2013-06-23 13:57:38 -04:00
|
|
|
|
import base64
|
2021-11-22 13:02:14 -05:00
|
|
|
|
import collections
|
2022-06-24 04:10:17 -04:00
|
|
|
|
import getpass
|
2014-01-17 08:47:46 -05:00
|
|
|
|
import hashlib
|
2022-06-24 07:06:16 -04:00
|
|
|
|
import http.client
|
|
|
|
|
import http.cookiejar
|
|
|
|
|
import http.cookies
|
2022-09-16 07:07:38 -04:00
|
|
|
|
import inspect
|
2021-10-08 16:39:55 -04:00
|
|
|
|
import itertools
|
2014-01-07 03:35:34 -05:00
|
|
|
|
import json
|
2022-04-11 18:32:57 -04:00
|
|
|
|
import math
|
2014-07-11 04:57:08 -04:00
|
|
|
|
import netrc
|
2013-06-23 13:57:38 -04:00
|
|
|
|
import os
|
2017-02-04 06:49:58 -05:00
|
|
|
|
import random
|
2022-07-17 20:20:54 -04:00
|
|
|
|
import re
|
2013-06-23 13:57:38 -04:00
|
|
|
|
import sys
|
2014-07-11 04:57:08 -04:00
|
|
|
|
import time
|
2022-07-31 21:22:03 -04:00
|
|
|
|
import types
|
2022-06-24 06:54:43 -04:00
|
|
|
|
import urllib.parse
|
2022-06-24 04:10:17 -04:00
|
|
|
|
import urllib.request
|
2022-04-11 18:32:57 -04:00
|
|
|
|
import xml.etree.ElementTree
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2022-07-17 20:20:54 -04:00
|
|
|
|
from ..compat import functools # isort: split
|
2022-06-24 06:54:43 -04:00
|
|
|
|
from ..compat import compat_etree_fromstring, compat_expanduser, compat_os_name
|
2022-09-16 13:02:00 -04:00
|
|
|
|
from ..cookies import LenientSimpleCookie
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
from ..downloader import FileDownloader
|
2022-04-11 18:32:57 -04:00
|
|
|
|
from ..downloader.f4m import get_base_url, remove_encrypted_media
|
2014-11-02 05:23:40 -05:00
|
|
|
|
from ..utils import (
|
2022-07-31 21:22:03 -04:00
|
|
|
|
IDENTITY,
|
2022-04-11 18:32:57 -04:00
|
|
|
|
JSON_LD_RE,
|
|
|
|
|
NO_DEFAULT,
|
|
|
|
|
ExtractorError,
|
|
|
|
|
GeoRestrictedError,
|
|
|
|
|
GeoUtils,
|
2022-06-03 11:32:31 -04:00
|
|
|
|
LenientJSONDecoder,
|
2022-04-11 18:32:57 -04:00
|
|
|
|
RegexNotFoundError,
|
2022-08-01 16:13:18 -04:00
|
|
|
|
RetryManager,
|
2022-04-11 18:32:57 -04:00
|
|
|
|
UnsupportedError,
|
2015-01-07 01:20:20 -05:00
|
|
|
|
age_restricted,
|
2016-11-01 15:14:01 -04:00
|
|
|
|
base_url,
|
2015-04-17 08:55:24 -04:00
|
|
|
|
bug_reports_message,
|
2022-05-11 11:54:44 -04:00
|
|
|
|
classproperty,
|
2013-06-23 13:57:38 -04:00
|
|
|
|
clean_html,
|
2015-07-15 15:15:15 -04:00
|
|
|
|
determine_ext,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
determine_protocol,
|
2019-03-16 22:09:32 -04:00
|
|
|
|
dict_get,
|
2022-01-19 17:23:55 -05:00
|
|
|
|
encode_data_uri,
|
2015-12-19 20:00:39 -05:00
|
|
|
|
error_to_compat_str,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
extract_attributes,
|
2022-03-27 22:51:45 -04:00
|
|
|
|
filter_dict,
|
2015-07-15 15:14:08 -04:00
|
|
|
|
fix_xml_ampersands,
|
2014-09-28 04:34:55 -04:00
|
|
|
|
float_or_none,
|
2021-08-07 07:20:46 -04:00
|
|
|
|
format_field,
|
2014-07-28 09:25:56 -04:00
|
|
|
|
int_or_none,
|
2021-11-05 21:05:24 -04:00
|
|
|
|
join_nonempty,
|
2017-02-16 10:42:36 -05:00
|
|
|
|
js_to_json,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
mimetype2ext,
|
2021-05-04 13:06:18 -04:00
|
|
|
|
network_exceptions,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
orderedSet,
|
2019-03-16 22:09:32 -04:00
|
|
|
|
parse_bitrate,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
parse_codecs,
|
|
|
|
|
parse_duration,
|
2016-01-15 13:36:02 -05:00
|
|
|
|
parse_iso8601,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
parse_m3u8_attributes,
|
2019-03-16 22:09:32 -04:00
|
|
|
|
parse_resolution,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
sanitize_filename,
|
2022-07-31 21:22:03 -04:00
|
|
|
|
sanitize_url,
|
2021-08-07 07:20:46 -04:00
|
|
|
|
sanitized_Request,
|
2022-10-09 06:39:36 -04:00
|
|
|
|
smuggle_url,
|
2019-03-16 22:09:32 -04:00
|
|
|
|
str_or_none,
|
2020-09-18 19:33:17 -04:00
|
|
|
|
str_to_int,
|
2019-05-23 12:52:11 -04:00
|
|
|
|
strip_or_none,
|
2021-06-25 10:05:41 -04:00
|
|
|
|
traverse_obj,
|
2022-06-28 20:37:21 -04:00
|
|
|
|
try_call,
|
2022-02-17 08:36:22 -05:00
|
|
|
|
try_get,
|
2013-07-17 04:38:23 -04:00
|
|
|
|
unescapeHTML,
|
2015-10-01 12:18:59 -04:00
|
|
|
|
unified_strdate,
|
2016-07-08 16:27:11 -04:00
|
|
|
|
unified_timestamp,
|
2017-03-08 18:13:54 -05:00
|
|
|
|
update_Request,
|
2022-06-02 20:45:54 -04:00
|
|
|
|
update_url_query,
|
2015-08-01 15:13:21 -04:00
|
|
|
|
url_basename,
|
2018-10-28 13:19:08 -04:00
|
|
|
|
url_or_none,
|
2021-08-07 07:20:46 -04:00
|
|
|
|
urljoin,
|
2021-07-10 17:59:44 -04:00
|
|
|
|
variadic,
|
2016-06-07 12:19:33 -04:00
|
|
|
|
xpath_element,
|
2015-08-09 09:07:18 -04:00
|
|
|
|
xpath_text,
|
|
|
|
|
xpath_with_ns,
|
2013-06-23 13:57:38 -04:00
|
|
|
|
)
|
2015-06-28 12:56:45 -04:00
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2022-04-11 11:10:28 -04:00
|
|
|
|
class InfoExtractor:
|
2013-06-23 13:57:38 -04:00
|
|
|
|
"""Information Extractor class.
|
|
|
|
|
|
|
|
|
|
Information extractors are the classes that, given a URL, extract
|
|
|
|
|
information about the video (or videos) the URL refers to. This
|
|
|
|
|
information includes the real video URL, the video title, author and
|
|
|
|
|
others. The information is stored in a dictionary which is then
|
2014-12-21 10:58:29 -05:00
|
|
|
|
passed to the YoutubeDL. The YoutubeDL processes this
|
2013-06-23 13:57:38 -04:00
|
|
|
|
information possibly downloading the video to the file system, among
|
|
|
|
|
other possible outcomes.
|
|
|
|
|
|
2015-04-29 11:03:10 -04:00
|
|
|
|
The type field determines the type of the result.
|
2014-11-20 10:47:59 -05:00
|
|
|
|
By far the most common value (and the default if _type is missing) is
|
|
|
|
|
"video", which indicates a single video.
|
|
|
|
|
|
|
|
|
|
For a video, the dictionaries must include the following fields:
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
|
|
|
|
id: Video identifier.
|
2022-05-07 10:15:00 -04:00
|
|
|
|
title: Video title, unescaped. Set to an empty string if video has
|
|
|
|
|
no title as opposed to "None" which signifies that the
|
|
|
|
|
extractor failed to obtain a title
|
2013-12-16 08:13:40 -05:00
|
|
|
|
|
2013-12-24 05:56:02 -05:00
|
|
|
|
Additionally, it must contain either a formats entry or a url one:
|
2013-12-16 08:13:40 -05:00
|
|
|
|
|
2013-12-24 05:56:02 -05:00
|
|
|
|
formats: A list of dictionaries for each format available, ordered
|
|
|
|
|
from worst to best quality.
|
|
|
|
|
|
|
|
|
|
Potential fields:
|
2019-03-04 12:39:15 -05:00
|
|
|
|
* url The mandatory URL representing the media:
|
|
|
|
|
for plain file media - HTTP URL of this file,
|
|
|
|
|
for RTMP - RTMP URL,
|
|
|
|
|
for HLS - URL of the M3U8 media playlist,
|
|
|
|
|
for HDS - URL of the F4M manifest,
|
2019-03-14 13:42:14 -04:00
|
|
|
|
for DASH
|
|
|
|
|
- HTTP URL to plain file media (in case of
|
|
|
|
|
unfragmented media)
|
|
|
|
|
- URL of the MPD manifest or base URL
|
|
|
|
|
representing the media if MPD manifest
|
2019-05-10 17:53:48 -04:00
|
|
|
|
is parsed from a string (in case of
|
2019-03-14 13:42:14 -04:00
|
|
|
|
fragmented media)
|
2019-03-04 12:39:15 -05:00
|
|
|
|
for MSS - URL of the ISM manifest.
|
2016-09-17 09:35:22 -04:00
|
|
|
|
* manifest_url
|
|
|
|
|
The URL of the manifest file in case of
|
2019-03-04 12:39:15 -05:00
|
|
|
|
fragmented media:
|
|
|
|
|
for HLS - URL of the M3U8 master playlist,
|
|
|
|
|
for HDS - URL of the F4M manifest,
|
|
|
|
|
for DASH - URL of the MPD manifest,
|
|
|
|
|
for MSS - URL of the ISM manifest.
|
2022-03-26 22:20:43 -04:00
|
|
|
|
* manifest_stream_number (For internal use only)
|
|
|
|
|
The index of the stream in the manifest file
|
2015-07-23 13:37:45 -04:00
|
|
|
|
* ext Will be calculated from URL if missing
|
2013-12-16 08:13:40 -05:00
|
|
|
|
* format A human-readable description of the format
|
|
|
|
|
("mp4 container with h264/opus").
|
|
|
|
|
Calculated from the format_id, width, height.
|
|
|
|
|
and format_note fields if missing.
|
|
|
|
|
* format_id A short description of the format
|
2013-12-26 15:19:00 -05:00
|
|
|
|
("mp4_h264_opus" or "19").
|
|
|
|
|
Technically optional, but strongly recommended.
|
2013-12-16 08:13:40 -05:00
|
|
|
|
* format_note Additional info about the format
|
|
|
|
|
("3D" or "DASH video")
|
|
|
|
|
* width Width of the video, if known
|
|
|
|
|
* height Height of the video, if known
|
2013-12-24 05:56:02 -05:00
|
|
|
|
* resolution Textual description of width and height
|
2021-10-18 09:04:21 -04:00
|
|
|
|
* dynamic_range The dynamic range of the video. One of:
|
|
|
|
|
"SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV"
|
2013-12-25 09:18:40 -05:00
|
|
|
|
* tbr Average bitrate of audio and video in KBit/s
|
2013-12-16 08:13:40 -05:00
|
|
|
|
* abr Average audio bitrate in KBit/s
|
|
|
|
|
* acodec Name of the audio codec in use
|
2014-01-18 23:47:20 -05:00
|
|
|
|
* asr Audio sampling rate in Hertz
|
2022-08-07 16:05:36 -04:00
|
|
|
|
* audio_channels Number of audio channels
|
2013-12-16 08:13:40 -05:00
|
|
|
|
* vbr Average video bitrate in KBit/s
|
2014-10-30 04:34:13 -04:00
|
|
|
|
* fps Frame rate
|
2013-12-16 08:13:40 -05:00
|
|
|
|
* vcodec Name of the video codec in use
|
2014-01-23 17:54:06 -05:00
|
|
|
|
* container Name of the container format
|
2013-12-16 08:13:40 -05:00
|
|
|
|
* filesize The number of bytes, if known in advance
|
2014-07-21 06:02:44 -04:00
|
|
|
|
* filesize_approx An estimate for the number of bytes
|
2013-12-16 08:13:40 -05:00
|
|
|
|
* player_url SWF Player URL (used for rtmpdump).
|
2013-12-24 17:32:04 -05:00
|
|
|
|
* protocol The protocol that will be used for the actual
|
2021-12-20 01:06:46 -05:00
|
|
|
|
download, lower-case. One of "http", "https" or
|
|
|
|
|
one of the protocols defined in downloader.PROTOCOL_MAP
|
2017-01-28 17:56:43 -05:00
|
|
|
|
* fragment_base_url
|
|
|
|
|
Base URL for fragments. Each fragment's path
|
|
|
|
|
value (if present) will be relative to
|
|
|
|
|
this URL.
|
|
|
|
|
* fragments A list of fragments of a fragmented media.
|
|
|
|
|
Each fragment entry must contain either an url
|
|
|
|
|
or a path. If an url is present it should be
|
|
|
|
|
considered by a client. Otherwise both path and
|
|
|
|
|
fragment_base_url must be present. Here is
|
|
|
|
|
the list of all potential fields:
|
|
|
|
|
* "url" - fragment's URL
|
|
|
|
|
* "path" - fragment's path relative to
|
|
|
|
|
fragment_base_url
|
2016-09-05 14:18:57 -04:00
|
|
|
|
* "duration" (optional, int or float)
|
|
|
|
|
* "filesize" (optional, int)
|
2021-12-20 01:06:46 -05:00
|
|
|
|
* is_from_start Is a live format that can be downloaded
|
|
|
|
|
from the start. Boolean
|
2013-12-24 05:56:02 -05:00
|
|
|
|
* preference Order number of this format. If this field is
|
2014-01-01 14:23:47 -05:00
|
|
|
|
present and not None, the formats get sorted
|
2014-03-23 12:41:43 -04:00
|
|
|
|
by this field, regardless of all other values.
|
2013-12-24 05:56:02 -05:00
|
|
|
|
-1 for default (order by other properties),
|
|
|
|
|
-2 or smaller for less than default.
|
2015-01-03 12:33:38 -05:00
|
|
|
|
< -1000 to hide the format (if there is
|
|
|
|
|
another one which is strictly better)
|
2016-01-01 07:28:45 -05:00
|
|
|
|
* language Language code, e.g. "de" or "en-US".
|
|
|
|
|
* language_preference Is this in the language mentioned in
|
|
|
|
|
the URL?
|
2014-11-20 06:06:33 -05:00
|
|
|
|
10 if it's what the URL is about,
|
|
|
|
|
-1 for default (don't know),
|
|
|
|
|
-10 otherwise, other values reserved for now.
|
2014-01-06 11:15:27 -05:00
|
|
|
|
* quality Order number of the video quality of this
|
|
|
|
|
format, irrespective of the file format.
|
|
|
|
|
-1 for default (order by other properties),
|
|
|
|
|
-2 or smaller for less than default.
|
2014-10-24 18:10:11 -04:00
|
|
|
|
* source_preference Order number for this video source
|
|
|
|
|
(quality takes higher priority)
|
|
|
|
|
-1 for default (order by other properties),
|
|
|
|
|
-2 or smaller for less than default.
|
2014-08-23 19:31:35 -04:00
|
|
|
|
* http_headers A dictionary of additional HTTP headers
|
|
|
|
|
to add to the request.
|
2015-01-09 23:45:51 -05:00
|
|
|
|
* stretched_ratio If given and not 1, indicates that the
|
2015-01-25 12:09:48 -05:00
|
|
|
|
video's pixels are not square.
|
|
|
|
|
width : height ratio as float.
|
|
|
|
|
* no_resume The server does not support resuming the
|
|
|
|
|
(HTTP or RTMP) download. Boolean.
|
2021-08-22 16:08:38 -04:00
|
|
|
|
* has_drm The format has DRM and cannot be downloaded. Boolean
|
2022-04-27 12:22:57 -04:00
|
|
|
|
* downloader_options A dictionary of downloader options
|
|
|
|
|
(For internal use only)
|
|
|
|
|
* http_chunk_size Chunk size for HTTP downloads
|
|
|
|
|
* ffmpeg_args Extra arguments for ffmpeg downloader
|
2021-06-08 10:43:41 -04:00
|
|
|
|
RTMP formats can also have the additional fields: page_url,
|
|
|
|
|
app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
|
|
|
|
|
rtmp_protocol, rtmp_real_time
|
2015-01-25 12:09:48 -05:00
|
|
|
|
|
2013-12-15 22:09:30 -05:00
|
|
|
|
url: Final video URL.
|
2013-06-23 13:57:38 -04:00
|
|
|
|
ext: Video filename extension.
|
2013-12-16 08:13:40 -05:00
|
|
|
|
format: The video format, defaults to ext (used for --get-format)
|
|
|
|
|
player_url: SWF Player URL (used for rtmpdump).
|
2013-10-04 05:09:43 -04:00
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
The following fields are optional:
|
|
|
|
|
|
2022-03-04 09:08:55 -05:00
|
|
|
|
direct: True if a direct video file was given (must only be set by GenericIE)
|
2014-12-11 21:34:28 -05:00
|
|
|
|
alt_title: A secondary title of the video.
