bac3455 | 1856 | base_url = '' |
1857 | for element in (representation, adaptation_set, period, mpd_doc): | |
f14be228 | 1858 | base_url_e = element.find(_add_ns('BaseURL')) |
1bac3455 | 1859 | if base_url_e is not None: |
1860 | base_url = base_url_e.text + base_url | |
1861 | if re.match(r'^https?://', base_url): | |
1862 | break | |
bb20526b S |
1863 | if mpd_base_url and not re.match(r'^https?://', base_url): |
1864 | if not mpd_base_url.endswith('/') and not base_url.startswith('/'): | |
1865 | mpd_base_url += '/' | |
1bac3455 | 1866 | base_url = mpd_base_url + base_url |
1867 | representation_id = representation_attrib.get('id') | |
d577c796 | 1868 | lang = representation_attrib.get('lang') |
51e9094f | 1869 | url_el = representation.find(_add_ns('BaseURL')) |
1870 | 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 | 1871 | bandwidth = int_or_none(representation_attrib.get('bandwidth')) |
1bac3455 | 1872 | f = { |
154c209e | 1873 | 'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id, |
1bac3455 | 1874 | 'url': base_url, |
86f4d14f | 1875 | 'manifest_url': mpd_url, |
a6c8b759 | 1876 | 'ext': mimetype2ext(mime_type), |
1bac3455 | 1877 | 'width': int_or_none(representation_attrib.get('width')), |
1878 | 'height': int_or_none(representation_attrib.get('height')), | |
9c99bef7 | 1879 | 'tbr': float_or_none(bandwidth, 1000), |
1bac3455 | 1880 | 'asr': int_or_none(representation_attrib.get('audioSamplingRate')), |
1881 | 'fps': int_or_none(representation_attrib.get('frameRate')), | |
d577c796 | 1882 | 'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None, |
1bac3455 | 1883 | 'format_note': 'DASH %s' % content_type, |
51e9094f | 1884 | 'filesize': filesize, |
1bac3455 | 1885 | } |
7fe15920 | 1886 | f.update(parse_codecs(representation_attrib.get('codecs'))) |
1bac3455 | 1887 | representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info) |
b4c1d6e8 | 1888 | |
e228616c S |
1889 | def prepare_template(template_name, identifiers): |
1890 | t = representation_ms_info[template_name] | |
1891 | t = t.replace('$RepresentationID$', representation_id) | |
1892 | t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t) | |
1893 | t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t) | |
1894 | t.replace('$$', '$') | |
1895 | return t | |
1896 | ||
1897 | # @initialization is a regular template like @media one | |
1898 | # so it should be handled just the same way (see | |
1899 | # https://github.com/rg3/youtube-dl/issues/11605) | |
1900 | if 'initialization' in representation_ms_info: | |
1901 | initialization_template = prepare_template( | |
1902 | 'initialization', | |
1903 | # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and | |
1904 | # $Time$ shall not be included for @initialization thus | |
1905 | # only $Bandwidth$ remains | |
1906 | ('Bandwidth', )) | |
1907 | representation_ms_info['initialization_url'] = initialization_template % { | |
1908 | 'Bandwidth': bandwidth, | |
1909 | } | |
1910 | ||
1141e910 S |
1911 | def location_key(location): |
1912 | return 'url' if re.match(r'^https?://', location) else 'path' | |
1913 | ||
e228616c S |
1914 | if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info: |
1915 | ||
1916 | media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time')) | |
1141e910 | 1917 | media_location_key = location_key(media_template) |
f0948348 S |
1918 | |
1919 | # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$ | |
1920 | # can't be used at the same time | |
b4c1d6e8 S |
1921 | if '%(Number' in media_template and 's' not in representation_ms_info: |
1922 | segment_duration = None | |
1923 | if 'total_number' not in representation_ms_info and 'segment_duration': | |
1924 | segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale']) | |
1925 | representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration)) | |
b4c1d6e8 | 1926 | representation_ms_info['fragments'] = [{ |
1141e910 | 1927 | media_location_key: media_template % { |
b4c1d6e8 | 1928 | 'Number': segment_number, |
e228616c | 1929 | 'Bandwidth': bandwidth, |
b4c1d6e8 S |
1930 | }, |
1931 | 'duration': segment_duration, | |
1932 | } for segment_number in range( | |
1933 | representation_ms_info['start_number'], | |
1934 | representation_ms_info['total_number'] + representation_ms_info['start_number'])] | |