|
2014-03-03 06:06:28 -05:00
|
|
|
|
display_id An alternative identifier for the video, not necessarily
|
|
|
|
|
unique, but available before title. Typically, id is
|
|
|
|
|
something like "4234987", title "Dancing naked mole rats",
|
|
|
|
|
and display_id "dancing-naked-mole-rats"
|
2014-06-07 09:33:45 -04:00
|
|
|
|
thumbnails: A list of dictionaries, with the following entries:
|
2015-01-24 20:38:47 -05:00
|
|
|
|
* "id" (optional, string) - Thumbnail format ID
|
2014-06-07 09:33:45 -04:00
|
|
|
|
* "url"
|
2015-01-24 20:38:47 -05:00
|
|
|
|
* "preference" (optional, int) - quality of the image
|
2014-06-07 09:33:45 -04:00
|
|
|
|
* "width" (optional, int)
|
|
|
|
|
* "height" (optional, int)
|
2019-07-17 11:47:53 -04:00
|
|
|
|
* "resolution" (optional, string "{width}x{height}",
|
2014-06-07 09:33:45 -04:00
|
|
|
|
deprecated)
|
2016-07-08 16:24:36 -04:00
|
|
|
|
* "filesize" (optional, int)
|
2022-02-11 13:00:48 -05:00
|
|
|
|
* "http_headers" (dict) - HTTP headers for the request
|
2013-06-23 13:57:38 -04:00
|
|
|
|
thumbnail: Full URL to a video thumbnail image.
|
2014-12-11 21:34:28 -05:00
|
|
|
|
description: Full video description.
|
2013-06-23 13:57:38 -04:00
|
|
|
|
uploader: Full name of the video uploader.
|
2016-03-02 12:06:39 -05:00
|
|
|
|
license: License name the video is licensed under.
|
2016-05-02 11:31:35 -04:00
|
|
|
|
creator: The creator of the video.
|
2021-03-14 19:22:06 -04:00
|
|
|
|
timestamp: UNIX timestamp of the moment the video was uploaded
|
2022-03-23 21:30:22 -04:00
|
|
|
|
upload_date: Video upload date in UTC (YYYYMMDD).
|
2022-01-07 06:03:02 -05:00
|
|
|
|
If not explicitly set, calculated from timestamp
|
|
|
|
|
release_timestamp: UNIX timestamp of the moment the video was released.
|
|
|
|
|
If it is not clear whether to use timestamp or this, use the former
|
2022-03-23 21:30:22 -04:00
|
|
|
|
release_date: The date (YYYYMMDD) when the video was released in UTC.
|
2022-01-07 06:03:02 -05:00
|
|
|
|
If not explicitly set, calculated from release_timestamp
|
|
|
|
|
modified_timestamp: UNIX timestamp of the moment the video was last modified.
|
2022-03-23 21:30:22 -04:00
|
|
|
|
modified_date: The date (YYYYMMDD) when the video was last modified in UTC.
|
2022-01-07 06:03:02 -05:00
|
|
|
|
If not explicitly set, calculated from modified_timestamp
|
2013-06-23 13:57:38 -04:00
|
|
|
|
uploader_id: Nickname or id of the video uploader.
|
2016-03-02 12:31:24 -05:00
|
|
|
|
uploader_url: Full URL to a personal webpage of the video uploader.
|
2018-09-14 14:23:36 -04:00
|
|
|
|
channel: Full name of the channel the video is uploaded on.
|
2018-09-14 14:53:01 -04:00
|
|
|
|
Note that channel fields may or may not repeat uploader
|
2018-09-14 14:23:36 -04:00
|
|
|
|
fields. This depends on a particular extractor.
|
|
|
|
|
channel_id: Id of the channel.
|
|
|
|
|
channel_url: Full URL to a channel webpage.
|
2022-01-21 01:04:36 -05:00
|
|
|
|
channel_follower_count: Number of followers of the channel.
|
2014-08-26 19:44:47 -04:00
|
|
|
|
location: Physical location where the video was filmed.
|
2015-02-15 12:03:41 -05:00
|
|
|
|
subtitles: The available subtitles as a dictionary in the format
|
2016-12-24 12:50:50 -05:00
|
|
|
|
{tag: subformats}. "tag" is usually a language code, and
|
|
|
|
|
"subformats" is a list sorted from lower to higher
|
|
|
|
|
preference, each element is a dictionary with the "ext"
|
|
|
|
|
entry and one of:
|
2015-02-15 12:03:41 -05:00
|
|
|
|
* "data": The subtitles file contents
|
2015-07-23 13:37:45 -04:00
|
|
|
|
* "url": A URL pointing to the subtitles file
|
2021-05-12 15:37:58 -04:00
|
|
|
|
It can optionally also have:
|
|
|
|
|
* "name": Name or description of the subtitles
|
2022-03-04 09:08:55 -05:00
|
|
|
|
* "http_headers": A dictionary of additional HTTP headers
|
2022-02-11 13:00:48 -05:00
|
|
|
|
to add to the request.
|
2015-10-04 10:33:42 -04:00
|
|
|
|
"ext" will be calculated from URL if missing
|
2021-03-18 11:34:09 -04:00
|
|
|
|
automatic_captions: Like 'subtitles'; contains automatically generated
|
|
|
|
|
captions instead of normal subtitles
|
2015-12-03 09:55:02 -05:00
|
|
|
|
duration: Length of the video in seconds, as an integer or float.
|
2013-06-29 10:32:28 -04:00
|
|
|
|
view_count: How many users have watched the video on the platform.
|
2022-10-07 03:00:40 -04:00
|
|
|
|
concurrent_view_count: How many users are currently watching the video on the platform.
|
2013-12-05 12:29:07 -05:00
|
|
|
|
like_count: Number of positive ratings of the video
|
|
|
|
|
dislike_count: Number of negative ratings of the video
|
2015-10-17 23:34:54 -04:00
|
|
|
|
repost_count: Number of reposts of the video
|
2015-02-11 12:39:31 -05:00
|
|
|
|
average_rating: Average rating give by users, the scale used depends on the webpage
|
2013-12-05 12:29:07 -05:00
|
|
|
|
comment_count: Number of comments on the video
|
2015-01-09 17:59:18 -05:00
|
|
|
|
comments: A list of comments, each with one or more of the following
|
|
|
|
|
properties (all but one of text or html optional):
|
|
|
|
|
* "author" - human-readable name of the comment author
|
|
|
|
|
* "author_id" - user ID of the comment author
|
2021-03-14 18:41:11 -04:00
|
|
|
|
* "author_thumbnail" - The thumbnail of the comment author
|
2015-01-09 17:59:18 -05:00
|
|
|
|
* "id" - Comment ID
|
|
|
|
|
* "html" - Comment as HTML
|
|
|
|
|
* "text" - Plain text of the comment
|
|
|
|
|
* "timestamp" - UNIX timestamp of comment
|
|
|
|
|
* "parent" - ID of the comment this one is replying to.
|
|
|
|
|
Set to "root" to indicate that this is a
|
|
|
|
|
comment to the original video.
|
2021-03-14 18:41:11 -04:00
|
|
|
|
* "like_count" - Number of positive ratings of the comment
|
|
|
|
|
* "dislike_count" - Number of negative ratings of the comment
|
|
|
|
|
* "is_favorited" - Whether the comment is marked as
|
|
|
|
|
favorite by the video uploader
|
|
|
|
|
* "author_is_uploader" - Whether the comment is made by
|
|
|
|
|
the video uploader
|
2013-10-06 00:06:30 -04:00
|
|
|
|
age_limit: Age restriction for the video, as an integer (years)
|
2021-02-24 13:45:56 -05:00
|
|
|
|
webpage_url: The URL to the video webpage, if given to yt-dlp it
|
2013-11-03 06:11:13 -05:00
|
|
|
|
should allow to get the same result again. (It will be set
|
|
|
|
|
by YoutubeDL if it's missing)
|
2014-05-15 06:41:42 -04:00
|
|
|
|
categories: A list of categories that the video falls in, for example
|
|
|
|
|
["Sports", "Berlin"]
|
2015-07-28 17:43:03 -04:00
|
|
|
|
tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"]
|
2021-06-13 12:06:47 -04:00
|
|
|
|
cast: A list of the video cast
|
2014-09-19 03:57:53 -04:00
|
|
|
|
is_live: True, False, or None (=unknown). Whether this video is a
|
|
|
|
|
live stream that goes on instead of a fixed-length video.
|
2021-02-21 16:41:24 -05:00
|
|
|
|
was_live: True, False, or None (=unknown). Whether this video was
|
|
|
|
|
originally a live stream.
|
2022-08-01 18:10:47 -04:00
|
|
|
|
live_status: None (=unknown), 'is_live', 'is_upcoming', 'was_live', 'not_live',
|
2022-07-31 16:12:04 -04:00
|
|
|
|
or 'post_live' (was live, but VOD is not yet processed)
|
2021-07-21 11:14:18 -04:00
|
|
|
|
If absent, automatically set from is_live, was_live
|
2015-07-20 15:10:28 -04:00
|
|
|
|
start_time: Time in seconds where the reproduction should start, as
|
2015-07-23 13:37:45 -04:00
|
|
|
|
specified in the URL.
|
2015-07-23 07:20:21 -04:00
|
|
|
|
end_time: Time in seconds where the reproduction should end, as
|
2015-07-23 13:37:45 -04:00
|
|
|
|
specified in the URL.
|
2016-05-05 16:40:19 -04:00
|
|
|
|
chapters: A list of dictionaries, with the following entries:
|
|
|
|
|
* "start_time" - The start time of the chapter in seconds
|
|
|
|
|
* "end_time" - The end time of the chapter in seconds
|
|
|
|
|
* "title" (optional, string)
|
2021-01-14 04:07:23 -05:00
|
|
|
|
playable_in_embed: Whether this video is allowed to play in embedded
|
|
|
|
|
players on other sites. Can be True (=always allowed),
|
|
|
|
|
False (=never allowed), None (=unknown), or a string
|
2022-08-14 08:04:13 -04:00
|
|
|
|
specifying the criteria for embedability; e.g. 'whitelist'
|
2021-03-21 17:23:34 -04:00
|
|
|
|
availability: Under what condition the video is available. One of
|
|
|
|
|
'private', 'premium_only', 'subscriber_only', 'needs_auth',
|
|
|
|
|
'unlisted' or 'public'. Use 'InfoExtractor._availability'
|
|
|
|
|
to set it
|
2022-07-13 05:33:05 -04:00
|
|
|
|
_old_archive_ids: A list of old archive ids needed for backward compatibility
|
2021-02-28 09:56:08 -05:00
|
|
|
|
__post_extractor: A function to be called just before the metadata is
|
|
|
|
|
written to either disk, logger or console. The function
|
|
|
|
|
must return a dict which will be added to the info_dict.
|
|
|
|
|
This is usefull for additional information that is
|
|
|
|
|
time-consuming to extract. Note that the fields thus
|
|
|
|
|
extracted will not be available to output template and
|
|
|
|
|
match_filter. So, only "comments" and "comment_count" are
|
|
|
|
|
currently allowed to be extracted via this method.
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2015-12-30 16:10:44 -05:00
|
|
|
|
The following fields should only be used when the video belongs to some logical
|
|
|
|
|
chapter or section:
|
|
|
|
|
|
|
|
|
|
chapter: Name or title of the chapter the video belongs to.
|
2016-01-01 09:26:56 -05:00
|
|
|
|
chapter_number: Number of the chapter the video belongs to, as an integer.
|
|
|
|
|
chapter_id: Id of the chapter the video belongs to, as a unicode string.
|
2015-12-30 16:10:44 -05:00
|
|
|
|
|
|
|
|
|
The following fields should only be used when the video is an episode of some
|
2016-10-16 07:37:17 -04:00
|
|
|
|
series, programme or podcast:
|
2015-12-30 16:10:44 -05:00
|
|
|
|
|
|
|
|
|
series: Title of the series or programme the video episode belongs to.
|
2021-11-13 13:19:14 -05:00
|
|
|
|
series_id: Id of the series or programme the video episode belongs to, as a unicode string.
|
2015-12-30 16:10:44 -05:00
|
|
|
|
season: Title of the season the video episode belongs to.
|
2016-01-01 09:26:56 -05:00
|
|
|
|
season_number: Number of the season the video episode belongs to, as an integer.
|
|
|
|
|
season_id: Id of the season the video episode belongs to, as a unicode string.
|
2015-12-30 16:10:44 -05:00
|
|
|
|
episode: Title of the video episode. Unlike mandatory video title field,
|
|
|
|
|
this field should denote the exact title of the video episode
|
|
|
|
|
without any kind of decoration.
|
2016-01-01 09:26:56 -05:00
|
|
|
|
episode_number: Number of the video episode within a season, as an integer.
|
|
|
|
|
episode_id: Id of the video episode, as a unicode string.
|
2015-12-30 16:10:44 -05:00
|
|
|
|
|
2016-04-06 16:53:53 -04:00
|
|
|
|
The following fields should only be used when the media is a track or a part of
|
|
|
|
|
a music album:
|
|
|
|
|
|
|
|
|
|
track: Title of the track.
|
|
|
|
|
track_number: Number of the track within an album or a disc, as an integer.
|
|
|
|
|
track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
|
|
|
|
|
as a unicode string.
|
|
|
|
|
artist: Artist(s) of the track.
|
|
|
|
|
genre: Genre(s) of the track.
|
|
|
|
|
album: Title of the album the track belongs to.
|
|
|
|
|
album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
|
|
|
|
|
album_artist: List of all artists appeared on the album (e.g.
|
|
|
|
|
"Ash Borer / Fell Voices" or "Various Artists", useful for splits
|
|
|
|
|
and compilations).
|
|
|
|
|
disc_number: Number of the disc or other physical medium the track belongs to,
|
|
|
|
|
as an integer.
|
|
|
|
|
release_year: Year (YYYY) when the album was released.
|
2022-01-10 16:36:05 -05:00
|
|
|
|
composer: Composer of the piece
|
2016-04-06 16:53:53 -04:00
|
|
|
|
|
2022-06-21 17:02:14 -04:00
|
|
|
|
The following fields should only be set for clips that should be cut from the original video:
|
|
|
|
|
|
|
|
|
|
section_start: Start time of the section in seconds
|
|
|
|
|
section_end: End time of the section in seconds
|
|
|
|
|
|
2022-07-12 11:16:45 -04:00
|
|
|
|
The following fields should only be set for storyboards:
|
|
|
|
|
rows: Number of rows in each storyboard fragment, as an integer
|
|
|
|
|
columns: Number of columns in each storyboard fragment, as an integer
|
|
|
|
|
|
2013-10-04 04:40:42 -04:00
|
|
|
|
Unless mentioned otherwise, the fields should be Unicode strings.
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2014-10-03 14:17:10 -04:00
|
|
|
|
Unless mentioned otherwise, None is equivalent to absence of information.
|
|
|
|
|
|
2014-11-20 10:47:59 -05:00
|
|
|
|
|
|
|
|
|
_type "playlist" indicates multiple videos.
|
2014-12-06 08:02:19 -05:00
|
|
|
|
There must be a key "entries", which is a list, an iterable, or a PagedList
|
|
|
|
|
object, each element of which is a valid dictionary by this specification.
|
2014-11-20 10:47:59 -05:00
|
|
|
|
|
2022-06-26 20:50:06 -04:00
|
|
|
|
Additionally, playlists can have "id", "title", and any other relevant
|
2021-02-02 11:21:32 -05:00
|
|
|
|
attributes with the same semantics as videos (see above).
|
2014-11-20 10:47:59 -05:00
|
|
|
|
|
2022-01-07 06:03:02 -05:00
|
|
|
|
It can also have the following optional fields:
|
|
|
|
|
|
|
|
|
|
playlist_count: The total number of videos in a playlist. If not given,
|
|
|
|
|
YoutubeDL tries to calculate it from "entries"
|
|
|
|
|
|
2014-11-20 10:47:59 -05:00
|
|
|
|
|
|
|
|
|
_type "multi_video" indicates that there are multiple videos that
|
|
|
|
|
form a single show, for examples multiple acts of an opera or TV episode.
|
|
|
|
|
It must have an entries key like a playlist and contain all the keys
|
|
|
|
|
required for a video at the same time.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
_type "url" indicates that the video must be extracted from another
|
|
|
|
|
location, possibly by a different extractor. Its only required key is:
|
|
|
|
|
"url" - the next URL to extract.
|
2014-12-09 04:58:06 -05:00
|
|
|
|
The key "ie_key" can be set to the class name (minus the trailing "IE",
|
|
|
|
|
e.g. "Youtube") if the extractor class is known in advance.
|
|
|
|
|
Additionally, the dictionary may have any properties of the resolved entity
|
|
|
|
|
known in advance, for example "title" if the title of the referred video is
|
2014-11-20 10:47:59 -05:00
|
|
|
|
known ahead of time.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
_type "url_transparent" entities have the same specification as "url", but
|
|
|
|
|
indicate that the given additional information is more precise than the one
|
|
|
|
|
associated with the resolved URL.
|
|
|
|
|
This is useful when a site employs a video service that hosts the video and
|
|
|
|
|
its technical metadata, but that video service does not embed a useful
|
|
|
|
|
title, description etc.
|
|
|
|
|
|
|
|
|
|
|
2022-07-31 21:22:03 -04:00
|
|
|
|
Subclasses of this should also be added to the list of extractors and
|
|
|
|
|
should define a _VALID_URL regexp and, re-define the _real_extract() and
|
|
|
|
|
(optionally) _real_initialize() methods.