f0948348 | 1935 | else: |
b4c1d6e8 S |
1936 | # $Number*$ or $Time$ in media template with S list available |
1937 | # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg | |
1938 | # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411 | |
b4c1d6e8 | 1939 | representation_ms_info['fragments'] = [] |
f0948348 | 1940 | segment_time = 0 |
b4c1d6e8 S |
1941 | segment_d = None |
1942 | segment_number = representation_ms_info['start_number'] | |
f0948348 S |
1943 | |
1944 | def add_segment_url(): | |
b4c1d6e8 S |
1945 | segment_url = media_template % { |
1946 | 'Time': segment_time, | |
e228616c | 1947 | 'Bandwidth': bandwidth, |
b4c1d6e8 S |
1948 | 'Number': segment_number, |
1949 | } | |
b4c1d6e8 | 1950 | representation_ms_info['fragments'].append({ |
1141e910 | 1951 | media_location_key: segment_url, |
b4c1d6e8 S |
1952 | 'duration': float_or_none(segment_d, representation_ms_info['timescale']), |
1953 | }) | |
f0948348 S |
1954 | |
1955 | for num, s in enumerate(representation_ms_info['s']): | |
1956 | segment_time = s.get('t') or segment_time | |
b4c1d6e8 | 1957 | segment_d = s['d'] |
f0948348 | 1958 | add_segment_url() |
b4c1d6e8 | 1959 | segment_number += 1 |
f0948348 | 1960 | for r in range(s.get('r', 0)): |
b4c1d6e8 | 1961 | segment_time += segment_d |
f0948348 | 1962 | add_segment_url() |
b4c1d6e8 S |
1963 | segment_number += 1 |
1964 | segment_time += segment_d | |
1965 | elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info: | |
1966 | # No media template | |
1967 | # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI | |
1968 | # or any YouTube dashsegments video | |
1969 | fragments = [] | |
d04621da S |
1970 | segment_index = 0 |
1971 | timescale = representation_ms_info['timescale'] | |
1972 | for s in representation_ms_info['s']: | |
1973 | duration = float_or_none(s['d'], timescale) | |
b4c1d6e8 | 1974 | for r in range(s.get('r', 0) + 1): |
1141e910 | 1975 | segment_uri = representation_ms_info['segment_urls'][segment_index] |
b4c1d6e8 | 1976 | fragments.append({ |
1141e910 | 1977 | location_key(segment_uri): segment_uri, |
d04621da | 1978 | 'duration': duration, |
b4c1d6e8 | 1979 | }) |
d04621da | 1980 | segment_index += 1 |
b4c1d6e8 | 1981 | representation_ms_info['fragments'] = fragments |
86f4d14f S |
1982 | # NB: MPD manifest may contain direct URLs to unfragmented media. |
1983 | # No fragments key is present in this case. | |
1984 | if 'fragments' in representation_ms_info: | |
1bac3455 | 1985 | f.update({ |
1141e910 | 1986 | 'fragment_base_url': base_url, |
b4c1d6e8 | 1987 | 'fragments': [], |
1bac3455 | 1988 | 'protocol': 'http_dash_segments', |
df374b52 | 1989 | }) |
1bac3455 | 1990 | if 'initialization_url' in representation_ms_info: |
e228616c | 1991 | initialization_url = representation_ms_info['initialization_url'] |
1bac3455 | 1992 | if not f.get('url'): |
1993 | f['url'] = initialization_url | |
1141e910 | 1994 | f['fragments'].append({location_key(initialization_url): initialization_url}) |
b4c1d6e8 | 1995 | f['fragments'].extend(representation_ms_info['fragments']) |
1bac3455 | 1996 | try: |
1997 | existing_format = next( | |
1998 | fo for fo in formats | |
1999 | if fo['format_id'] == representation_id) | |
2000 | except StopIteration: | |
2001 | full_info = formats_dict.get(representation_id, {}).copy() | |
2002 | full_info.update(f) | |
2003 | formats.append(full_info) | |
2004 | else: | |
2005 | existing_format.update(f) | |
17b598d3 | 2006 | else: |
1bac3455 | 2007 | self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) |
17b598d3 YCH |
2008 | return formats |
2009 | ||
b2758123 RA |
2010 | def _extract_ism_formats(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True): |
2011 | res = self._download_webpage_handle( | |
2012 | ism_url, video_id, | |
2013 | note=note or 'Downloading ISM manifest', | |
2014 | errnote=errnote or 'Failed to download ISM manifest', | |
2015 | fatal=fatal) | |
2016 | if res is False: | |
2017 | return [] | |
2018 | ism, urlh = res | |
2019 | ||
2020 | return self._parse_ism_formats( | |
2021 | compat_etree_fromstring(ism.encode('utf-8')), urlh.geturl(), ism_id) | |
2022 | ||
2023 | def _parse_ism_formats(self, ism_doc, ism_url, ism_id=None): | |
76d5a363 S |
2024 | """ |
2025 | Parse formats from ISM manifest. | |