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2021-09-17 14:23:55 -04:00
|
|
|
|
Subclasses may also override suitable() if necessary, but ensure the function
|
|
|
|
|
signature is preserved and that this function imports everything it needs
|
2022-03-18 16:53:33 -04:00
|
|
|
|
(except other extractors), so that lazy_extractors works correctly.
|
|
|
|
|
|
2022-07-31 21:22:03 -04:00
|
|
|
|
Subclasses can define a list of _EMBED_REGEX, which will be searched for in
|
|
|
|
|
the HTML of Generic webpages. It may also override _extract_embed_urls
|
|
|
|
|
or _extract_from_webpage as necessary. While these are normally classmethods,
|
|
|
|
|
_extract_from_webpage is allowed to be an instance method.
|
|
|
|
|
|
|
|
|
|
_extract_from_webpage may raise self.StopExtraction() to stop further
|
|
|
|
|
processing of the webpage and obtain exclusive rights to it. This is useful
|
2022-08-14 08:04:13 -04:00
|
|
|
|
when the extractor cannot reliably be matched using just the URL,
|
|
|
|
|
e.g. invidious/peertube instances
|
2022-07-31 21:22:03 -04:00
|
|
|
|
|
|
|
|
|
Embed-only extractors can be defined by setting _VALID_URL = False.
|
|
|
|
|
|
2022-03-18 16:53:33 -04:00
|
|
|
|
To support username + password (or netrc) login, the extractor must define a
|
|
|
|
|
_NETRC_MACHINE and re-define _perform_login(username, password) and
|
|
|
|
|
(optionally) _initialize_pre_login() methods. The _perform_login method will
|
|
|
|
|
be called between _initialize_pre_login and _real_initialize if credentials
|
|
|
|
|
are passed by the user. In cases where it is necessary to have the login
|
|
|
|
|
process as part of the extraction rather than initialization, _perform_login
|
|
|
|
|
can be left undefined.
|
2021-09-17 14:23:55 -04:00
|
|
|
|
|
2017-02-18 15:53:23 -05:00
|
|
|
|
_GEO_BYPASS attribute may be set to False in order to disable
|
2017-02-04 06:49:58 -05:00
|
|
|
|
geo restriction bypass mechanisms for a particular extractor.
|
|
|
|
|
Though it won't disable explicit geo restriction bypass based on
|
2018-05-19 12:53:24 -04:00
|
|
|
|
country code provided with geo_bypass_country.
|
2017-02-18 15:53:23 -05:00
|
|
|
|
|
|
|
|
|
_GEO_COUNTRIES attribute may contain a list of presumably geo unrestricted
|
|
|
|
|
countries for this extractor. One of these countries will be used by
|
|
|
|
|
geo restriction bypass mechanism right away in order to bypass
|
2018-05-19 12:53:24 -04:00
|
|
|
|
geo restriction, of course, if the mechanism is not disabled.
|
2017-02-04 06:49:58 -05:00
|
|
|
|
|
2018-05-01 20:18:01 -04:00
|
|
|
|
_GEO_IP_BLOCKS attribute may contain a list of presumably geo unrestricted
|
|
|
|
|
IP blocks in CIDR notation for this extractor. One of these IP blocks
|
|
|
|
|
will be used by geo restriction bypass mechanism similarly
|
2018-05-19 12:53:24 -04:00
|
|
|
|
to _GEO_COUNTRIES.
|
2017-02-20 11:21:15 -05:00
|
|
|
|
|
2022-08-23 20:12:16 -04:00
|
|
|
|
The _ENABLED attribute should be set to False for IEs that
|
|
|
|
|
are disabled by default and must be explicitly enabled.
|
|
|
|
|
|
2021-09-17 14:23:55 -04:00
|
|
|
|
The _WORKING attribute should be set to False for broken IEs
|
2013-06-23 13:57:38 -04:00
|
|
|
|
in order to warn the users and skip the tests.
|
|
|
|
|
"""
|
|
|
|
|
|
|
|
|
|
_ready = False
|
|
|
|
|
_downloader = None
|
2017-02-04 06:49:58 -05:00
|
|
|
|
_x_forwarded_for_ip = None
|
2017-02-18 15:53:23 -05:00
|
|
|
|
_GEO_BYPASS = True
|
|
|
|
|
_GEO_COUNTRIES = None
|
2018-05-01 20:18:01 -04:00
|
|
|
|
_GEO_IP_BLOCKS = None
|
2013-06-23 13:57:38 -04:00
|
|
|
|
_WORKING = True
|
2022-08-23 20:12:16 -04:00
|
|
|
|
_ENABLED = True
|
2022-03-18 16:53:33 -04:00
|
|
|
|
_NETRC_MACHINE = None
|
2022-03-23 21:23:11 -04:00
|
|
|
|
IE_DESC = None
|
2022-05-09 00:32:17 -04:00
|
|
|
|
SEARCH_KEY = None
|
2022-07-31 21:22:03 -04:00
|
|
|
|
_VALID_URL = None
|
|
|
|
|
_EMBED_REGEX = []
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2022-05-09 00:32:17 -04:00
|
|
|
|
def _login_hint(self, method=NO_DEFAULT, netrc=None):
|
|
|
|
|
password_hint = f'--username and --password, or --netrc ({netrc or self._NETRC_MACHINE}) to provide account credentials'
|
|
|
|
|
return {
|
|
|
|
|
None: '',
|
|
|
|
|
'any': f'Use --cookies, --cookies-from-browser, {password_hint}',
|
|
|
|
|
'password': f'Use {password_hint}',
|
|
|
|
|
'cookies': (
|
|
|
|
|
'Use --cookies-from-browser or --cookies for the authentication. '
|
2022-09-07 08:05:45 -04:00
|
|
|
|
'See https://github.com/yt-dlp/yt-dlp/wiki/FAQ#how-do-i-pass-cookies-to-yt-dlp for how to manually pass cookies'),
|
2022-05-09 00:32:17 -04:00
|
|
|
|
}[method if method is not NO_DEFAULT else 'any' if self.supports_login() else 'cookies']
|
2021-05-19 09:41:44 -04:00
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
def __init__(self, downloader=None):
|
2021-10-22 16:37:20 -04:00
|
|
|
|
"""Constructor. Receives an optional downloader (a YoutubeDL instance).
|
|
|
|
|
If a downloader is not passed during initialization,
|
|
|
|
|
it must be set using "set_downloader()" before "extract()" is called"""
|
2013-06-23 13:57:38 -04:00
|
|
|
|
self._ready = False
|
2017-02-04 06:49:58 -05:00
|
|
|
|
self._x_forwarded_for_ip = None
|
2021-08-05 18:17:11 -04:00
|
|
|
|
self._printed_messages = set()
|
2013-06-23 13:57:38 -04:00
|
|
|
|
self.set_downloader(downloader)
|
|
|
|
|
|
|
|
|
|
@classmethod
|
2021-08-18 21:41:24 -04:00
|
|
|
|
def _match_valid_url(cls, url):
|
2022-07-31 21:22:03 -04:00
|
|
|
|
if cls._VALID_URL is False:
|
|
|
|
|
return None
|
2013-08-20 22:06:46 -04:00
|
|
|
|
# This does not use has/getattr intentionally - we want to know whether
|
|
|
|
|
# we have cached the regexp for *this* class, whereas getattr would also
|
|
|
|
|
# match the superclass
|
|
|
|
|
if '_VALID_URL_RE' not in cls.__dict__:
|
|
|
|
|
cls._VALID_URL_RE = re.compile(cls._VALID_URL)
|
2021-08-18 21:41:24 -04:00
|
|
|
|
return cls._VALID_URL_RE.match(url)
|
|
|
|
|
|
|
|
|
|
@classmethod
|
|
|
|
|
def suitable(cls, url):
|
|
|
|
|
"""Receives a URL and returns True if suitable for this IE."""
|
2021-08-22 15:19:23 -04:00
|
|
|
|
# This function must import everything it needs (except other extractors),
|
|
|
|
|
# so that lazy_extractors works correctly
|
2021-08-18 21:41:24 -04:00
|
|
|
|
return cls._match_valid_url(url) is not None
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2014-09-28 03:31:58 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def _match_id(cls, url):
|
2021-08-18 21:41:24 -04:00
|
|
|
|
return cls._match_valid_url(url).group('id')
|
2014-09-28 03:31:58 -04:00
|
|
|
|
|
2021-08-18 21:49:23 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def get_temp_id(cls, url):
|
|
|
|
|
try:
|
|
|
|
|
return cls._match_id(url)
|
|
|
|
|
except (IndexError, AttributeError):
|
|
|
|
|
return None
|
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def working(cls):
|
|
|
|
|
"""Getter method for _WORKING."""
|
|
|
|
|
return cls._WORKING
|
|
|
|
|
|
2022-03-18 16:53:33 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def supports_login(cls):
|
|
|
|
|
return bool(cls._NETRC_MACHINE)
|
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
def initialize(self):
|
|
|
|
|
"""Initializes an instance (authentication, etc)."""
|
2021-08-05 18:17:11 -04:00
|
|
|
|
self._printed_messages = set()
|
2018-05-01 20:18:01 -04:00
|
|
|
|
self._initialize_geo_bypass({
|
|
|
|
|
'countries': self._GEO_COUNTRIES,
|
|
|
|
|
'ip_blocks': self._GEO_IP_BLOCKS,
|
|
|
|
|
})
|
2017-02-18 15:53:23 -05:00
|
|
|
|
if not self._ready:
|
2022-03-18 16:53:33 -04:00
|
|
|
|
self._initialize_pre_login()
|
|
|
|
|
if self.supports_login():
|
|
|
|
|
username, password = self._get_login_info()
|
|
|
|
|
if username:
|
|
|
|
|
self._perform_login(username, password)
|
|
|
|
|
elif self.get_param('username') and False not in (self.IE_DESC, self._NETRC_MACHINE):
|
2022-05-09 00:32:17 -04:00
|
|
|
|
self.report_warning(f'Login with password is not supported for this website. {self._login_hint("cookies")}')
|
2017-02-18 15:53:23 -05:00
|
|
|
|
self._real_initialize()
|
|
|
|
|
self._ready = True
|
|
|
|
|
|
2018-05-01 20:18:01 -04:00
|
|
|
|
def _initialize_geo_bypass(self, geo_bypass_context):
|
2017-02-21 11:00:43 -05:00
|
|
|
|
"""
|
|
|
|
|
Initialize geo restriction bypass mechanism.
|
|
|
|
|
|
|
|
|
|
This method is used to initialize geo bypass mechanism based on faking
|
|
|
|
|
X-Forwarded-For HTTP header. A random country from provided country list
|
2017-02-21 11:05:31 -05:00
|
|
|
|
is selected and a random IP belonging to this country is generated. This
|
2017-02-21 11:00:43 -05:00
|
|
|
|
IP will be passed as X-Forwarded-For HTTP header in all subsequent
|
|
|
|
|
HTTP requests.
|
|
|
|
|
|
|
|
|
|
This method will be used for initial geo bypass mechanism initialization
|
2018-05-01 20:18:01 -04:00
|
|
|
|
during the instance initialization with _GEO_COUNTRIES and
|
|
|
|
|
_GEO_IP_BLOCKS.
|
2017-02-21 11:00:43 -05:00
|
|
|
|
|
2018-05-01 20:18:01 -04:00
|
|
|
|
You may also manually call it from extractor's code if geo bypass
|
2017-02-21 11:00:43 -05:00
|
|
|
|
information is not available beforehand (e.g. obtained during
|
2018-05-01 20:18:01 -04:00
|
|
|
|
extraction) or due to some other reason. In this case you should pass
|
|
|
|
|
this information in geo bypass context passed as first argument. It may
|
|
|
|
|
contain following fields:
|
|
|
|
|
|
|
|
|
|
countries: List of geo unrestricted countries (similar
|
|
|
|
|
to _GEO_COUNTRIES)
|
|
|
|
|
ip_blocks: List of geo unrestricted IP blocks in CIDR notation
|
|
|
|
|
(similar to _GEO_IP_BLOCKS)
|
|
|
|
|
|
2017-02-21 11:00:43 -05:00
|
|
|
|
"""
|
2017-02-04 06:49:58 -05:00
|
|
|
|
if not self._x_forwarded_for_ip:
|
2018-05-01 20:18:01 -04:00
|
|
|
|
|
|
|
|
|
# Geo bypass mechanism is explicitly disabled by user
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if not self.get_param('geo_bypass', True):
|
2018-05-01 20:18:01 -04:00
|
|
|
|
return
|
|
|
|
|
|
|
|
|
|
if not geo_bypass_context:
|
|
|
|
|
geo_bypass_context = {}
|
|
|
|
|
|
|
|
|
|
# Backward compatibility: previously _initialize_geo_bypass
|
|
|
|
|
# expected a list of countries, some 3rd party code may still use
|
|
|
|
|
# it this way
|
|
|
|
|
if isinstance(geo_bypass_context, (list, tuple)):
|
|
|
|
|
geo_bypass_context = {
|
|
|
|
|
'countries': geo_bypass_context,
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
# The whole point of geo bypass mechanism is to fake IP
|
|
|
|
|
# as X-Forwarded-For HTTP header based on some IP block or
|
|
|
|
|
# country code.
|
|
|
|
|
|
|
|
|
|
# Path 1: bypassing based on IP block in CIDR notation
|
|
|
|
|
|
|
|
|
|
# Explicit IP block specified by user, use it right away
|
|
|
|
|
# regardless of whether extractor is geo bypassable or not
|
2021-05-17 08:23:08 -04:00
|
|
|
|
ip_block = self.get_param('geo_bypass_ip_block', None)
|
2018-05-01 20:18:01 -04:00
|
|
|
|
|
|
|
|
|
# Otherwise use random IP block from geo bypass context but only
|
|
|
|
|
# if extractor is known as geo bypassable
|
|
|
|
|
if not ip_block:
|
|
|
|
|
ip_blocks = geo_bypass_context.get('ip_blocks')
|
|
|
|
|
if self._GEO_BYPASS and ip_blocks:
|
|
|
|
|
ip_block = random.choice(ip_blocks)
|
|
|
|
|
|
|
|
|
|
if ip_block:
|
|
|
|
|
self._x_forwarded_for_ip = GeoUtils.random_ipv4(ip_block)
|
2022-05-26 19:06:23 -04:00
|
|
|
|
self.write_debug(f'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For')
|
2018-05-01 20:18:01 -04:00
|
|
|
|
return
|
|
|
|
|
|
|
|
|
|
# Path 2: bypassing based on country code
|
|
|
|
|
|
|
|
|
|
# Explicit country code specified by user, use it right away
|
|
|
|
|
# regardless of whether extractor is geo bypassable or not
|
2021-05-17 08:23:08 -04:00
|
|
|
|
country = self.get_param('geo_bypass_country', None)
|
2018-05-01 20:18:01 -04:00
|
|
|
|
|
|
|
|
|
# Otherwise use random country code from geo bypass context but
|
|
|
|
|
# only if extractor is known as geo bypassable
|
|
|
|
|
if not country:
|
|
|
|
|
countries = geo_bypass_context.get('countries')
|
|
|
|
|
if self._GEO_BYPASS and countries:
|
|
|
|
|
country = random.choice(countries)
|
|
|
|
|
|
|
|
|
|
if country:
|
|
|
|
|
self._x_forwarded_for_ip = GeoUtils.random_ipv4(country)
|
2021-05-14 03:45:29 -04:00
|
|
|
|
self._downloader.write_debug(
|
2022-04-11 11:10:28 -04:00
|
|
|
|
f'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For')
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
|
|
|
|
def extract(self, url):
|
|
|
|
|
"""Extracts URL information and returns it in list of dicts."""
|
2015-02-09 19:13:57 -05:00
|
|
|
|
try:
|
2017-02-04 06:49:58 -05:00
|
|
|
|
for _ in range(2):
|
|
|
|
|
try:
|
|
|
|
|
self.initialize()
|
2021-05-17 08:23:08 -04:00
|
|
|
|
self.write_debug('Extracting URL: %s' % url)
|
2017-02-04 09:06:07 -05:00
|
|
|
|
ie_result = self._real_extract(url)
|
2021-05-18 14:20:59 -04:00
|
|
|
|
if ie_result is None:
|
|
|
|
|
return None
|
2017-02-04 09:06:07 -05:00
|
|
|
|
if self._x_forwarded_for_ip:
|
|
|
|
|
ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip
|
2022-07-19 05:57:11 -04:00
|
|
|
|
subtitles = ie_result.get('subtitles') or {}
|
|
|
|
|
if 'no-live-chat' in self.get_param('compat_opts'):
|
|
|
|
|
for lang in ('live_chat', 'comments', 'danmaku'):
|
|
|
|
|
subtitles.pop(lang, None)
|
2017-02-04 09:06:07 -05:00
|
|
|
|
return ie_result
|
2017-02-04 06:49:58 -05:00
|
|
|
|
except GeoRestrictedError as e:
|
2017-02-18 15:53:23 -05:00
|
|
|
|
if self.__maybe_fake_ip_and_retry(e.countries):
|
|
|
|
|
continue
|
2017-02-04 06:49:58 -05:00
|
|
|
|
raise
|
2021-10-26 10:47:29 -04:00
|
|
|
|
except UnsupportedError:
|
|
|
|
|
raise
|
2021-08-18 21:49:23 -04:00
|
|
|
|
except ExtractorError as e:
|
2021-10-26 10:47:29 -04:00
|
|
|
|
kwargs = {
|
|
|
|
|
'video_id': e.video_id or self.get_temp_id(url),
|
|
|
|
|
'ie': self.IE_NAME,
|
2021-12-22 20:42:26 -05:00
|
|
|
|
'tb': e.traceback or sys.exc_info()[2],
|
2021-10-26 10:47:29 -04:00
|
|
|
|
'expected': e.expected,
|
|
|
|
|
'cause': e.cause
|
|
|
|
|
}
|
|
|
|
|
if hasattr(e, 'countries'):
|
|
|
|
|
kwargs['countries'] = e.countries
|
2022-03-04 09:07:43 -05:00
|
|
|
|
raise type(e)(e.orig_msg, **kwargs)
|
2022-06-24 04:10:17 -04:00
|
|
|
|
except http.client.IncompleteRead as e:
|
2021-08-18 21:49:23 -04:00
|
|
|
|
raise ExtractorError('A network error has occurred.', cause=e, expected=True, video_id=self.get_temp_id(url))
|
2015-02-10 09:55:51 -05:00
|
|
|
|
except (KeyError, StopIteration) as e:
|
2021-08-18 21:49:23 -04:00
|
|
|
|
raise ExtractorError('An extractor error has occurred.', cause=e, video_id=self.get_temp_id(url))
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2017-02-18 15:53:23 -05:00
|
|
|
|
def __maybe_fake_ip_and_retry(self, countries):
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if (not self.get_param('geo_bypass_country', None)
|
2019-05-10 16:56:22 -04:00
|
|
|
|
and self._GEO_BYPASS
|
2021-05-17 08:23:08 -04:00
|
|
|
|
and self.get_param('geo_bypass', True)
|
2019-05-10 16:56:22 -04:00
|
|
|
|
and not self._x_forwarded_for_ip
|
|
|
|
|
and countries):
|
2017-02-21 11:14:33 -05:00
|
|
|
|
country_code = random.choice(countries)
|
|
|
|
|
self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code)
|
2017-02-18 15:53:23 -05:00
|
|
|
|
if self._x_forwarded_for_ip:
|
|
|
|
|
self.report_warning(
|
2017-02-21 11:14:33 -05:00
|
|
|
|
'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.'