2026 | References: | |
2027 | 1. [MS-SSTR]: Smooth Streaming Protocol, | |
2028 | https://msdn.microsoft.com/en-us/library/ff469518.aspx | |
2029 | """ | |
b2758123 RA |
2030 | if ism_doc.get('IsLive') == 'TRUE' or ism_doc.find('Protection') is not None: |
2031 | return [] | |
2032 | ||
b2758123 RA |
2033 | duration = int(ism_doc.attrib['Duration']) |
2034 | timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000 | |
2035 | ||
2036 | formats = [] | |
2037 | for stream in ism_doc.findall('StreamIndex'): | |
2038 | stream_type = stream.get('Type') | |
2039 | if stream_type not in ('video', 'audio'): | |
2040 | continue | |
2041 | url_pattern = stream.attrib['Url'] | |
2042 | stream_timescale = int_or_none(stream.get('TimeScale')) or timescale | |
2043 | stream_name = stream.get('Name') | |
2044 | for track in stream.findall('QualityLevel'): | |
2045 | fourcc = track.get('FourCC') | |
2046 | # TODO: add support for WVC1 and WMAP | |
2047 | if fourcc not in ('H264', 'AVC1', 'AACL'): | |
2048 | self.report_warning('%s is not a supported codec' % fourcc) | |
2049 | continue | |
2050 | tbr = int(track.attrib['Bitrate']) // 1000 | |
76d5a363 S |
2051 | # [1] does not mention Width and Height attributes. However, |
2052 | # they're often present while MaxWidth and MaxHeight are | |
2053 | # missing, so should be used as fallbacks | |
2054 | width = int_or_none(track.get('MaxWidth') or track.get('Width')) | |
2055 | height = int_or_none(track.get('MaxHeight') or track.get('Height')) | |
b2758123 RA |
2056 | sampling_rate = int_or_none(track.get('SamplingRate')) |
2057 | ||
2058 | track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern) | |
2059 | track_url_pattern = compat_urlparse.urljoin(ism_url, track_url_pattern) | |
2060 | ||
2061 | fragments = [] | |
2062 | fragment_ctx = { | |
2063 | 'time': 0, | |
2064 | } | |
2065 | stream_fragments = stream.findall('c') | |
2066 | for stream_fragment_index, stream_fragment in enumerate(stream_fragments): | |
2067 | fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time'] | |
2068 | fragment_repeat = int_or_none(stream_fragment.get('r')) or 1 | |
2069 | fragment_ctx['duration'] = int_or_none(stream_fragment.get('d')) | |
2070 | if not fragment_ctx['duration']: | |
2071 | try: | |
2072 | next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t']) | |
2073 | except IndexError: | |
2074 | next_fragment_time = duration | |
1616f9b4 | 2075 | fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat |
b2758123 RA |
2076 | for _ in range(fragment_repeat): |
2077 | fragments.append({ | |
1616f9b4 | 2078 | 'url': re.sub(r'{start[ _]time}', compat_str(fragment_ctx['time']), track_url_pattern), |
b2758123 RA |
2079 | 'duration': fragment_ctx['duration'] / stream_timescale, |
2080 | }) | |
2081 | fragment_ctx['time'] += fragment_ctx['duration'] | |
2082 | ||
2083 | format_id = [] | |
2084 | if ism_id: | |
2085 | format_id.append(ism_id) | |
2086 | if stream_name: | |
2087 | format_id.append(stream_name) | |
2088 | format_id.append(compat_str(tbr)) | |
2089 | ||
2090 | formats.append({ | |
2091 | 'format_id': '-'.join(format_id), | |
2092 | 'url': ism_url, | |
2093 | 'manifest_url': ism_url, | |
2094 | 'ext': 'ismv' if stream_type == 'video' else 'isma', | |
2095 | 'width': width, | |
2096 | 'height': height, | |
2097 | 'tbr': tbr, | |
2098 | 'asr': sampling_rate, | |
2099 | 'vcodec': 'none' if stream_type == 'audio' else fourcc, | |
2100 | 'acodec': 'none' if stream_type == 'video' else fourcc, | |
2101 | 'protocol': 'ism', | |
2102 | 'fragments': fragments, | |
2103 | '_download_params': { | |
2104 | 'duration': duration, | |
2105 | 'timescale': stream_timescale, | |
2106 | 'width': width or 0, | |
2107 | 'height': height or 0, | |
2108 | 'fourcc': fourcc, | |
2109 | 'codec_private_data': track.get('CodecPrivateData'), | |
2110 | 'sampling_rate': sampling_rate, | |
2111 | 'channels': int_or_none(track.get('Channels', 2)), | |
2112 | 'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)), | |
2113 | 'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)), | |
2114 | }, | |
2115 | }) | |
2116 | return formats | |
2117 | ||
eeb0a956 | 2118 | def _parse_html5_media_entries(self, base_url, webpage, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8', mpd_id=None, preference=None): |