|
|
|
|
|
% (self._x_forwarded_for_ip, country_code.upper()))
|
2017-02-18 15:53:23 -05:00
|
|
|
|
return True
|
|
|
|
|
return False
|
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
def set_downloader(self, downloader):
|
2022-03-04 09:08:55 -05:00
|
|
|
|
"""Sets a YoutubeDL instance as the downloader for this IE."""
|
2013-06-23 13:57:38 -04:00
|
|
|
|
self._downloader = downloader
|
|
|
|
|
|
2022-06-23 00:14:22 -04:00
|
|
|
|
@property
|
|
|
|
|
def cache(self):
|
|
|
|
|
return self._downloader.cache
|
|
|
|
|
|
|
|
|
|
@property
|
|
|
|
|
def cookiejar(self):
|
|
|
|
|
return self._downloader.cookiejar
|
|
|
|
|
|
2022-03-18 16:53:33 -04:00
|
|
|
|
def _initialize_pre_login(self):
|
2022-06-26 20:50:06 -04:00
|
|
|
|
""" Initialization before login. Redefine in subclasses."""
|
2022-03-18 16:53:33 -04:00
|
|
|
|
pass
|
|
|
|
|
|
|
|
|
|
def _perform_login(self, username, password):
|
|
|
|
|
""" Login with username and password. Redefine in subclasses."""
|
|
|
|
|
pass
|
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
def _real_initialize(self):
|
|
|
|
|
"""Real initialization process. Redefine in subclasses."""
|
|
|
|
|
pass
|
|
|
|
|
|
|
|
|
|
def _real_extract(self, url):
|
|
|
|
|
"""Real extraction process. Redefine in subclasses."""
|
2022-03-04 09:08:55 -05:00
|
|
|
|
raise NotImplementedError('This method must be implemented by subclasses')
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2013-07-08 09:14:27 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def ie_key(cls):
|
|
|
|
|
"""A string for getting the InfoExtractor with get_info_extractor"""
|
2021-08-22 15:19:23 -04:00
|
|
|
|
return cls.__name__[:-2]
|
2013-07-08 09:14:27 -04:00
|
|
|
|
|
2022-05-11 11:54:44 -04:00
|
|
|
|
@classproperty
|
|
|
|
|
def IE_NAME(cls):
|
|
|
|
|
return cls.__name__[:-2]
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2018-06-17 17:01:48 -04:00
|
|
|
|
@staticmethod
|
|
|
|
|
def __can_accept_status_code(err, expected_status):
|
2022-06-24 04:10:17 -04:00
|
|
|
|
assert isinstance(err, urllib.error.HTTPError)
|
2018-06-17 17:01:48 -04:00
|
|
|
|
if expected_status is None:
|
|
|
|
|
return False
|
|
|
|
|
elif callable(expected_status):
|
|
|
|
|
return expected_status(err.code) is True
|
|
|
|
|
else:
|
2021-07-10 17:59:44 -04:00
|
|
|
|
return err.code in variadic(expected_status)
|
2018-06-17 17:01:48 -04:00
|
|
|
|
|
2022-06-25 10:11:22 -04:00
|
|
|
|
def _create_request(self, url_or_request, data=None, headers=None, query=None):
|
2022-06-24 04:10:17 -04:00
|
|
|
|
if isinstance(url_or_request, urllib.request.Request):
|
2022-06-02 20:45:54 -04:00
|
|
|
|
return update_Request(url_or_request, data=data, headers=headers, query=query)
|
|
|
|
|
if query:
|
|
|
|
|
url_or_request = update_url_query(url_or_request, query)
|
2022-06-25 10:11:22 -04:00
|
|
|
|
return sanitized_Request(url_or_request, data, headers or {})
|
2022-05-31 16:33:22 -04:00
|
|
|
|
|
2022-06-25 10:11:22 -04:00
|
|
|
|
def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers=None, query=None, expected_status=None):
|
2018-06-17 17:01:48 -04:00
|
|
|
|
"""
|
|
|
|
|
Return the response handle.
|
|
|
|
|
|
|
|
|
|
See _download_webpage docstring for arguments specification.
|
|
|
|
|
"""
|
2021-02-27 07:41:23 -05:00
|
|
|
|
if not self._downloader._first_webpage_request:
|
2021-10-22 16:37:20 -04:00
|
|
|
|
sleep_interval = self.get_param('sleep_interval_requests') or 0
|
2021-02-27 07:41:23 -05:00
|
|
|
|
if sleep_interval > 0:
|
2021-02-28 19:09:50 -05:00
|
|
|
|
self.to_screen('Sleeping %s seconds ...' % sleep_interval)
|
2021-02-27 07:41:23 -05:00
|
|
|
|
time.sleep(sleep_interval)
|
|
|
|
|
else:
|
|
|
|
|
self._downloader._first_webpage_request = False
|
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
if note is None:
|
|
|
|
|
self.report_download_webpage(video_id)
|
|
|
|
|
elif note is not False:
|
2013-12-08 19:49:01 -05:00
|
|
|
|
if video_id is None:
|
2022-04-11 11:10:28 -04:00
|
|
|
|
self.to_screen(str(note))
|
2013-12-08 19:49:01 -05:00
|
|
|
|
else:
|
2022-04-11 11:10:28 -04:00
|
|
|
|
self.to_screen(f'{video_id}: {note}')
|
2017-12-23 08:57:35 -05:00
|
|
|
|
|
|
|
|
|
# Some sites check X-Forwarded-For HTTP header in order to figure out
|
|
|
|
|
# the origin of the client behind proxy. This allows bypassing geo
|
|
|
|
|
# restriction by faking this header's value to IP that belongs to some
|
|
|
|
|
# geo unrestricted country. We will do so once we encounter any
|
|
|
|
|
# geo restriction error.
|
|
|
|
|
if self._x_forwarded_for_ip:
|
2022-06-25 10:11:22 -04:00
|
|
|
|
headers = (headers or {}).copy()
|
|
|
|
|
headers.setdefault('X-Forwarded-For', self._x_forwarded_for_ip)
|
2017-12-23 08:57:35 -05:00
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
try:
|
2022-05-31 16:33:22 -04:00
|
|
|
|
return self._downloader.urlopen(self._create_request(url_or_request, data, headers, query))
|
2021-05-04 13:06:18 -04:00
|
|
|
|
except network_exceptions as err:
|
2022-06-24 04:10:17 -04:00
|
|
|
|
if isinstance(err, urllib.error.HTTPError):
|
2018-06-17 17:01:48 -04:00
|
|
|
|
if self.__can_accept_status_code(err, expected_status):
|
2018-11-02 14:18:20 -04:00
|
|
|
|
# Retain reference to error to prevent file object from
|
|
|
|
|
# being closed before it can be read. Works around the
|
|
|
|
|
# effects of <https://bugs.python.org/issue15002>
|
|
|
|
|
# introduced in Python 3.4.1.
|
|
|
|
|
err.fp._error = err
|
2018-06-17 17:01:48 -04:00
|
|
|
|
return err.fp
|
|
|
|
|
|
2013-12-20 11:05:28 -05:00
|
|
|
|
if errnote is False:
|
|
|
|
|
return False
|
2013-06-23 13:57:38 -04:00
|
|
|
|
if errnote is None:
|
2014-08-27 19:04:43 -04:00
|
|
|
|
errnote = 'Unable to download webpage'
|
2015-12-19 18:27:38 -05:00
|
|
|
|
|
2022-04-11 11:10:28 -04:00
|
|
|
|
errmsg = f'{errnote}: {error_to_compat_str(err)}'
|
2013-12-08 19:49:01 -05:00
|
|
|
|
if fatal:
|
2022-03-08 01:34:49 -05:00
|
|
|
|
raise ExtractorError(errmsg, cause=err)
|
2013-12-08 19:49:01 -05:00
|
|
|
|
else:
|
2021-04-16 06:01:10 -04:00
|
|
|
|
self.report_warning(errmsg)
|
2013-12-08 19:49:01 -05:00
|
|
|
|
return False
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2022-06-03 11:59:03 -04:00
|
|
|
|
def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True,
|
|
|
|
|
encoding=None, data=None, headers={}, query={}, expected_status=None):
|
2018-06-17 17:01:48 -04:00
|
|
|
|
"""
|
|
|
|
|
Return a tuple (page content as string, URL handle).
|
|
|
|
|
|
2022-05-31 13:43:26 -04:00
|
|
|
|
Arguments:
|
|
|
|
|
url_or_request -- plain text URL as a string or
|
2022-06-24 04:10:17 -04:00
|
|
|
|
a urllib.request.Request object
|
2022-05-31 13:43:26 -04:00
|
|
|
|
video_id -- Video/playlist/item identifier (string)
|
|
|
|
|
|
|
|
|
|
Keyword arguments:
|
|
|
|
|
note -- note printed before downloading (string)
|
|
|
|
|
errnote -- note printed in case of an error (string)
|
|
|
|
|
fatal -- flag denoting whether error should be considered fatal,
|
|
|
|
|
i.e. whether it should cause ExtractionError to be raised,
|
|
|
|
|
otherwise a warning will be reported and extraction continued
|
|
|
|
|
encoding -- encoding for a page content decoding, guessed automatically
|
|
|
|
|
when not explicitly specified
|
|
|
|
|
data -- POST data (bytes)
|
|
|
|
|
headers -- HTTP headers (dict)
|
|
|
|
|
query -- URL query (dict)
|
|
|
|
|
expected_status -- allows to accept failed HTTP requests (non 2xx
|
|
|
|
|
status code) by explicitly specifying a set of accepted status
|
|
|
|
|
codes. Can be any of the following entities:
|
|
|
|
|
- an integer type specifying an exact failed status code to
|
|
|
|
|
accept
|
|
|
|
|
- a list or a tuple of integer types specifying a list of
|
|
|
|
|
failed status codes to accept
|
|
|
|
|
- a callable accepting an actual failed status code and
|
|
|
|
|
returning True if it should be accepted
|
|
|
|
|
Note that this argument does not affect success status codes (2xx)
|
|
|
|
|
which are always accepted.
|
2018-06-17 17:01:48 -04:00
|
|
|
|
"""
|
2022-05-31 13:43:26 -04:00
|
|
|
|
|
2013-07-13 16:52:12 -04:00
|
|
|
|
# Strip hashes from the URL (#1038)
|
2022-06-24 06:54:43 -04:00
|
|
|
|
if isinstance(url_or_request, str):
|
2013-07-13 16:52:12 -04:00
|
|
|
|
url_or_request = url_or_request.partition('#')[0]
|
|
|
|
|
|
2018-06-17 17:01:48 -04:00
|
|
|
|
urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query, expected_status=expected_status)
|
2013-12-08 19:49:01 -05:00
|
|
|
|
if urlh is False:
|
|
|
|
|
assert not fatal
|
|
|
|
|
return False
|
2015-03-21 00:21:27 -04:00
|
|
|
|
content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
|
2014-10-26 12:05:44 -04:00
|
|
|
|
return (content, urlh)
|
|
|
|
|
|
2015-03-21 00:21:27 -04:00
|
|
|
|
@staticmethod
|
|
|
|
|
def _guess_encoding_from_content(content_type, webpage_bytes):
|
2013-06-23 13:57:38 -04:00
|
|
|
|
m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
|
|
|
|
|
if m:
|
|
|
|
|
encoding = m.group(1)
|
|
|
|
|
else:
|
2013-08-29 05:35:15 -04:00
|
|
|
|
m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
|
2013-08-28 07:59:08 -04:00
|
|
|
|
webpage_bytes[:1024])
|
|
|
|
|
if m:
|
|
|
|
|
encoding = m.group(1).decode('ascii')
|
2014-01-20 19:39:39 -05:00
|
|
|
|
elif webpage_bytes.startswith(b'\xff\xfe'):
|
|
|
|
|
encoding = 'utf-16'
|
2013-08-28 07:59:08 -04:00
|
|
|
|
else:
|
|
|
|
|
encoding = 'utf-8'
|
2015-03-21 00:21:27 -04:00
|
|
|
|
|
|
|
|
|
return encoding
|
|
|
|
|
|
2017-04-01 16:56:49 -04:00
|
|
|
|
def __check_blocked(self, content):
|
|
|
|
|
first_block = content[:512]
|
2019-05-10 16:56:22 -04:00
|
|
|
|
if ('<title>Access to this site is blocked</title>' in content
|
|
|
|
|
and 'Websense' in first_block):
|
2017-04-01 16:56:49 -04:00
|
|
|
|
msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
|
|
|
|
|
blocked_iframe = self._html_search_regex(
|
|
|
|
|
r'<iframe src="([^"]+)"', content,
|
|
|
|
|
'Websense information URL', default=None)
|
|
|
|
|
if blocked_iframe:
|
|
|
|
|
msg += ' Visit %s for more details' % blocked_iframe
|
|
|
|
|
raise ExtractorError(msg, expected=True)
|
|
|
|
|
if '<title>The URL you requested has been blocked</title>' in first_block:
|
|
|
|
|
msg = (
|
|
|
|
|
'Access to this webpage has been blocked by Indian censorship. '
|
|
|
|
|
'Use a VPN or proxy server (with --proxy) to route around it.')