59bbe491 | 2119 | def absolute_url(video_url): |
2120 | return compat_urlparse.urljoin(base_url, video_url) | |
2121 | ||
2122 | def parse_content_type(content_type): | |
2123 | if not content_type: | |
2124 | return {} | |
2125 | ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type) | |
2126 | if ctr: | |
2127 | mimetype, codecs = ctr.groups() | |
2128 | f = parse_codecs(codecs) | |
2129 | f['ext'] = mimetype2ext(mimetype) | |
2130 | return f | |
2131 | return {} | |
2132 | ||
868f79db | 2133 | def _media_formats(src, cur_media_type, type_info={}): |
520251c0 | 2134 | full_url = absolute_url(src) |
82889d4a | 2135 | ext = type_info.get('ext') or determine_ext(full_url) |
87a449c1 | 2136 | if ext == 'm3u8': |
520251c0 YCH |
2137 | is_plain_url = False |
2138 | formats = self._extract_m3u8_formats( | |
ad120ae1 | 2139 | full_url, video_id, ext='mp4', |
eeb0a956 | 2140 | entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id, |
b359e977 | 2141 | preference=preference, fatal=False) |
87a449c1 S |
2142 | elif ext == 'mpd': |
2143 | is_plain_url = False | |
2144 | formats = self._extract_mpd_formats( | |
b359e977 | 2145 | full_url, video_id, mpd_id=mpd_id, fatal=False) |
520251c0 YCH |
2146 | else: |
2147 | is_plain_url = True | |
2148 | formats = [{ | |
2149 | 'url': full_url, | |
2150 | 'vcodec': 'none' if cur_media_type == 'audio' else None, | |
2151 | }] | |
2152 | return is_plain_url, formats | |
2153 | ||
59bbe491 | 2154 | entries = [] |
4328ddf8 S |
2155 | # amp-video and amp-audio are very similar to their HTML5 counterparts |
2156 | # so we wll include them right here (see | |
2157 | # https://www.ampproject.org/docs/reference/components/amp-video) | |
cea364f7 YCH |
2158 | media_tags = [(media_tag, media_type, '') |
2159 | for media_tag, media_type | |
4328ddf8 | 2160 | in re.findall(r'(?s)(<(?:amp-)?(video|audio)[^>]*/>)', webpage)] |
2aec7256 S |
2161 | media_tags.extend(re.findall( |
2162 | # We only allow video|audio followed by a whitespace or '>'. | |
2163 | # Allowing more characters may end up in significant slow down (see | |
2164 | # https://github.com/rg3/youtube-dl/issues/11979, example URL: | |
2165 | # http://www.porntrex.com/maps/videositemap.xml). | |
4328ddf8 | 2166 | r'(?s)(<(?P<tag>(?:amp-)?(?:video|audio))(?:\s+[^>]*)?>)(.*?)</(?P=tag)>', webpage)) |
cea364f7 | 2167 | for media_tag, media_type, media_content in media_tags: |
59bbe491 | 2168 | media_info = { |
2169 | 'formats': [], | |
2170 | 'subtitles': {}, | |
2171 | } | |
2172 | media_attributes = extract_attributes(media_tag) | |
2173 | src = media_attributes.get('src') | |
2174 | if src: | |
dedb1770 | 2175 | _, formats = _media_formats(src, media_type) |
520251c0 | 2176 | media_info['formats'].extend(formats) |
59bbe491 | 2177 | media_info['thumbnail'] = media_attributes.get('poster') |
2178 | if media_content: | |
2179 | for source_tag in re.findall(r'<source[^>]+>', media_content): | |
2180 | source_attributes = extract_attributes(source_tag) | |
2181 | src = source_attributes.get('src') | |
2182 | if not src: | |
2183 | continue | |
82889d4a | 2184 | f = parse_content_type(source_attributes.get('type')) |
868f79db | 2185 | is_plain_url, formats = _media_formats(src, media_type, f) |
520251c0 | 2186 | if is_plain_url: |
520251c0 YCH |
2187 | f.update(formats[0]) |
2188 | media_info['formats'].append(f) | |
2189 | else: | |
2190 | media_info['formats'].extend(formats) | |
59bbe491 | 2191 | for track_tag in re.findall(r'<track[^>]+>', media_content): |
2192 | track_attributes = extract_attributes(track_tag) | |
2193 | kind = track_attributes.get('kind') | |
5968d7d2 | 2194 | if not kind or kind in ('subtitles', 'captions'): |
59bbe491 | 2195 | src = track_attributes.get('src') |
2196 | if not src: | |
2197 | continue | |
2198 | lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label') | |
2199 | media_info['subtitles'].setdefault(lang, []).append({ | |
2200 | 'url': absolute_url(src), | |
2201 | }) | |
5968d7d2 | 2202 | if media_info['formats'] or media_info['subtitles']: |
59bbe491 | 2203 | entries.append(media_info) |
2204 | return entries | |
2205 | ||
c4251b9a | 2206 | def _extract_akamai_formats(self, manifest_url, video_id, hosts={}): |
c7c43a93 | 2207 | formats = [] |
e71a4509 | 2208 | hdcore_sign = 'hdcore=3.7.0' |