|
|
|
|
|
block_msg = self._html_search_regex(
|
|
|
|
|
r'</h1><p>(.*?)</p>',
|
|
|
|
|
content, 'block message', default=None)
|
|
|
|
|
if block_msg:
|
|
|
|
|
msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
|
|
|
|
|
raise ExtractorError(msg, expected=True)
|
2019-05-10 16:56:22 -04:00
|
|
|
|
if ('<title>TTK :: Доступ к ресурсу ограничен</title>' in content
|
|
|
|
|
and 'blocklist.rkn.gov.ru' in content):
|
2017-04-01 16:56:49 -04:00
|
|
|
|
raise ExtractorError(
|
|
|
|
|
'Access to this webpage has been blocked by decision of the Russian government. '
|
|
|
|
|
'Visit http://blocklist.rkn.gov.ru/ for a block reason.',
|
|
|
|
|
expected=True)
|
|
|
|
|
|
2022-05-31 16:33:22 -04:00
|
|
|
|
def _request_dump_filename(self, url, video_id):
|
|
|
|
|
basen = f'{video_id}_{url}'
|
|
|
|
|
trim_length = self.get_param('trim_file_name') or 240
|
|
|
|
|
if len(basen) > trim_length:
|
|
|
|
|
h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
|
|
|
|
|
basen = basen[:trim_length - len(h)] + h
|
|
|
|
|
filename = sanitize_filename(f'{basen}.dump', restricted=True)
|
|
|
|
|
# Working around MAX_PATH limitation on Windows (see
|
|
|
|
|
# http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
|
|
|
|
|
if compat_os_name == 'nt':
|
|
|
|
|
absfilepath = os.path.abspath(filename)
|
|
|
|
|
if len(absfilepath) > 259:
|
|
|
|
|
filename = fR'\\?\{absfilepath}'
|
|
|
|
|
return filename
|
|
|
|
|
|
|
|
|
|
def __decode_webpage(self, webpage_bytes, encoding, headers):
|
|
|
|
|
if not encoding:
|
|
|
|
|
encoding = self._guess_encoding_from_content(headers.get('Content-Type', ''), webpage_bytes)
|
|
|
|
|
try:
|
|
|
|
|
return webpage_bytes.decode(encoding, 'replace')
|
|
|
|
|
except LookupError:
|
|
|
|
|
return webpage_bytes.decode('utf-8', 'replace')
|
|
|
|
|
|
2015-03-21 00:21:27 -04:00
|
|
|
|
def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
|
|
|
|
|
webpage_bytes = urlh.read()
|
|
|
|
|
if prefix is not None:
|
|
|
|
|
webpage_bytes = prefix + webpage_bytes
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if self.get_param('dump_intermediate_pages', False):
|
2017-11-18 07:02:56 -05:00
|
|
|
|
self.to_screen('Dumping request to ' + urlh.geturl())
|
2013-06-23 13:57:38 -04:00
|
|
|
|
dump = base64.b64encode(webpage_bytes).decode('ascii')
|
|
|
|
|
self._downloader.to_screen(dump)
|
2022-05-31 16:33:22 -04:00
|
|
|
|
if self.get_param('write_pages'):
|
2022-06-17 21:57:22 -04:00
|
|
|
|
filename = self._request_dump_filename(urlh.geturl(), video_id)
|
2022-05-31 16:33:22 -04:00
|
|
|
|
self.to_screen(f'Saving request to {filename}')
|
2013-10-28 05:44:02 -04:00
|
|
|
|
with open(filename, 'wb') as outf:
|
|
|
|
|
outf.write(webpage_bytes)
|
|
|
|
|
|
2022-05-31 16:33:22 -04:00
|
|
|
|
content = self.__decode_webpage(webpage_bytes, encoding, urlh.headers)
|
2017-04-01 16:56:49 -04:00
|
|
|
|
self.__check_blocked(content)
|
2014-04-03 00:07:35 -04:00
|
|
|
|
|
2014-10-26 12:05:44 -04:00
|
|
|
|
return content
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2022-07-15 06:35:00 -04:00
|
|
|
|
def __print_error(self, errnote, fatal, video_id, err):
|
|
|
|
|
if fatal:
|
2022-07-17 18:56:50 -04:00
|
|
|
|
raise ExtractorError(f'{video_id}: {errnote}', cause=err)
|
2022-07-15 06:35:00 -04:00
|
|
|
|
elif errnote:
|
2022-07-17 18:56:50 -04:00
|
|
|
|
self.report_warning(f'{video_id}: {errnote}: {err}')
|
2022-07-15 06:35:00 -04:00
|
|
|
|
|
|
|
|
|
def _parse_xml(self, xml_string, video_id, transform_source=None, fatal=True, errnote=None):
|
2013-12-10 06:45:22 -05:00
|
|
|
|
if transform_source:
|
|
|
|
|
xml_string = transform_source(xml_string)
|
2017-08-22 13:32:41 -04:00
|
|
|
|
try:
|
|
|
|
|
return compat_etree_fromstring(xml_string.encode('utf-8'))
|
2022-04-11 16:09:26 -04:00
|
|
|
|
except xml.etree.ElementTree.ParseError as ve:
|
2022-07-15 06:35:00 -04:00
|
|
|
|
self.__print_error('Failed to parse XML' if errnote is None else errnote, fatal, video_id, ve)
|
2013-11-24 08:59:19 -05:00
|
|
|
|
|
2022-07-15 06:35:00 -04:00
|
|
|
|
def _parse_json(self, json_string, video_id, transform_source=None, fatal=True, errnote=None, **parser_kwargs):
|
2014-01-07 03:35:34 -05:00
|
|
|
|
try:
|
2022-06-03 11:32:31 -04:00
|
|
|
|
return json.loads(
|
|
|
|
|
json_string, cls=LenientJSONDecoder, strict=False, transform_source=transform_source, **parser_kwargs)
|
2014-01-07 03:35:34 -05:00
|
|
|
|
except ValueError as ve:
|
2022-07-15 06:35:00 -04:00
|
|
|
|
self.__print_error('Failed to parse JSON' if errnote is None else errnote, fatal, video_id, ve)
|
2014-01-07 03:35:34 -05:00
|
|
|
|
|
2022-07-15 06:35:00 -04:00
|
|
|
|
def _parse_socket_response_as_json(self, data, *args, **kwargs):
|
|
|
|
|
return self._parse_json(data[data.find('{'):data.rfind('}') + 1], *args, **kwargs)
|
2021-05-30 04:17:39 -04:00
|
|
|
|
|
2022-05-31 13:43:26 -04:00
|
|
|
|
def __create_download_methods(name, parser, note, errnote, return_value):
|
|
|
|
|
|
2022-07-15 06:35:00 -04:00
|
|
|
|
def parse(ie, content, *args, errnote=errnote, **kwargs):
|
2022-05-31 13:43:26 -04:00
|
|
|
|
if parser is None:
|
|
|
|
|
return content
|
2022-07-15 06:35:00 -04:00
|
|
|
|
if errnote is False:
|
|
|
|
|
kwargs['errnote'] = errnote
|
2022-05-31 13:43:26 -04:00
|
|
|
|
# parser is fetched by name so subclasses can override it
|
|
|
|
|
return getattr(ie, parser)(content, *args, **kwargs)
|
|
|
|
|
|
2022-06-03 07:55:05 -04:00
|
|
|
|
def download_handle(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
|
|
|
|
|
fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
|
|
|
|
|
res = self._download_webpage_handle(
|
|
|
|
|
url_or_request, video_id, note=note, errnote=errnote, fatal=fatal, encoding=encoding,
|
|
|
|
|
data=data, headers=headers, query=query, expected_status=expected_status)
|
2022-05-31 13:43:26 -04:00
|
|
|
|
if res is False:
|
|
|
|
|
return res
|
|
|
|
|
content, urlh = res
|
2022-07-15 06:35:00 -04:00
|
|
|
|
return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote), urlh
|
2022-05-31 13:43:26 -04:00
|
|
|
|
|
2022-05-31 16:33:22 -04:00
|
|
|
|
def download_content(self, url_or_request, video_id, note=note, errnote=errnote, transform_source=None,
|
2022-06-03 07:55:05 -04:00
|
|
|
|
fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None):
|
2022-05-31 16:33:22 -04:00
|
|
|
|
if self.get_param('load_pages'):
|
|
|
|
|
url_or_request = self._create_request(url_or_request, data, headers, query)
|
|
|
|
|
filename = self._request_dump_filename(url_or_request.full_url, video_id)
|
|
|
|
|
self.to_screen(f'Loading request from {filename}')
|
|
|
|
|
try:
|
|
|
|
|
with open(filename, 'rb') as dumpf:
|
|
|
|
|
webpage_bytes = dumpf.read()
|
|
|
|
|
except OSError as e:
|
|
|
|
|
self.report_warning(f'Unable to load request from disk: {e}')
|
|
|
|
|
else:
|
|
|
|
|
content = self.__decode_webpage(webpage_bytes, encoding, url_or_request.headers)
|
2022-07-15 06:35:00 -04:00
|
|
|
|
return parse(self, content, video_id, transform_source=transform_source, fatal=fatal, errnote=errnote)
|
2022-06-03 07:55:05 -04:00
|
|
|
|
kwargs = {
|
|
|
|
|
'note': note,
|
|
|
|
|
'errnote': errnote,
|
|
|
|
|
'transform_source': transform_source,
|
|
|
|
|
'fatal': fatal,
|
|
|
|
|
'encoding': encoding,
|
|
|
|
|
'data': data,
|
|
|
|
|
'headers': headers,
|
|
|
|
|
'query': query,
|
|
|
|
|
'expected_status': expected_status,
|
|
|
|
|
}
|
2022-05-31 13:43:26 -04:00
|
|
|
|
if parser is None:
|
2022-06-03 07:55:05 -04:00
|
|
|
|
kwargs.pop('transform_source')
|
2022-05-31 13:43:26 -04:00
|
|
|
|
# The method is fetched by name so subclasses can override _download_..._handle
|
2022-06-03 07:55:05 -04:00
|
|
|
|
res = getattr(self, download_handle.__name__)(url_or_request, video_id, **kwargs)
|
2022-05-31 13:43:26 -04:00
|
|
|
|
return res if res is False else res[0]
|
|
|
|
|
|
|
|
|
|
def impersonate(func, name, return_value):
|
|
|
|
|
func.__name__, func.__qualname__ = name, f'InfoExtractor.{name}'
|
|
|
|
|
func.__doc__ = f'''
|
|
|
|
|
@param transform_source Apply this transformation before parsing
|
|
|
|
|
@returns {return_value}
|
|
|
|
|
|
|
|
|
|
See _download_webpage_handle docstring for other arguments specification
|
|
|
|
|
'''
|
|
|
|
|
|
|
|
|
|
impersonate(download_handle, f'_download_{name}_handle', f'({return_value}, URL handle)')
|
|
|
|
|
impersonate(download_content, f'_download_{name}', f'{return_value}')
|
|
|
|
|
return download_handle, download_content
|
|
|
|
|
|
|
|
|
|
_download_xml_handle, _download_xml = __create_download_methods(
|
|
|
|
|
'xml', '_parse_xml', 'Downloading XML', 'Unable to download XML', 'xml as an xml.etree.ElementTree.Element')
|
|
|
|
|
_download_json_handle, _download_json = __create_download_methods(
|
|
|
|
|
'json', '_parse_json', 'Downloading JSON metadata', 'Unable to download JSON metadata', 'JSON object as a dict')
|
|
|
|
|
_download_socket_json_handle, _download_socket_json = __create_download_methods(
|
|
|
|
|
'socket_json', '_parse_socket_response_as_json', 'Polling socket', 'Unable to poll socket', 'JSON object as a dict')
|
|
|
|
|
__download_webpage = __create_download_methods('webpage', None, None, None, 'data of the page as a string')[1]
|
2021-05-30 04:17:39 -04:00
|
|
|
|
|
2022-05-31 13:43:26 -04:00
|
|
|
|
def _download_webpage(
|
|
|
|
|
self, url_or_request, video_id, note=None, errnote=None,
|
|
|
|
|
fatal=True, tries=1, timeout=NO_DEFAULT, *args, **kwargs):
|
2021-05-30 04:17:39 -04:00
|
|
|
|
"""
|
2022-05-31 13:43:26 -04:00
|
|
|
|
Return the data of the page as a string.
|
2021-05-30 04:17:39 -04:00
|
|
|
|
|
2022-05-31 13:43:26 -04:00
|
|
|
|
Keyword arguments:
|
|
|
|
|
tries -- number of tries
|
|
|
|
|
timeout -- sleep interval between tries
|
|
|
|
|
|
|
|
|
|
See _download_webpage_handle docstring for other arguments specification.
|
2021-05-30 04:17:39 -04:00
|
|
|
|
"""
|
2022-05-31 13:43:26 -04:00
|
|
|
|
|
|
|
|
|
R''' # NB: These are unused; should they be deprecated?
|
|
|
|
|
if tries != 1:
|
|
|
|
|
self._downloader.deprecation_warning('tries argument is deprecated in InfoExtractor._download_webpage')
|
|
|
|
|
if timeout is NO_DEFAULT:
|
|
|
|
|
timeout = 5
|
|
|
|
|
else:
|
|
|
|
|
self._downloader.deprecation_warning('timeout argument is deprecated in InfoExtractor._download_webpage')
|
|
|
|
|
'''
|
|
|
|
|
|
|
|
|
|
try_count = 0
|
|
|
|
|
while True:
|
|
|
|
|
try:
|
|
|
|
|
return self.__download_webpage(url_or_request, video_id, note, errnote, None, fatal, *args, **kwargs)
|
2022-06-24 04:10:17 -04:00
|
|
|
|
except http.client.IncompleteRead as e:
|
2022-05-31 13:43:26 -04:00
|
|
|
|
try_count += 1
|
|
|
|
|
if try_count >= tries:
|
|
|
|
|
raise e
|
|
|
|
|
self._sleep(timeout, video_id)
|
2021-05-30 04:17:39 -04:00
|
|
|
|
|
2021-08-05 18:17:11 -04:00
|
|
|
|
def report_warning(self, msg, video_id=None, *args, only_once=False, **kwargs):
|
2022-06-17 22:00:12 -04:00
|
|
|
|
idstr = format_field(video_id, None, '%s: ')
|
2021-08-05 18:17:11 -04:00
|
|
|
|
msg = f'[{self.IE_NAME}] {idstr}{msg}'
|
|
|
|
|
if only_once:
|
|
|
|
|
if f'WARNING: {msg}' in self._printed_messages:
|
|
|
|
|
return
|
|
|
|
|
self._printed_messages.add(f'WARNING: {msg}')
|
|
|
|
|
self._downloader.report_warning(msg, *args, **kwargs)
|
2013-12-23 09:57:43 -05:00
|
|
|
|
|
2021-05-17 08:23:08 -04:00
|
|
|
|
def to_screen(self, msg, *args, **kwargs):
|
2013-06-23 13:57:38 -04:00
|
|
|
|
"""Print msg to screen, prefixing it with '[ie_name]'"""
|
2022-04-11 11:10:28 -04:00
|
|
|
|
self._downloader.to_screen(f'[{self.IE_NAME}] {msg}', *args, **kwargs)
|
2021-05-17 08:23:08 -04:00
|
|
|
|
|
|
|
|
|
def write_debug(self, msg, *args, **kwargs):
|
2022-04-11 11:10:28 -04:00
|
|
|
|
self._downloader.write_debug(f'[{self.IE_NAME}] {msg}', *args, **kwargs)
|
2021-05-17 08:23:08 -04:00
|
|
|
|
|
|
|
|
|
def get_param(self, name, default=None, *args, **kwargs):
|
|
|
|
|
if self._downloader:
|
|
|
|
|
return self._downloader.params.get(name, default, *args, **kwargs)
|
|
|
|
|
return default
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2022-10-18 13:58:57 -04:00
|
|
|
|
def report_drm(self, video_id, partial=NO_DEFAULT):
|
|
|
|
|
if partial is not NO_DEFAULT:
|
|
|
|
|
self._downloader.deprecation_warning('InfoExtractor.report_drm no longer accepts the argument partial')
|
2021-08-22 16:08:38 -04:00
|
|
|
|
self.raise_no_formats('This video is DRM protected', expected=True, video_id=video_id)
|
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
def report_extraction(self, id_or_name):
|
|
|
|
|
"""Report information extraction."""
|
2014-08-27 19:04:43 -04:00
|
|
|
|
self.to_screen('%s: Extracting information' % id_or_name)
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
|
|
|
|
def report_download_webpage(self, video_id):
|
|
|
|
|
"""Report webpage download."""
|
2014-08-27 19:04:43 -04:00
|
|
|
|
self.to_screen('%s: Downloading webpage' % video_id)
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
|
|
|
|
def report_age_confirmation(self):
|
|
|
|
|
"""Report attempt to confirm age."""
|
2014-08-27 19:04:43 -04:00
|
|
|
|
self.to_screen('Confirming age')
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2013-07-07 17:24:34 -04:00
|
|
|
|
def report_login(self):
|
|
|
|
|
"""Report attempt to log in."""
|
2014-08-27 19:04:43 -04:00
|
|
|
|
self.to_screen('Logging in')
|
2013-07-07 17:24:34 -04:00
|
|
|
|
|
2021-04-16 20:09:58 -04:00
|
|
|
|
def raise_login_required(
|
2021-05-19 09:41:44 -04:00
|
|
|
|
self, msg='This video is only available for registered users',
|
2022-03-18 16:53:33 -04:00
|
|
|
|
metadata_available=False, method=NO_DEFAULT):
|
2021-11-28 13:57:44 -05:00
|
|
|
|
if metadata_available and (
|
|
|
|
|
self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')):
|
2021-04-16 20:09:58 -04:00
|
|
|
|
self.report_warning(msg)
|
2022-03-04 09:07:43 -05:00
|
|
|
|
return
|
2022-06-17 22:00:12 -04:00
|
|
|
|
msg += format_field(self._login_hint(method), None, '. %s')
|
2021-07-06 16:57:53 -04:00
|
|
|
|
raise ExtractorError(msg, expected=True)
|
2015-08-26 11:24:47 -04:00
|
|
|
|
|
2021-04-16 20:09:58 -04:00
|
|
|
|
def raise_geo_restricted(
|
|
|
|
|
self, msg='This video is not available from your location due to geo restriction',
|
|
|
|
|
countries=None, metadata_available=False):
|
2021-11-28 13:57:44 -05:00
|
|
|
|
if metadata_available and (
|
|
|
|
|
self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')):
|
2021-04-16 20:09:58 -04:00
|
|
|
|
self.report_warning(msg)
|
|
|
|
|
else:
|
|
|
|
|
raise GeoRestrictedError(msg, countries=countries)
|
|
|
|
|
|
|
|
|
|
def raise_no_formats(self, msg, expected=False, video_id=None):
|
2021-11-28 13:57:44 -05:00
|
|
|
|
if expected and (
|
|
|
|
|
self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')):
|
2021-04-16 20:09:58 -04:00
|
|
|
|
self.report_warning(msg, video_id)
|
2021-08-01 22:16:12 -04:00
|
|
|
|
elif isinstance(msg, ExtractorError):
|
|
|
|
|
raise msg
|
2021-04-16 20:09:58 -04:00
|
|
|
|
else:
|
|
|
|
|
raise ExtractorError(msg, expected=expected, video_id=video_id)
|
2015-09-22 11:50:20 -04:00
|
|
|
|
|
2014-11-23 14:41:03 -05:00
|
|
|
|
# Methods for following #608
|
2013-12-19 14:28:52 -05:00
|
|
|
|
@staticmethod
|
2022-01-19 20:36:42 -05:00
|
|
|
|
def url_result(url, ie=None, video_id=None, video_title=None, *, url_transparent=False, **kwargs):
|
2015-07-23 13:37:45 -04:00
|
|
|
|
"""Returns a URL that points to a page that should be processed"""
|
2022-01-19 20:36:42 -05:00
|
|
|
|
if ie is not None:
|
|
|
|
|
kwargs['ie_key'] = ie if isinstance(ie, str) else ie.ie_key()
|
2013-11-22 16:46:46 -05:00
|
|
|
|
if video_id is not None:
|
2022-01-19 20:36:42 -05:00
|
|
|
|
kwargs['id'] = video_id
|
2015-04-12 13:11:47 -04:00
|
|
|
|
if video_title is not None:
|
2022-01-19 20:36:42 -05:00
|
|
|
|
kwargs['title'] = video_title
|
|
|
|
|
return {
|
|
|
|
|
**kwargs,
|
|
|
|
|
'_type': 'url_transparent' if url_transparent else 'url',
|
|
|
|
|
'url': url,
|
|
|
|
|
}
|
|
|
|
|
|
2022-07-31 21:22:03 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def playlist_from_matches(cls, matches, playlist_id=None, playlist_title=None,
|
|
|
|
|
getter=IDENTITY, ie=None, video_kwargs=None, **kwargs):
|
|
|
|
|
return cls.playlist_result(
|
|
|
|
|
(cls.url_result(m, ie, **(video_kwargs or {})) for m in orderedSet(map(getter, matches), lazy=True)),
|
|
|
|
|
playlist_id, playlist_title, **kwargs)
|
2017-03-08 18:13:54 -05:00
|
|
|
|
|
2013-12-19 14:28:52 -05:00
|
|
|
|
@staticmethod
|
2022-01-19 20:36:42 -05:00
|
|
|
|
def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None, *, multi_video=False, **kwargs):
|
2013-06-23 13:57:38 -04:00
|
|
|
|
"""Returns a playlist"""
|
|
|
|
|
if playlist_id:
|
2022-01-19 20:36:42 -05:00
|
|
|
|
kwargs['id'] = playlist_id
|
2013-06-23 13:57:38 -04:00
|
|
|
|
if playlist_title:
|
2022-01-19 20:36:42 -05:00
|
|
|
|
kwargs['title'] = playlist_title
|
2021-02-07 09:44:44 -05:00
|
|
|
|
if playlist_description is not None:
|
2022-01-19 20:36:42 -05:00
|
|
|
|
kwargs['description'] = playlist_description
|
|
|
|
|
return {
|
|
|
|
|
**kwargs,
|
|
|
|
|
'_type': 'multi_video' if multi_video else 'playlist',
|
|
|
|
|
'entries': entries,
|
|
|
|
|
}
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2015-06-28 12:56:45 -04:00
|
|
|
|
def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
|
2013-06-23 13:57:38 -04:00
|
|
|
|
"""
|
|
|
|
|
Perform a regex search on the given string, using a single or a list of
|
|
|
|
|
patterns returning the first matching group.
|
|
|
|
|
In case of failure return a default value or raise a WARNING or a
|
2013-10-23 08:38:03 -04:00
|
|
|
|
RegexNotFoundError, depending on fatal, specifying the field name.
|
2013-06-23 13:57:38 -04:00
|
|
|
|
"""
|
2022-04-07 04:11:16 -04:00
|
|
|
|
if string is None:
|
|
|
|
|
mobj = None
|
2022-02-04 08:37:02 -05:00
|
|
|
|
elif isinstance(pattern, (str, re.Pattern)):
|
2013-06-23 13:57:38 -04:00
|
|
|
|
mobj = re.search(pattern, string, flags)
|
|
|
|
|
else:
|
|
|
|
|
for p in pattern:
|
|
|
|
|
mobj = re.search(p, string, flags)
|
2014-07-25 04:43:03 -04:00
|
|
|
|
if mobj:
|
|
|
|
|
break
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
2021-10-20 12:37:32 -04:00
|
|
|
|
_name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
|
2013-06-23 13:57:38 -04:00
|
|
|
|
|
|
|
|
|
if mobj:
|
2014-11-04 17:14:16 -05:00
|
|
|
|
if group is None:
|
|
|
|
|
# return the first matching group
|
|
|
|
|
return next(g for g in mobj.groups() if g is not None)
|
2021-07-13 03:17:39 -04:00
|
|
|
|
elif isinstance(group, (list, tuple)):
|
|
|
|
|
return tuple(mobj.group(g) for g in group)
|
2014-11-04 17:14:16 -05:00
|
|
|
|
else:
|
|
|
|
|
return mobj.group(group)
|
2015-06-28 12:56:45 -04:00
|
|
|
|
elif default is not NO_DEFAULT:
|
2013-06-23 13:57:38 -04:00
|
|
|
|
return default
|
|
|
|
|
elif fatal:
|
2014-08-27 19:04:43 -04:00
|
|
|
|
raise RegexNotFoundError('Unable to extract %s' % _name)
|
2013-06-23 13:57:38 -04:00
|
|
|
|
else:
|
2021-04-16 06:01:10 -04:00
|
|
|
|
self.report_warning('unable to extract %s' % _name + bug_reports_message())
|
2013-06-23 13:57:38 -04:00
|
|
|
|
return None
|
|
|
|
|
|
2022-06-18 20:55:18 -04:00
|
|
|
|
def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
|
2022-10-03 07:20:27 -04:00
|
|
|
|
contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT, **kwargs):
|
2022-06-03 11:32:31 -04:00
|
|
|
|
"""Searches string for the JSON object specified by start_pattern"""
|
|
|
|
|
# NB: end_pattern is only used to reduce the size of the initial match
|
2022-06-18 20:55:18 -04:00
|
|
|
|
if default is NO_DEFAULT:
|
|
|
|
|
default, has_default = {}, False
|
|
|
|
|
else:
|
|
|
|
|
fatal, has_default = False, True
|
|
|
|
|
|
|
|
|
|
json_string = self._search_regex(
|
2022-10-03 07:20:27 -04:00
|
|
|
|
rf'(?:{start_pattern})\s*(?P<json>{contains_pattern})\s*(?:{end_pattern})',
|
2022-06-18 20:55:18 -04:00
|
|
|
|
string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT)
|
|
|
|
|
if not json_string:
|
|
|
|
|
return default
|
|
|
|
|
|
|
|
|
|
_name = self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
|
|
|
|
|
try:
|
|
|
|
|
return self._parse_json(json_string, video_id, ignore_extra=True, **kwargs)
|
|
|
|
|
except ExtractorError as e:
|
|
|
|
|
if fatal:
|
|
|
|
|
raise ExtractorError(
|
|
|
|
|
f'Unable to extract {_name} - Failed to parse JSON', cause=e.cause, video_id=video_id)
|
|
|
|
|
elif not has_default:
|
|
|
|
|
self.report_warning(
|
|
|
|
|
f'Unable to extract {_name} - Failed to parse JSON: {e}', video_id=video_id)
|
|
|
|
|
return default
|
2022-06-03 11:32:31 -04:00
|
|
|
|
|
2015-06-28 12:56:45 -04:00
|
|
|
|
def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
|
2013-06-23 13:57:38 -04:00
|
|
|
|
"""
|
|
|
|
|
Like _search_regex, but strips HTML tags and unescapes entities.
|
|
|
|
|
"""
|
2014-11-04 17:14:16 -05:00
|
|
|
|
res = self._search_regex(pattern, string, name, default, fatal, flags, group)
|
2013-06-23 13:57:38 -04:00
|
|
|
|
if res:
|
|
|
|
|
return clean_html(res).strip()
|
|
|
|
|
else:
|
|
|
|
|
return res
|
|
|
|
|
|
2016-08-14 06:48:13 -04:00
|
|
|
|
def _get_netrc_login_info(self, netrc_machine=None):
|
|
|
|
|
username = None
|
|
|
|
|
password = None
|
|
|
|
|
netrc_machine = netrc_machine or self._NETRC_MACHINE
|
|
|
|
|
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if self.get_param('usenetrc', False):
|
2016-08-14 06:48:13 -04:00
|
|
|
|
try:
|
2021-09-15 15:21:40 -04:00
|
|
|
|
netrc_file = compat_expanduser(self.get_param('netrc_location') or '~')
|
|
|
|
|
if os.path.isdir(netrc_file):
|
|
|
|
|
netrc_file = os.path.join(netrc_file, '.netrc')
|
|
|
|
|
info = netrc.netrc(file=netrc_file).authenticators(netrc_machine)
|
2016-08-14 06:48:13 -04:00
|
|
|
|
if info is not None:
|
|
|
|
|
username = info[0]
|
|
|
|
|
password = info[2]
|
|
|
|
|
else:
|
2016-09-15 11:35:12 -04:00
|
|
|
|
raise netrc.NetrcParseError(
|
|
|
|
|
'No authenticators for %s' % netrc_machine)
|
2022-04-11 11:10:28 -04:00
|
|
|
|
except (OSError, netrc.NetrcParseError) as err:
|
2021-04-16 06:01:10 -04:00
|
|
|
|
self.report_warning(
|
2016-09-15 11:35:12 -04:00
|
|
|
|
'parsing .netrc: %s' % error_to_compat_str(err))
|
2016-08-14 06:48:13 -04:00
|
|
|
|
|
2016-09-15 11:35:12 -04:00
|
|
|
|
return username, password
|
2016-08-14 06:48:13 -04:00
|
|
|
|
|
2016-09-13 17:16:01 -04:00
|
|
|
|
def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None):
|
2013-07-07 17:24:34 -04:00
|
|
|
|
"""
|
2015-04-29 11:03:10 -04:00
|
|
|
|
Get the login info as (username, password)
|
2016-09-15 11:34:29 -04:00
|
|
|
|
First look for the manually specified credentials using username_option
|
|
|
|
|
and password_option as keys in params dictionary. If no such credentials
|
|
|
|
|
available look in the netrc file using the netrc_machine or _NETRC_MACHINE
|
|
|
|
|
value.
|
2013-07-07 17:24:34 -04:00
|
|
|
|
If there's no info available, return (None, None)
|
|
|
|
|
"""
|
|
|
|
|
|
|
|
|
|
# Attempt to use provided username and password or .netrc data
|
2021-05-17 08:23:08 -04:00
|
|
|
|
username = self.get_param(username_option)
|
|
|
|
|
if username is not None:
|
|
|
|
|
password = self.get_param(password_option)
|
2016-08-14 06:48:13 -04:00
|
|
|
|
else:
|
2016-09-13 17:16:01 -04:00
|
|
|
|
username, password = self._get_netrc_login_info(netrc_machine)
|
2014-11-23 14:41:03 -05:00
|
|
|
|
|
2016-09-15 11:26:37 -04:00
|
|
|
|
return username, password
|
2013-07-07 17:24:34 -04:00
|
|
|
|
|
2015-08-15 11:55:07 -04:00
|
|
|
|
def _get_tfa_info(self, note='two-factor verification code'):
|
2014-08-16 17:28:41 -04:00
|
|
|
|
"""
|
|
|
|
|
Get the two-factor authentication info
|
|
|
|
|
TODO - asking the user will be required for sms/phone verify
|
|
|
|
|
currently just uses the command line option
|
|
|
|
|
If there's no info available, return None
|
|
|
|
|
"""
|
|
|
|
|
|
2021-05-17 08:23:08 -04:00
|
|
|
|
tfa = self.get_param('twofactor')
|
|
|
|
|
if tfa is not None:
|
|
|
|
|
return tfa
|
2014-08-16 17:28:41 -04:00
|
|
|
|
|
2022-06-24 04:10:17 -04:00
|
|
|
|
return getpass.getpass('Type %s and press [Return]: ' % note)
|
2014-08-16 17:28:41 -04:00
|
|
|
|
|
2013-07-12 13:00:19 -04:00
|
|
|
|
# Helper functions for extracting OpenGraph info
|
|
|
|
|
@staticmethod
|
2013-11-15 06:24:54 -05:00
|
|
|
|
def _og_regexes(prop):
|
2015-10-17 23:11:02 -04:00
|
|
|
|
content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))'
|
2022-04-03 11:11:50 -04:00
|
|
|
|
property_re = (r'(?:name|property)=(?:\'og%(sep)s%(prop)s\'|"og%(sep)s%(prop)s"|\s*og%(sep)s%(prop)s\b)'
|
|
|
|
|
% {'prop': re.escape(prop), 'sep': '(?::|[:-])'})
|
2013-11-15 06:54:13 -05:00
|
|
|
|
template = r'<meta[^>]+?%s[^>]+?%s'
|
2013-11-15 06:24:54 -05:00
|
|
|
|
return [
|
2013-11-15 06:54:13 -05:00
|
|
|
|
template % (property_re, content_re),
|
|
|
|
|
template % (content_re, property_re),
|
2013-11-15 06:24:54 -05:00
|
|
|
|
]
|
2013-07-12 13:00:19 -04:00
|
|
|
|
|
2015-07-28 17:43:03 -04:00
|
|
|
|
@staticmethod
|
|
|
|
|
def _meta_regex(prop):
|
|
|
|
|
return r'''(?isx)<meta
|
2015-08-15 05:58:30 -04:00
|
|
|
|
(?=[^>]+(?:itemprop|name|property|id|http-equiv)=(["\']?)%s\1)
|
2015-07-28 17:43:03 -04:00
|
|
|
|
[^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop)
|
|
|
|
|
|
2013-07-13 14:39:47 -04:00
|
|
|
|
def _og_search_property(self, prop, html, name=None, **kargs):
|
2021-07-10 17:59:44 -04:00
|
|
|
|
prop = variadic(prop)
|
2013-07-12 13:00:19 -04:00
|
|
|
|
if name is None:
|
2016-08-02 11:55:14 -04:00
|
|
|
|
name = 'OpenGraph %s' % prop[0]
|
|
|
|
|
og_regexes = []
|
|
|
|
|
for p in prop:
|
|
|
|
|
og_regexes.extend(self._og_regexes(p))
|
|
|
|
|
escaped = self._search_regex(og_regexes, html, name, flags=re.DOTALL, **kargs)
|
2013-11-12 04:36:23 -05:00
|
|
|
|
if escaped is None:
|
|
|
|
|
return None
|
|
|
|
|
return unescapeHTML(escaped)
|
2013-07-12 13:00:19 -04:00
|
|
|
|
|
|
|
|
|
def _og_search_thumbnail(self, html, **kargs):
|
2015-07-23 13:37:45 -04:00
|
|
|
|
return self._og_search_property('image', html, 'thumbnail URL', fatal=False, **kargs)
|
2013-07-12 13:00:19 -04:00
|
|
|
|
|
|
|
|
|
def _og_search_description(self, html, **kargs):
|
|
|
|
|
return self._og_search_property('description', html, fatal=False, **kargs)
|
|
|
|
|
|
2022-04-04 04:27:35 -04:00
|
|
|
|
def _og_search_title(self, html, *, fatal=False, **kargs):
|
|
|
|
|
return self._og_search_property('title', html, fatal=fatal, **kargs)
|
2013-07-12 13:00:19 -04:00
|
|
|
|
|
2013-10-28 02:34:29 -04:00
|
|
|
|
def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
|
2014-08-21 07:05:24 -04:00
|
|
|
|
regexes = self._og_regexes('video') + self._og_regexes('video:url')
|
|
|
|
|
if secure:
|
|
|
|
|
regexes = self._og_regexes('video:secure_url') + regexes
|
2013-10-28 02:34:29 -04:00
|
|
|
|
return self._html_search_regex(regexes, html, name, **kargs)
|
2013-07-12 13:00:19 -04:00
|
|
|
|
|
2014-06-26 10:34:36 -04:00
|
|
|
|
def _og_search_url(self, html, **kargs):
|
|
|
|
|
return self._og_search_property('url', html, **kargs)
|
|
|
|
|
|
2022-04-04 04:27:35 -04:00
|
|
|
|
def _html_extract_title(self, html, name='title', *, fatal=False, **kwargs):
|
2022-05-17 23:34:30 -04:00
|
|
|
|
return self._html_search_regex(r'(?s)<title\b[^>]*>([^<]+)</title>', html, name, fatal=fatal, **kwargs)
|
2022-02-22 11:15:08 -05:00
|
|
|
|
|
2014-07-11 09:38:18 -04:00
|
|
|
|
def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
|
2021-07-10 17:59:44 -04:00
|
|
|
|
name = variadic(name)
|
2013-11-20 00:13:19 -05:00
|
|
|
|
if display_name is None:
|
2016-06-26 05:57:14 -04:00
|
|
|
|
display_name = name[0]
|
2013-11-20 00:13:19 -05:00
|
|
|
|
return self._html_search_regex(
|
2016-06-26 05:57:14 -04:00
|
|
|
|
[self._meta_regex(n) for n in name],
|
2014-11-04 17:14:16 -05:00
|
|
|
|
html, display_name, fatal=fatal, group='content', **kwargs)
|
2013-11-20 00:13:19 -05:00
|
|
|
|
|
|
|
|
|
def _dc_search_uploader(self, html):
|
|
|
|
|
return self._html_search_meta('dc.creator', html, 'uploader')
|
|
|
|
|
|
2022-07-31 21:22:03 -04:00
|
|
|
|
@staticmethod
|
|
|
|
|
def _rta_search(html):
|
2013-10-06 00:06:30 -04:00
|
|
|
|
# See http://www.rtalabel.org/index.php?content=howtofaq#single
|
|
|
|
|
if re.search(r'(?ix)<meta\s+name="rating"\s+'
|
|
|
|
|
r' content="RTA-5042-1996-1400-1577-RTA"',
|
|
|
|
|
html):