ff6f9a67 | 2209 | f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') |
c4251b9a RA |
2210 | hds_host = hosts.get('hds') |
2211 | if hds_host: | |
2212 | f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url) | |
e71a4509 RA |
2213 | if 'hdcore=' not in f4m_url: |
2214 | f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign | |
2215 | f4m_formats = self._extract_f4m_formats( | |
2216 | f4m_url, video_id, f4m_id='hds', fatal=False) | |
2217 | for entry in f4m_formats: | |
2218 | entry.update({'extra_param_to_segment_url': hdcore_sign}) | |
2219 | formats.extend(f4m_formats) | |
c4251b9a RA |
2220 | m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8') |
2221 | hls_host = hosts.get('hls') | |
2222 | if hls_host: | |
2223 | m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url) | |
c7c43a93 RA |
2224 | formats.extend(self._extract_m3u8_formats( |
2225 | m3u8_url, video_id, 'mp4', 'm3u8_native', | |
2226 | m3u8_id='hls', fatal=False)) | |
2227 | return formats | |
2228 | ||
6ad02195 RA |
2229 | def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]): |
2230 | url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url) | |
6f766798 S |
2231 | url_base = self._search_regex( |
2232 | r'(?:(?:https?|rtmp|rtsp):)?(//[^?]+)', url, 'format url') | |
f2e2f0c7 | 2233 | http_base_url = '%s:%s' % ('http', url_base) |
6ad02195 RA |
2234 | formats = [] |
2235 | if 'm3u8' not in skip_protocols: | |
2236 | formats.extend(self._extract_m3u8_formats( | |
2237 | http_base_url + '/playlist.m3u8', video_id, 'mp4', | |
2238 | m3u8_entry_protocol, m3u8_id='hls', fatal=False)) | |
2239 | if 'f4m' not in skip_protocols: | |
2240 | formats.extend(self._extract_f4m_formats( | |
2241 | http_base_url + '/manifest.f4m', | |
2242 | video_id, f4m_id='hds', fatal=False)) | |
0384932e RA |
2243 | if 'dash' not in skip_protocols: |
2244 | formats.extend(self._extract_mpd_formats( | |
2245 | http_base_url + '/manifest.mpd', | |
2246 | video_id, mpd_id='dash', fatal=False)) | |
6ad02195 | 2247 | if re.search(r'(?:/smil:|\.smil)', url_base): |
6ad02195 RA |
2248 | if 'smil' not in skip_protocols: |
2249 | rtmp_formats = self._extract_smil_formats( | |
2250 | http_base_url + '/jwplayer.smil', | |
2251 | video_id, fatal=False) | |
2252 | for rtmp_format in rtmp_formats: | |
2253 | rtsp_format = rtmp_format.copy() | |
2254 | rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path']) | |
2255 | del rtsp_format['play_path'] | |
2256 | del rtsp_format['ext'] | |
2257 | rtsp_format.update({ | |
2258 | 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'), | |
2259 | 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'), | |
2260 | 'protocol': 'rtsp', | |
2261 | }) | |
2262 | formats.extend([rtmp_format, rtsp_format]) | |
2263 | else: | |
2264 | for protocol in ('rtmp', 'rtsp'): | |
2265 | if protocol not in skip_protocols: | |
2266 | formats.append({ | |
f2e2f0c7 | 2267 | 'url': '%s:%s' % (protocol, url_base), |
6ad02195 RA |
2268 | 'format_id': protocol, |
2269 | 'protocol': protocol, | |
2270 | }) | |
2271 | return formats | |
2272 | ||
c73e330e | 2273 | def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json): |
a4a554a7 | 2274 | mobj = re.search( |
ac9c69ac | 2275 | r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)', |
a4a554a7 YCH |
2276 | webpage) |
2277 | if mobj: | |
c73e330e RU |
2278 | try: |
2279 | jwplayer_data = self._parse_json(mobj.group('options'), | |
2280 | video_id=video_id, | |
2281 | transform_source=transform_source) | |
2282 | except ExtractorError: | |
2283 | pass | |
2284 | else: | |
2285 | if isinstance(jwplayer_data, dict): | |
2286 | return jwplayer_data | |
a4a554a7 YCH |
2287 | |
2288 | def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs): | |
c73e330e RU |
2289 | jwplayer_data = self._find_jwplayer_data( |
2290 | webpage, video_id, transform_source=js_to_json) | |
a4a554a7 YCH |
2291 | return self._parse_jwplayer_data( |
2292 | jwplayer_data, video_id, *args, **kwargs) | |
2293 | ||
2294 | def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True, | |
2295 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
2296 | # JWPlayer backward compatibility: flattened playlists | |
2297 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96 | |
2298 | if 'playlist' not in jwplayer_data: | |
2299 | jwplayer_data = {'playlist': [jwplayer_data]} | |
2300 | ||
2301 | entries = [] | |
2302 | ||
2303 | # JWPlayer backward compatibility: single playlist item | |
2304 | # https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10 | |
2305 | if not isinstance(jwplayer_data['playlist'], list): | |
2306 | jwplayer_data['playlist'] = [jwplayer_data['playlist']] | |
2307 | ||
2308 | for video_data in jwplayer_data['playlist']: | |
2309 | # JWPlayer backward compatibility: flattened sources | |
2310 | # https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35 | |
2311 | if 'sources' not in video_data: | |
2312 | video_data['sources'] = [video_data] | |
2313 | ||
2314 | this_video_id = video_id or video_data['mediaid'] | |
2315 | ||
1a2192cb S |
2316 | formats = self._parse_jwplayer_formats( |
2317 | video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id, | |
2318 | mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url) | |
a4a554a7 YCH |
2319 | self._sort_formats(formats) |
2320 | ||
2321 | subtitles = {} | |
2322 | tracks = video_data.get('tracks') | |
2323 | if tracks and isinstance(tracks, list): | |
2324 | for track in tracks: | |
96a2daa1 S |
2325 | if not isinstance(track, dict): |
2326 | continue | |
a4a554a7 YCH |
2327 | if track.get('kind') != 'captions': |
2328 | continue | |
2329 | track_url = urljoin(base_url, track.get('file')) | |
2330 | if not track_url: | |
2331 | continue | |
2332 | subtitles.setdefault(track.get('label') or 'en', []).append({ | |
2333 | 'url': self._proto_relative_url(track_url) | |
2334 | }) | |
2335 | ||
2336 | entries.append({ | |
2337 | 'id': this_video_id, | |
2338 | 'title': video_data['title'] if require_title else video_data.get('title'), | |
2339 | 'description': video_data.get('description'), | |
2340 | 'thumbnail': self._proto_relative_url(video_data.get('image')), | |
2341 | 'timestamp': int_or_none(video_data.get('pubdate')), | |
2342 | 'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')), | |
2343 | 'subtitles': subtitles, | |
2344 | 'formats': formats, | |
2345 | }) | |
2346 | if len(entries) == 1: | |
2347 | return entries[0] | |
2348 | else: | |
2349 | return self.playlist_result(entries) | |
2350 | ||
ed0cf9b3 S |
2351 | def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None, |
2352 | m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None): | |
bf1b87cd | 2353 | urls = [] |
ed0cf9b3 | 2354 | formats = [] |
1a2192cb | 2355 | for source in jwplayer_sources_data: |
0a268c6e S |
2356 | if not isinstance(source, dict): |
2357 | continue | |
bf1b87cd RA |
2358 | source_url = self._proto_relative_url(source.get('file')) |
2359 | if not source_url: | |
2360 | continue | |
ed0cf9b3 S |
2361 | if base_url: |
2362 | source_url = compat_urlparse.urljoin(base_url, source_url) | |
bf1b87cd RA |
2363 | if source_url in urls: |
2364 | continue | |
2365 | urls.append(source_url) | |
ed0cf9b3 S |
2366 | source_type = source.get('type') or '' |
2367 | ext = mimetype2ext(source_type) or determine_ext(source_url) | |
2368 | if source_type == 'hls' or ext == 'm3u8': | |
2369 | formats.extend(self._extract_m3u8_formats( | |
0236cd0d S |
2370 | source_url, video_id, 'mp4', entry_protocol='m3u8_native', |
2371 | m3u8_id=m3u8_id, fatal=False)) | |
ed0cf9b3 S |
2372 | elif ext == 'mpd': |
2373 | formats.extend(self._extract_mpd_formats( | |
2374 | source_url, video_id, mpd_id=mpd_id, fatal=False)) | |
b51dc9db S |
2375 | elif ext == 'smil': |
2376 | formats.extend(self._extract_smil_formats( | |
2377 | source_url, video_id, fatal=False)) | |
ed0cf9b3 | 2378 | # https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67 |
0236cd0d S |
2379 | elif source_type.startswith('audio') or ext in ( |
2380 | 'oga', 'aac', 'mp3', 'mpeg', 'vorbis'): | |
ed0cf9b3 S |
2381 | formats.append({ |
2382 | 'url': source_url, | |
2383 | 'vcodec': 'none', | |
2384 | 'ext': ext, | |
2385 | }) | |
2386 | else: | |
2387 | height = int_or_none(source.get('height')) | |
2388 | if height is None: | |
2389 | # Often no height is provided but there is a label in | |
0236cd0d | 2390 | # format like "1080p", "720p SD", or 1080. |
ed0cf9b3 | 2391 | height = int_or_none(self._search_regex( |
0236cd0d | 2392 | r'^(\d{3,4})[pP]?(?:\b|$)', compat_str(source.get('label') or ''), |
ed0cf9b3 S |