|
|
|
|
|
return 18
|
2022-07-31 21:22:03 -04:00
|
|
|
|
|
|
|
|
|
# And then there are the jokers who advertise that they use RTA, but actually don't.
|
|
|
|
|
AGE_LIMIT_MARKERS = [
|
|
|
|
|
r'Proudly Labeled <a href="http://www\.rtalabel\.org/" title="Restricted to Adults">RTA</a>',
|
|
|
|
|
]
|
|
|
|
|
if any(re.search(marker, html) for marker in AGE_LIMIT_MARKERS):
|
|
|
|
|
return 18
|
2013-10-06 00:06:30 -04:00
|
|
|
|
return 0
|
|
|
|
|
|
2013-11-20 00:13:19 -05:00
|
|
|
|
def _media_rating_search(self, html):
|
|
|
|
|
# See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
|
|
|
|
|
rating = self._html_search_meta('rating', html)
|
|
|
|
|
|
|
|
|
|
if not rating:
|
|
|
|
|
return None
|
|
|
|
|
|
|
|
|
|
RATING_TABLE = {
|
|
|
|
|
'safe for kids': 0,
|
|
|
|
|
'general': 8,
|
|
|
|
|
'14 years': 14,
|
|
|
|
|
'mature': 17,
|
|
|
|
|
'restricted': 19,
|
|
|
|
|
}
|
2016-02-14 03:25:04 -05:00
|
|
|
|
return RATING_TABLE.get(rating.lower())
|
2013-11-20 00:13:19 -05:00
|
|
|
|
|
2015-02-08 10:39:00 -05:00
|
|
|
|
def _family_friendly_search(self, html):
|
2015-02-14 16:20:24 -05:00
|
|
|
|
# See http://schema.org/VideoObject
|
2017-08-12 06:11:35 -04:00
|
|
|
|
family_friendly = self._html_search_meta(
|
|
|
|
|
'isFamilyFriendly', html, default=None)
|
2015-02-08 10:39:00 -05:00
|
|
|
|
|
|
|
|
|
if not family_friendly:
|
|
|
|
|
return None
|
|
|
|
|
|
|
|
|
|
RATING_TABLE = {
|
|
|
|
|
'1': 0,
|
|
|
|
|
'true': 0,
|
|
|
|
|
'0': 18,
|
|
|
|
|
'false': 18,
|
|
|
|
|
}
|
2016-02-14 03:25:04 -05:00
|
|
|
|
return RATING_TABLE.get(family_friendly.lower())
|
2015-02-08 10:39:00 -05:00
|
|
|
|
|
2014-01-29 12:03:32 -05:00
|
|
|
|
def _twitter_search_player(self, html):
|
|
|
|
|
return self._html_search_meta('twitter:player', html,
|
2014-11-23 15:39:15 -05:00
|
|
|
|
'twitter card player')
|
2014-01-29 12:03:32 -05:00
|
|
|
|
|
2022-06-22 16:12:39 -04:00
|
|
|
|
def _yield_json_ld(self, html, video_id, *, fatal=True, default=NO_DEFAULT):
|
|
|
|
|
"""Yield all json ld objects in the html"""
|
|
|
|
|
if default is not NO_DEFAULT:
|
|
|
|
|
fatal = False
|
|
|
|
|
for mobj in re.finditer(JSON_LD_RE, html):
|
|
|
|
|
json_ld_item = self._parse_json(mobj.group('json_ld'), video_id, fatal=fatal)
|
|
|
|
|
for json_ld in variadic(json_ld_item):
|
|
|
|
|
if isinstance(json_ld, dict):
|
|
|
|
|
yield json_ld
|
|
|
|
|
|
|
|
|
|
def _search_json_ld(self, html, video_id, expected_type=None, *, fatal=True, default=NO_DEFAULT):
|
|
|
|
|
"""Search for a video in any json ld in the html"""
|
|
|
|
|
if default is not NO_DEFAULT:
|
|
|
|
|
fatal = False
|
|
|
|
|
info = self._json_ld(
|
|
|
|
|
list(self._yield_json_ld(html, video_id, fatal=fatal, default=default)),
|
|
|
|
|
video_id, fatal=fatal, expected_type=expected_type)
|
|
|
|
|
if info:
|
|
|
|
|
return info
|
2020-05-02 12:40:30 -04:00
|
|
|
|
if default is not NO_DEFAULT:
|
|
|
|
|
return default
|
|
|
|
|
elif fatal:
|
|
|
|
|
raise RegexNotFoundError('Unable to extract JSON-LD')
|
|
|
|
|
else:
|
2021-04-16 06:01:10 -04:00
|
|
|
|
self.report_warning('unable to extract JSON-LD %s' % bug_reports_message())
|
2020-05-02 12:40:30 -04:00
|
|
|
|
return {}
|
2016-01-15 13:36:02 -05:00
|
|
|
|
|
2016-07-08 16:28:04 -04:00
|
|
|
|
def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None):
|
2022-06-24 06:54:43 -04:00
|
|
|
|
if isinstance(json_ld, str):
|
2016-01-15 13:36:02 -05:00
|
|
|
|
json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
|
|
|
|
|
if not json_ld:
|
|
|
|
|
return {}
|
|
|
|
|
info = {}
|
2017-04-18 11:21:38 -04:00
|
|
|
|
|
2018-04-27 15:48:03 -04:00
|
|
|
|
INTERACTION_TYPE_MAP = {
|
|
|
|
|
'CommentAction': 'comment',
|
|
|
|
|
'AgreeAction': 'like',
|
|
|
|
|
'DisagreeAction': 'dislike',
|
|
|
|
|
'LikeAction': 'like',
|
|
|
|
|
'DislikeAction': 'dislike',
|
|
|
|
|
'ListenAction': 'view',
|
|
|
|
|
'WatchAction': 'view',
|
|
|
|
|
'ViewAction': 'view',
|
|
|
|
|
}
|
|
|
|
|
|
2022-06-13 09:09:58 -04:00
|
|
|
|
def is_type(e, *expected_types):
|
|
|
|
|
type = variadic(traverse_obj(e, '@type'))
|
|
|
|
|
return any(x in type for x in expected_types)
|
|
|
|
|
|
2021-01-01 07:26:37 -05:00
|
|
|
|
def extract_interaction_type(e):
|
|
|
|
|
interaction_type = e.get('interactionType')
|
|
|
|
|
if isinstance(interaction_type, dict):
|
|
|
|
|
interaction_type = interaction_type.get('@type')
|
|
|
|
|
return str_or_none(interaction_type)
|
|
|
|
|
|
2018-04-27 15:48:03 -04:00
|
|
|
|
def extract_interaction_statistic(e):
|
|
|
|
|
interaction_statistic = e.get('interactionStatistic')
|
2021-01-01 07:26:37 -05:00
|
|
|
|
if isinstance(interaction_statistic, dict):
|
|
|
|
|
interaction_statistic = [interaction_statistic]
|
2018-04-27 15:48:03 -04:00
|
|
|
|
if not isinstance(interaction_statistic, list):
|
|
|
|
|
return
|
|
|
|
|
for is_e in interaction_statistic:
|
2022-06-13 09:09:58 -04:00
|
|
|
|
if not is_type(is_e, 'InteractionCounter'):
|
2018-04-27 15:48:03 -04:00
|
|
|
|
continue
|
2021-01-01 07:26:37 -05:00
|
|
|
|
interaction_type = extract_interaction_type(is_e)
|
|
|
|
|
if not interaction_type:
|
2018-04-27 15:48:03 -04:00
|
|
|
|
continue
|
2020-09-18 19:33:17 -04:00
|
|
|
|
# For interaction count some sites provide string instead of
|
|
|
|
|
# an integer (as per spec) with non digit characters (e.g. ",")
|
|
|
|
|
# so extracting count with more relaxed str_to_int
|
|
|
|
|
interaction_count = str_to_int(is_e.get('userInteractionCount'))
|
2018-04-27 15:48:03 -04:00
|
|
|
|
if interaction_count is None:
|
|
|
|
|
continue
|
|
|
|
|
count_kind = INTERACTION_TYPE_MAP.get(interaction_type.split('/')[-1])
|
|
|
|
|
if not count_kind:
|
|
|
|
|
continue
|
|
|
|
|
count_key = '%s_count' % count_kind
|
|
|
|
|
if info.get(count_key) is not None:
|
|
|
|
|
continue
|
|
|
|
|
info[count_key] = interaction_count
|
|
|
|
|
|
2022-01-01 16:07:00 -05:00
|
|
|
|
def extract_chapter_information(e):
|
|
|
|
|
chapters = [{
|
|
|
|
|
'title': part.get('name'),
|
|
|
|
|
'start_time': part.get('startOffset'),
|
|
|
|
|
'end_time': part.get('endOffset'),
|
2022-02-01 20:58:01 -05:00
|
|
|
|
} for part in variadic(e.get('hasPart') or []) if part.get('@type') == 'Clip']
|
2022-01-01 16:07:00 -05:00
|
|
|
|
for idx, (last_c, current_c, next_c) in enumerate(zip(
|
|
|
|
|
[{'end_time': 0}] + chapters, chapters, chapters[1:])):
|
|
|
|
|
current_c['end_time'] = current_c['end_time'] or next_c['start_time']
|
|
|
|
|
current_c['start_time'] = current_c['start_time'] or last_c['end_time']
|
|
|
|
|
if None in current_c.values():
|
|
|
|
|
self.report_warning(f'Chapter {idx} contains broken data. Not extracting chapters')
|
|
|
|
|
return
|
|
|
|
|
if chapters:
|
|
|
|
|
chapters[-1]['end_time'] = chapters[-1]['end_time'] or info['duration']
|
|
|
|
|
info['chapters'] = chapters
|
|
|
|
|
|
2017-04-18 11:21:38 -04:00
|
|
|
|
def extract_video_object(e):
|
2021-04-10 12:47:11 -04:00
|
|
|
|
author = e.get('author')
|
2017-04-18 11:21:38 -04:00
|
|
|
|
info.update({
|
2022-06-22 16:12:39 -04:00
|
|
|
|
'url': url_or_none(e.get('contentUrl')),
|
2022-09-26 17:00:50 -04:00
|
|
|
|
'ext': mimetype2ext(e.get('encodingFormat')),
|
2017-04-18 11:21:38 -04:00
|
|
|
|
'title': unescapeHTML(e.get('name')),
|
|
|
|
|
'description': unescapeHTML(e.get('description')),
|
2022-07-13 14:29:44 -04:00
|
|
|
|
'thumbnails': [{'url': unescapeHTML(url)}
|
2022-05-17 23:34:30 -04:00
|
|
|
|
for url in variadic(traverse_obj(e, 'thumbnailUrl', 'thumbnailURL'))
|
|
|
|
|
if url_or_none(url)],
|
2017-04-18 11:21:38 -04:00
|
|
|
|
'duration': parse_duration(e.get('duration')),
|
|
|
|
|
'timestamp': unified_timestamp(e.get('uploadDate')),
|
2021-04-10 12:47:11 -04:00
|
|
|
|
# author can be an instance of 'Organization' or 'Person' types.
|
|
|
|
|
# both types can have 'name' property(inherited from 'Thing' type). [1]
|
|
|
|
|
# however some websites are using 'Text' type instead.
|
|
|
|
|
# 1. https://schema.org/VideoObject
|
2022-06-24 06:54:43 -04:00
|
|
|
|
'uploader': author.get('name') if isinstance(author, dict) else author if isinstance(author, str) else None,
|
2022-09-26 17:00:50 -04:00
|
|
|
|
'artist': traverse_obj(e, ('byArtist', 'name'), expected_type=str),
|
2022-06-10 15:03:54 -04:00
|
|
|
|
'filesize': int_or_none(float_or_none(e.get('contentSize'))),
|
2017-04-18 11:21:38 -04:00
|
|
|
|
'tbr': int_or_none(e.get('bitrate')),
|
|
|
|
|
'width': int_or_none(e.get('width')),
|
|
|
|
|
'height': int_or_none(e.get('height')),
|
2017-04-30 10:11:55 -04:00
|
|
|
|
'view_count': int_or_none(e.get('interactionCount')),
|
2022-09-26 17:00:50 -04:00
|
|
|
|
'tags': try_call(lambda: e.get('keywords').split(',')),
|
2017-04-18 11:21:38 -04:00
|
|
|
|
})
|
2022-09-26 17:00:50 -04:00
|
|
|
|
if is_type(e, 'AudioObject'):
|
|
|
|
|
info.update({
|
|
|
|
|
'vcodec': 'none',
|
|
|
|
|
'abr': int_or_none(e.get('bitrate')),
|
|
|
|
|
})
|
2018-04-27 15:48:03 -04:00
|
|
|
|
extract_interaction_statistic(e)
|
2022-01-01 16:07:00 -05:00
|
|
|
|
extract_chapter_information(e)
|
2017-04-18 11:21:38 -04:00
|
|
|
|
|
2021-12-16 15:46:30 -05:00
|
|
|
|
def traverse_json_ld(json_ld, at_top_level=True):
|
2022-10-08 23:17:58 -04:00
|
|
|
|
for e in variadic(json_ld):
|
|
|
|
|
if not isinstance(e, dict):
|
|
|
|
|
continue
|
2021-12-16 15:46:30 -05:00
|
|
|
|
if at_top_level and '@context' not in e:
|
|
|
|
|
continue
|
|
|
|
|
if at_top_level and set(e.keys()) == {'@context', '@graph'}:
|
2022-10-08 23:17:58 -04:00
|
|
|
|
traverse_json_ld(e['@graph'], at_top_level=False)
|
2022-10-19 17:47:18 -04:00
|
|
|
|
continue
|
2022-06-13 09:09:58 -04:00
|
|
|
|
if expected_type is not None and not is_type(e, expected_type):
|
2020-05-02 12:40:30 -04:00
|
|
|
|
continue
|
2021-11-23 02:41:28 -05:00
|
|
|
|
rating = traverse_obj(e, ('aggregateRating', 'ratingValue'), expected_type=float_or_none)
|
|
|
|
|
if rating is not None:
|
|
|
|
|
info['average_rating'] = rating
|
2022-06-13 09:09:58 -04:00
|
|
|
|
if is_type(e, 'TVEpisode', 'Episode'):
|
2019-01-07 22:02:49 -05:00
|
|
|
|
episode_name = unescapeHTML(e.get('name'))
|
2016-08-05 12:14:32 -04:00
|
|
|
|
info.update({
|
2019-01-07 22:02:49 -05:00
|
|
|
|
'episode': episode_name,
|
2016-08-05 12:14:32 -04:00
|
|
|
|
'episode_number': int_or_none(e.get('episodeNumber')),
|
|
|
|
|
'description': unescapeHTML(e.get('description')),
|
|
|
|
|
})
|
2019-01-07 22:02:49 -05:00
|
|
|
|
if not info.get('title') and episode_name:
|
|
|
|
|
info['title'] = episode_name
|
2016-08-05 12:14:32 -04:00
|
|
|
|
part_of_season = e.get('partOfSeason')
|
2022-06-13 09:09:58 -04:00
|
|
|
|
if is_type(part_of_season, 'TVSeason', 'Season', 'CreativeWorkSeason'):
|
2019-01-26 16:36:58 -05:00
|
|
|
|
info.update({
|
|
|
|
|
'season': unescapeHTML(part_of_season.get('name')),
|
|
|
|
|
'season_number': int_or_none(part_of_season.get('seasonNumber')),
|
|
|
|
|
})
|
2016-08-06 13:58:38 -04:00
|
|
|
|
part_of_series = e.get('partOfSeries') or e.get('partOfTVSeries')
|
2022-06-13 09:09:58 -04:00
|
|
|
|
if is_type(part_of_series, 'TVSeries', 'Series', 'CreativeWorkSeries'):
|
2016-08-05 12:14:32 -04:00
|
|
|
|
info['series'] = unescapeHTML(part_of_series.get('name'))
|
2022-06-13 09:09:58 -04:00
|
|
|
|
elif is_type(e, 'Movie'):
|
2019-01-07 22:02:00 -05:00
|
|
|
|
info.update({
|
|
|
|
|
'title': unescapeHTML(e.get('name')),
|
|
|
|
|
'description': unescapeHTML(e.get('description')),
|
|
|
|
|
'duration': parse_duration(e.get('duration')),
|
|
|
|
|
'timestamp': unified_timestamp(e.get('dateCreated')),
|
|
|
|
|
})
|
2022-06-13 09:09:58 -04:00
|
|
|
|
elif is_type(e, 'Article', 'NewsArticle'):
|
2016-08-05 12:14:32 -04:00
|
|
|
|
info.update({
|
|
|
|
|
'timestamp': parse_iso8601(e.get('datePublished')),
|
|
|
|
|
'title': unescapeHTML(e.get('headline')),
|
2021-12-16 15:46:30 -05:00
|
|
|
|
'description': unescapeHTML(e.get('articleBody') or e.get('description')),
|
2016-08-05 12:14:32 -04:00
|
|
|
|
})
|
2022-06-13 09:09:58 -04:00
|
|
|
|
if is_type(traverse_obj(e, ('video', 0)), 'VideoObject'):
|
2022-01-23 15:10:05 -05:00
|
|
|
|
extract_video_object(e['video'][0])
|
2022-06-13 09:09:58 -04:00
|
|
|
|
elif is_type(traverse_obj(e, ('subjectOf', 0)), 'VideoObject'):
|
2022-06-02 20:51:11 -04:00
|
|
|
|
extract_video_object(e['subjectOf'][0])
|
2022-09-26 17:00:50 -04:00
|
|
|
|
elif is_type(e, 'VideoObject', 'AudioObject'):
|
2017-04-18 11:21:38 -04:00
|
|
|
|
extract_video_object(e)
|
2020-05-02 12:40:30 -04:00
|
|
|
|
if expected_type is None:
|
|
|
|
|
continue
|
|
|
|
|
else:
|
|
|
|
|
break
|
2017-06-30 11:19:06 -04:00
|
|
|
|
video = e.get('video')
|
2022-06-13 09:09:58 -04:00
|
|
|
|
if is_type(video, 'VideoObject'):
|
2017-06-30 11:19:06 -04:00
|
|
|
|
extract_video_object(video)
|
2020-05-02 12:40:30 -04:00
|
|
|
|
if expected_type is None:
|
|
|
|
|
continue
|
|
|
|
|
else:
|
|
|
|
|
break
|
2021-12-16 15:46:30 -05:00
|
|
|
|
|
2022-10-08 23:17:58 -04:00
|
|
|
|
traverse_json_ld(json_ld)
|
2022-03-27 22:51:45 -04:00
|
|
|
|
return filter_dict(info)
|
2016-01-15 13:36:02 -05:00
|
|
|
|
|
2022-01-19 16:55:15 -05:00
|
|
|
|
def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw):
|
2021-10-22 22:02:23 -04:00
|
|
|
|
return self._parse_json(
|
|
|
|
|
self._search_regex(
|
|
|
|
|
r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>',
|
2022-01-19 16:55:15 -05:00
|
|
|
|
webpage, 'next.js data', fatal=fatal, **kw),
|
|
|
|
|
video_id, transform_source=transform_source, fatal=fatal)
|
2021-10-22 22:02:23 -04:00
|
|
|
|
|
2022-06-19 17:33:19 -04:00
|
|
|
|
def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__', *, fatal=True, traverse=('data', 0)):
|
|
|
|
|
"""Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
|
2021-12-07 11:38:50 -05:00
|
|
|
|
rectx = re.escape(context_name)
|
2022-06-19 17:33:19 -04:00
|
|
|
|
FUNCTION_RE = r'\(function\((?P<arg_keys>.*?)\){return\s+(?P<js>{.*?})\s*;?\s*}\((?P<arg_vals>.*?)\)'
|
2021-12-07 11:38:50 -05:00
|
|
|
|
js, arg_keys, arg_vals = self._search_regex(
|
2022-06-19 17:33:19 -04:00
|
|
|
|
(rf'<script>\s*window\.{rectx}={FUNCTION_RE}\s*\)\s*;?\s*</script>', rf'{rectx}\(.*?{FUNCTION_RE}'),