2393 | 'height', default=None)) |
2394 | a_format = { | |
2395 | 'url': source_url, | |
2396 | 'width': int_or_none(source.get('width')), | |
2397 | 'height': height, | |
0236cd0d | 2398 | 'tbr': int_or_none(source.get('bitrate')), |
ed0cf9b3 S |
2399 | 'ext': ext, |
2400 | } | |
2401 | if source_url.startswith('rtmp'): | |
2402 | a_format['ext'] = 'flv' | |
ed0cf9b3 S |
2403 | # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as |
2404 | # of jwplayer.flash.swf | |
2405 | rtmp_url_parts = re.split( | |
2406 | r'((?:mp4|mp3|flv):)', source_url, 1) | |
2407 | if len(rtmp_url_parts) == 3: | |
2408 | rtmp_url, prefix, play_path = rtmp_url_parts | |
2409 | a_format.update({ | |
2410 | 'url': rtmp_url, | |
2411 | 'play_path': prefix + play_path, | |
2412 | }) | |
2413 | if rtmp_params: | |
2414 | a_format.update(rtmp_params) | |
2415 | formats.append(a_format) | |
2416 | return formats | |
2417 | ||
f4b1c7ad PH |
2418 | def _live_title(self, name): |
2419 | """ Generate the title for a live video """ | |
2420 | now = datetime.datetime.now() | |
611c1dd9 | 2421 | now_str = now.strftime('%Y-%m-%d %H:%M') |
f4b1c7ad PH |
2422 | return name + ' ' + now_str |
2423 | ||
b14f3a4c PH |
2424 | def _int(self, v, name, fatal=False, **kwargs): |
2425 | res = int_or_none(v, **kwargs) | |
2426 | if 'get_attr' in kwargs: | |
2427 | print(getattr(v, kwargs['get_attr'])) | |
2428 | if res is None: | |
2429 | msg = 'Failed to extract %s: Could not parse value %r' % (name, v) | |
2430 | if fatal: | |
2431 | raise ExtractorError(msg) | |
2432 | else: | |
2433 | self._downloader.report_warning(msg) | |
2434 | return res | |
2435 | ||
2436 | def _float(self, v, name, fatal=False, **kwargs): | |
2437 | res = float_or_none(v, **kwargs) | |
2438 | if res is None: | |
2439 | msg = 'Failed to extract %s: Could not parse value %r' % (name, v) | |
2440 | if fatal: | |
2441 | raise ExtractorError(msg) | |
2442 | else: | |
2443 | self._downloader.report_warning(msg) | |
2444 | return res | |
2445 | ||
42939b61 | 2446 | def _set_cookie(self, domain, name, value, expire_time=None): |
810fb84d PH |
2447 | cookie = compat_cookiejar.Cookie( |
2448 | 0, name, value, None, None, domain, None, | |
42939b61 JMF |
2449 | None, '/', True, False, expire_time, '', None, None, None) |
2450 | self._downloader.cookiejar.set_cookie(cookie) | |
2451 | ||
799207e8 | 2452 | def _get_cookies(self, url): |
2453 | """ Return a compat_cookies.SimpleCookie with the cookies for the url """ | |
5c2266df | 2454 | req = sanitized_Request(url) |
799207e8 | 2455 | self._downloader.cookiejar.add_cookie_header(req) |
2456 | return compat_cookies.SimpleCookie(req.get_header('Cookie')) | |
2457 | ||
05900629 PH |
2458 | def get_testcases(self, include_onlymatching=False): |
2459 | t = getattr(self, '_TEST', None) | |
2460 | if t: | |
2461 | assert not hasattr(self, '_TESTS'), \ | |
2462 | '%s has _TEST and _TESTS' % type(self).__name__ | |
2463 | tests = [t] | |
2464 | else: | |
2465 | tests = getattr(self, '_TESTS', []) | |
2466 | for t in tests: | |
2467 | if not include_onlymatching and t.get('only_matching', False): | |
2468 | continue | |
2469 | t['name'] = type(self).__name__[:-len('IE')] | |
2470 | yield t | |
2471 | ||
2472 | def is_suitable(self, age_limit): | |
2473 | """ Test whether the extractor is generally suitable for the given | |
2474 | age limit (i.e. pornographic sites are not, all others usually are) """ | |
2475 | ||
2476 | any_restricted = False | |
2477 | for tc in self.get_testcases(include_onlymatching=False): | |
40090e8d | 2478 | if tc.get('playlist', []): |
05900629 PH |
2479 | tc = tc['playlist'][0] |
2480 | is_restricted = age_restricted( | |
2481 | tc.get('info_dict', {}).get('age_limit'), age_limit) | |
2482 | if not is_restricted: | |
2483 | return True | |
2484 | any_restricted = any_restricted or is_restricted | |
2485 | return not any_restricted | |
2486 | ||
a504ced0 | 2487 | def extract_subtitles(self, *args, **kwargs): |
9868ea49 JMF |
2488 | if (self._downloader.params.get('writesubtitles', False) or |
2489 | self._downloader.params.get('listsubtitles')): | |
2490 | return self._get_subtitles(*args, **kwargs) | |
2491 | return {} | |
a504ced0 JMF |
2492 | |
2493 | def _get_subtitles(self, *args, **kwargs): | |
611c1dd9 | 2494 | raise NotImplementedError('This method must be implemented by subclasses') |
a504ced0 | 2495 | |
912e0b7e YCH |
2496 | @staticmethod |
2497 | def _merge_subtitle_items(subtitle_list1, subtitle_list2): | |
2498 | """ Merge subtitle items for one language. Items with duplicated URLs | |
2499 | will be dropped. """ | |
2500 | list1_urls = set([item['url'] for item in subtitle_list1]) | |
2501 | ret = list(subtitle_list1) | |
2502 | ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls]) | |
2503 | return ret | |
2504 | ||
2505 | @classmethod | |
8c97f819 | 2506 | def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2): |
912e0b7e | 2507 | """ Merge two subtitle dictionaries, language by language. """ |
912e0b7e YCH |
2508 | ret = dict(subtitle_dict1) |
2509 | for lang in subtitle_dict2: | |
8c97f819 | 2510 | ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang]) |
912e0b7e YCH |
2511 | return ret |
2512 | ||
360e1ca5 | 2513 | def extract_automatic_captions(self, *args, **kwargs): |
9868ea49 JMF |
2514 | if (self._downloader.params.get('writeautomaticsub', False) or |
2515 | self._downloader.params.get('listsubtitles')): | |
2516 | return self._get_automatic_captions(*args, **kwargs) | |
2517 | return {} | |
360e1ca5 JMF |
2518 | |
2519 | def _get_automatic_captions(self, *args, **kwargs): | |
611c1dd9 | 2520 | raise NotImplementedError('This method must be implemented by subclasses') |
360e1ca5 | 2521 | |
d77ab8e2 S |
2522 | def mark_watched(self, *args, **kwargs): |
2523 | if (self._downloader.params.get('mark_watched', False) and | |
2524 | (self._get_login_info()[0] is not None or | |
2525 | self._downloader.params.get('cookiefile') is not None)): | |
2526 | self._mark_watched(*args, **kwargs) | |
2527 | ||
2528 | def _mark_watched(self, *args, **kwargs): | |
2529 | raise NotImplementedError('This method must be implemented by subclasses') | |
2530 | ||
38cce791 YCH |
2531 | def geo_verification_headers(self): |
2532 | headers = {} | |
2533 | geo_verification_proxy = self._downloader.params.get('geo_verification_proxy') | |
2534 | if geo_verification_proxy: | |
2535 | headers['Ytdl-request-proxy'] = geo_verification_proxy | |
2536 | return headers | |
2537 | ||
98763ee3 YCH |
2538 | def _generic_id(self, url): |
2539 | return compat_urllib_parse_unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0]) | |
2540 | ||
2541 | def _generic_title(self, url): | |
2542 | return compat_urllib_parse_unquote(os.path.splitext(url_basename(url))[0]) | |
2543 | ||
8dbe9899 | 2544 | |
d6983cb4 PH |
2545 | class SearchInfoExtractor(InfoExtractor): |
2546 | """ | |
2547 | Base class for paged search queries extractors. | |
10952eb2 | 2548 | They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query} |
d6983cb4 PH |
2549 | Instances should define _SEARCH_KEY and _MAX_RESULTS. |
2550 | """ | |
2551 | ||
2552 | @classmethod | |
2553 | def _make_valid_url(cls): | |
2554 | return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY | |
2555 | ||
2556 | @classmethod | |
2557 | def suitable(cls, url): | |
2558 | return re.match(cls._make_valid_url(), url) is not None | |
2559 | ||
2560 | def _real_extract(self, query): | |
2561 | mobj = re.match(self._make_valid_url(), query) | |
2562 | if mobj is None: | |
f1a9d64e | 2563 | raise ExtractorError('Invalid search query "%s"' % query) |
d6983cb4 PH |
2564 | |
2565 | prefix = mobj.group('prefix') | |
2566 | query = mobj.group('query') | |
2567 | if prefix == '': | |
2568 | return self._get_n_results(query, 1) | |
2569 | elif prefix == 'all': | |
2570 | return self._get_n_results(query, self._MAX_RESULTS) | |
2571 | else: | |
2572 | n = int(prefix) | |
2573 | if n <= 0: | |
f1a9d64e | 2574 | raise ExtractorError('invalid download number %s for query "%s"' % (n, query)) |
d6983cb4 | 2575 | elif n > self._MAX_RESULTS: |
f1a9d64e | 2576 | self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n)) |
d6983cb4 PH |
2577 | n = self._MAX_RESULTS |
2578 | return self._get_n_results(query, n) | |
2579 | ||
2580 | def _get_n_results(self, query, n): | |
2581 | """Get a specified number of results for a query""" | |
611c1dd9 | 2582 | raise NotImplementedError('This method must be implemented by subclasses') |
0f818663 PH |
2583 | |
2584 | @property | |
2585 | def SEARCH_KEY(self): | |
2586 | return self._SEARCH_KEY |