|
|
|
|
|
webpage, context_name, group=('js', 'arg_keys', 'arg_vals'), fatal=fatal)
|
2021-12-07 11:38:50 -05:00
|
|
|
|
|
|
|
|
|
args = dict(zip(arg_keys.split(','), arg_vals.split(',')))
|
|
|
|
|
|
|
|
|
|
for key, val in args.items():
|
|
|
|
|
if val in ('undefined', 'void 0'):
|
|
|
|
|
args[key] = 'null'
|
|
|
|
|
|
2022-06-19 17:33:19 -04:00
|
|
|
|
ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal)
|
|
|
|
|
return traverse_obj(ret, traverse) or {}
|
2021-12-07 11:38:50 -05:00
|
|
|
|
|
2015-07-10 11:49:09 -04:00
|
|
|
|
@staticmethod
|
2015-07-14 12:36:30 -04:00
|
|
|
|
def _hidden_inputs(html):
|
2015-09-11 11:07:32 -04:00
|
|
|
|
html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
|
2015-08-15 11:52:22 -04:00
|
|
|
|
hidden_inputs = {}
|
2016-09-15 10:54:48 -04:00
|
|
|
|
for input in re.findall(r'(?i)(<input[^>]+>)', html):
|
|
|
|
|
attrs = extract_attributes(input)
|
|
|
|
|
if not input:
|
2015-08-15 11:52:22 -04:00
|
|
|
|
continue
|
2016-09-15 10:54:48 -04:00
|
|
|
|
if attrs.get('type') not in ('hidden', 'submit'):
|
2015-08-15 11:52:22 -04:00
|
|
|
|
continue
|
2016-09-15 10:54:48 -04:00
|
|
|
|
name = attrs.get('name') or attrs.get('id')
|
|
|
|
|
value = attrs.get('value')
|
|
|
|
|
if name and value is not None:
|
|
|
|
|
hidden_inputs[name] = value
|
2015-08-15 11:52:22 -04:00
|
|
|
|
return hidden_inputs
|
2015-07-10 11:49:09 -04:00
|
|
|
|
|
2015-07-14 12:38:10 -04:00
|
|
|
|
def _form_hidden_inputs(self, form_id, html):
|
|
|
|
|
form = self._search_regex(
|
2015-09-11 10:43:05 -04:00
|
|
|
|
r'(?is)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
|
2015-07-14 12:38:10 -04:00
|
|
|
|
html, '%s form' % form_id, group='form')
|
|
|
|
|
return self._hidden_inputs(form)
|
|
|
|
|
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
class FormatSort:
|
2021-04-01 04:46:10 -04:00
|
|
|
|
regex = r' *((?P<reverse>\+)?(?P<field>[a-zA-Z0-9_]+)((?P<separator>[~:])(?P<limit>.*?))?)? *$'
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
|
2021-06-12 16:02:19 -04:00
|
|
|
|
default = ('hidden', 'aud_or_vid', 'hasvid', 'ie_pref', 'lang', 'quality',
|
2022-08-10 21:52:36 -04:00
|
|
|
|
'res', 'fps', 'hdr:12', 'vcodec:vp9.2', 'channels', 'acodec',
|
|
|
|
|
'size', 'br', 'asr', 'proto', 'ext', 'hasaud', 'source', 'id') # These must not be aliases
|
2021-08-25 22:04:08 -04:00
|
|
|
|
ytdl_default = ('hasaud', 'lang', 'quality', 'tbr', 'filesize', 'vbr',
|
2021-05-11 04:00:48 -04:00
|
|
|
|
'height', 'width', 'proto', 'vext', 'abr', 'aext',
|
2021-11-28 16:22:52 -05:00
|
|
|
|
'fps', 'fs_approx', 'source', 'id')
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
|
|
|
|
|
settings = {
|
|
|
|
|
'vcodec': {'type': 'ordered', 'regex': True,
|
2021-02-18 09:24:30 -05:00
|
|
|
|
'order': ['av0?1', 'vp0?9.2', 'vp0?9', '[hx]265|he?vc?', '[hx]264|avc', 'vp0?8', 'mp4v|h263', 'theora', '', None, 'none']},
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'acodec': {'type': 'ordered', 'regex': True,
|
2022-03-04 09:11:44 -05:00
|
|
|
|
'order': ['[af]lac', 'wav|aiff', 'opus', 'vorbis|ogg', 'aac', 'mp?4a?', 'mp3', 'e-?a?c-?3', 'ac-?3', 'dts', '', None, 'none']},
|
2021-10-18 09:04:21 -04:00
|
|
|
|
'hdr': {'type': 'ordered', 'regex': True, 'field': 'dynamic_range',
|
|
|
|
|
'order': ['dv', '(hdr)?12', r'(hdr)?10\+', '(hdr)?10', 'hlg', '', 'sdr', None]},
|
2021-01-25 08:58:39 -05:00
|
|
|
|
'proto': {'type': 'ordered', 'regex': True, 'field': 'protocol',
|
2021-11-28 16:22:52 -05:00
|
|
|
|
'order': ['(ht|f)tps', '(ht|f)tp$', 'm3u8.*', '.*dash', 'websocket_frag', 'rtmpe?', '', 'mms|rtsp', 'ws|websocket', 'f4']},
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'vext': {'type': 'ordered', 'field': 'video_ext',
|
Change defaults
* Enabled --ignore by default
* Disabled --video-multistreams and --audio-multistreams by default
* Changed default format selection to 'bv*+ba/b' when --audio-multistreams is disabled
* Changed default format sort order to 'res,fps,codec,size,br,asr,proto,ext,has_audio,source,format_id'
* Changed default output template to '%(title)s [%(id)s].%(ext)s'
* Enabled `--list-formats-as-table` by default
2021-01-04 11:40:47 -05:00
|
|
|
|
'order': ('mp4', 'webm', 'flv', '', 'none'),
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'order_free': ('webm', 'mp4', 'flv', '', 'none')},
|
|
|
|
|
'aext': {'type': 'ordered', 'field': 'audio_ext',
|
|
|
|
|
'order': ('m4a', 'aac', 'mp3', 'ogg', 'opus', 'webm', '', 'none'),
|
2022-08-31 13:19:14 -04:00
|
|
|
|
'order_free': ('ogg', 'opus', 'webm', 'mp3', 'm4a', 'aac', '', 'none')},
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'hidden': {'visible': False, 'forced': True, 'type': 'extractor', 'max': -1000},
|
2021-07-31 06:29:52 -04:00
|
|
|
|
'aud_or_vid': {'visible': False, 'forced': True, 'type': 'multiple',
|
2021-06-12 16:02:19 -04:00
|
|
|
|
'field': ('vcodec', 'acodec'),
|
|
|
|
|
'function': lambda it: int(any(v != 'none' for v in it))},
|
2021-02-18 17:03:16 -05:00
|
|
|
|
'ie_pref': {'priority': True, 'type': 'extractor'},
|
2021-01-21 07:50:09 -05:00
|
|
|
|
'hasvid': {'priority': True, 'field': 'vcodec', 'type': 'boolean', 'not_in_list': ('none',)},
|
|
|
|
|
'hasaud': {'field': 'acodec', 'type': 'boolean', 'not_in_list': ('none',)},
|
2021-10-29 15:43:51 -04:00
|
|
|
|
'lang': {'convert': 'float', 'field': 'language_preference', 'default': -1},
|
|
|
|
|
'quality': {'convert': 'float', 'default': -1},
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'filesize': {'convert': 'bytes'},
|
2021-01-25 08:58:39 -05:00
|
|
|
|
'fs_approx': {'convert': 'bytes', 'field': 'filesize_approx'},
|
|
|
|
|
'id': {'convert': 'string', 'field': 'format_id'},
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'height': {'convert': 'float_none'},
|
|
|
|
|
'width': {'convert': 'float_none'},
|
|
|
|
|
'fps': {'convert': 'float_none'},
|
2022-08-07 16:05:36 -04:00
|
|
|
|
'channels': {'convert': 'float_none', 'field': 'audio_channels'},
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'tbr': {'convert': 'float_none'},
|
|
|
|
|
'vbr': {'convert': 'float_none'},
|
|
|
|
|
'abr': {'convert': 'float_none'},
|
|
|
|
|
'asr': {'convert': 'float_none'},
|
2021-10-29 15:43:51 -04:00
|
|
|
|
'source': {'convert': 'float', 'field': 'source_preference', 'default': -1},
|
2021-01-21 07:50:09 -05:00
|
|
|
|
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'codec': {'type': 'combined', 'field': ('vcodec', 'acodec')},
|
2021-01-21 07:50:09 -05:00
|
|
|
|
'br': {'type': 'combined', 'field': ('tbr', 'vbr', 'abr'), 'same_limit': True},
|
|
|
|
|
'size': {'type': 'combined', 'same_limit': True, 'field': ('filesize', 'fs_approx')},
|
|
|
|
|
'ext': {'type': 'combined', 'field': ('vext', 'aext')},
|
2021-07-31 06:29:52 -04:00
|
|
|
|
'res': {'type': 'multiple', 'field': ('height', 'width'),
|
2021-07-31 06:51:01 -04:00
|
|
|
|
'function': lambda it: (lambda l: min(l) if l else 0)(tuple(filter(None, it)))},
|
2021-01-21 07:50:09 -05:00
|
|
|
|
|
2022-08-07 16:05:36 -04:00
|
|
|
|
# Actual field names
|
2021-12-09 17:48:36 -05:00
|
|
|
|
'format_id': {'type': 'alias', 'field': 'id'},
|
|
|
|
|
'preference': {'type': 'alias', 'field': 'ie_pref'},
|
|
|
|
|
'language_preference': {'type': 'alias', 'field': 'lang'},
|
2021-01-21 07:50:09 -05:00
|
|
|
|
'source_preference': {'type': 'alias', 'field': 'source'},
|
2022-03-04 09:08:55 -05:00
|
|
|
|
'protocol': {'type': 'alias', 'field': 'proto'},
|
2021-01-21 07:50:09 -05:00
|
|
|
|
'filesize_approx': {'type': 'alias', 'field': 'fs_approx'},
|
2022-08-07 16:05:36 -04:00
|
|
|
|
'audio_channels': {'type': 'alias', 'field': 'channels'},
|
2022-03-04 09:08:55 -05:00
|
|
|
|
|
|
|
|
|
# Deprecated
|
|
|
|
|
'dimension': {'type': 'alias', 'field': 'res', 'deprecated': True},
|
|
|
|
|
'resolution': {'type': 'alias', 'field': 'res', 'deprecated': True},
|
|
|
|
|
'extension': {'type': 'alias', 'field': 'ext', 'deprecated': True},
|
|
|
|
|
'bitrate': {'type': 'alias', 'field': 'br', 'deprecated': True},
|
|
|
|
|
'total_bitrate': {'type': 'alias', 'field': 'tbr', 'deprecated': True},
|
|
|
|
|
'video_bitrate': {'type': 'alias', 'field': 'vbr', 'deprecated': True},
|
|
|
|
|
'audio_bitrate': {'type': 'alias', 'field': 'abr', 'deprecated': True},
|
|
|
|
|
'framerate': {'type': 'alias', 'field': 'fps', 'deprecated': True},
|
|
|
|
|
'filesize_estimate': {'type': 'alias', 'field': 'size', 'deprecated': True},
|
|
|
|
|
'samplerate': {'type': 'alias', 'field': 'asr', 'deprecated': True},
|
|
|
|
|
'video_ext': {'type': 'alias', 'field': 'vext', 'deprecated': True},
|
|
|
|
|
'audio_ext': {'type': 'alias', 'field': 'aext', 'deprecated': True},
|
|
|
|
|
'video_codec': {'type': 'alias', 'field': 'vcodec', 'deprecated': True},
|
|
|
|
|
'audio_codec': {'type': 'alias', 'field': 'acodec', 'deprecated': True},
|
|
|
|
|
'video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
|
|
|
|
|
'has_video': {'type': 'alias', 'field': 'hasvid', 'deprecated': True},
|
|
|
|
|
'audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
|
|
|
|
|
'has_audio': {'type': 'alias', 'field': 'hasaud', 'deprecated': True},
|
|
|
|
|
'extractor': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
|
|
|
|
|
'extractor_preference': {'type': 'alias', 'field': 'ie_pref', 'deprecated': True},
|
2021-01-21 07:50:09 -05:00
|
|
|
|
}
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
|
2021-11-28 16:22:52 -05:00
|
|
|
|
def __init__(self, ie, field_preference):
|
|
|
|
|
self._order = []
|
|
|
|
|
self.ydl = ie._downloader
|
|
|
|
|
self.evaluate_params(self.ydl.params, field_preference)
|
|
|
|
|
if ie.get_param('verbose'):
|
|
|
|
|
self.print_verbose_info(self.ydl.write_debug)
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
|
|
|
|
|
def _get_field_setting(self, field, key):
|
|
|
|
|
if field not in self.settings:
|
2021-11-29 12:46:06 -05:00
|
|
|
|
if key in ('forced', 'priority'):
|
|
|
|
|
return False
|
2022-08-30 11:28:28 -04:00
|
|
|
|
self.ydl.deprecated_feature(f'Using arbitrary fields ({field}) for format sorting is '
|
|
|
|
|
'deprecated and may be removed in a future version')
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
self.settings[field] = {}
|
|
|
|
|
propObj = self.settings[field]
|
|
|
|
|
if key not in propObj:
|
|
|
|
|
type = propObj.get('type')
|
|
|
|
|
if key == 'field':
|
|
|
|
|
default = 'preference' if type == 'extractor' else (field,) if type in ('combined', 'multiple') else field
|
|
|
|
|
elif key == 'convert':
|
|
|
|
|
default = 'order' if type == 'ordered' else 'float_string' if field else 'ignore'
|
2013-12-24 06:25:22 -05:00
|
|
|
|
else:
|
2021-07-31 06:29:52 -04:00
|
|
|
|
default = {'type': 'field', 'visible': True, 'order': [], 'not_in_list': (None,)}.get(key, None)
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
propObj[key] = default
|
|
|
|
|
return propObj[key]
|
|
|
|
|
|
|
|
|
|
def _resolve_field_value(self, field, value, convertNone=False):
|
|
|
|
|
if value is None:
|
|
|
|
|
if not convertNone:
|
|
|
|
|
return None
|
2013-12-24 06:25:22 -05:00
|
|
|
|
else:
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
value = value.lower()
|
|
|
|
|
conversion = self._get_field_setting(field, 'convert')
|
|
|
|
|
if conversion == 'ignore':
|
|
|
|
|
return None
|
|
|
|
|
if conversion == 'string':
|
|
|
|
|
return value
|
|
|
|
|
elif conversion == 'float_none':
|
|
|
|
|
return float_or_none(value)
|
|
|
|
|
elif conversion == 'bytes':
|
|
|
|
|
return FileDownloader.parse_bytes(value)
|
|
|
|
|
elif conversion == 'order':
|
2021-02-18 09:17:46 -05:00
|
|
|
|
order_list = (self._use_free_order and self._get_field_setting(field, 'order_free')) or self._get_field_setting(field, 'order')
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
use_regex = self._get_field_setting(field, 'regex')
|
|
|
|
|
list_length = len(order_list)
|
|
|
|
|
empty_pos = order_list.index('') if '' in order_list else list_length + 1
|
|
|
|
|
if use_regex and value is not None:
|
2021-02-18 09:17:46 -05:00
|
|
|
|
for i, regex in enumerate(order_list):
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
if regex and re.match(regex, value):
|
|
|
|
|
return list_length - i
|
|
|
|
|
return list_length - empty_pos # not in list
|
|
|
|
|
else: # not regex or value = None
|
|
|
|
|
return list_length - (order_list.index(value) if value in order_list else empty_pos)
|
|
|
|
|
else:
|
|
|
|
|
if value.isnumeric():
|
|
|
|
|
return float(value)
|
2013-12-24 06:25:22 -05:00
|
|
|
|
else:
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
self.settings[field]['convert'] = 'string'
|
|
|
|
|
return value
|
|
|
|
|
|
|
|
|
|
def evaluate_params(self, params, sort_extractor):
|
|
|
|
|
self._use_free_order = params.get('prefer_free_formats', False)
|
|
|
|
|
self._sort_user = params.get('format_sort', [])
|
|
|
|
|
self._sort_extractor = sort_extractor
|
|
|
|
|
|
|
|
|
|
def add_item(field, reverse, closest, limit_text):
|
|
|
|
|
field = field.lower()
|
|
|
|
|
if field in self._order:
|
|
|
|
|
return
|
|
|
|
|
self._order.append(field)
|
|
|
|
|
limit = self._resolve_field_value(field, limit_text)
|
|
|
|
|
data = {
|
|
|
|
|
'reverse': reverse,
|
|
|
|
|
'closest': False if limit is None else closest,
|
|
|
|
|
'limit_text': limit_text,
|
|
|
|
|
'limit': limit}
|
|
|
|
|
if field in self.settings:
|
|
|
|
|
self.settings[field].update(data)
|
|
|
|
|
else:
|
|
|
|
|
self.settings[field] = data
|
|
|
|
|
|
|
|
|
|
sort_list = (
|
|
|
|
|
tuple(field for field in self.default if self._get_field_setting(field, 'forced'))
|
|
|
|
|
+ (tuple() if params.get('format_sort_force', False)
|
|
|
|
|
else tuple(field for field in self.default if self._get_field_setting(field, 'priority')))
|
|
|
|
|
+ tuple(self._sort_user) + tuple(sort_extractor) + self.default)
|
|
|
|
|
|
|
|
|
|
for item in sort_list:
|
|
|
|
|
match = re.match(self.regex, item)
|
|
|
|
|
if match is None:
|
|
|
|
|
raise ExtractorError('Invalid format sort string "%s" given by extractor' % item)
|
|
|
|
|
field = match.group('field')
|
|
|
|
|
if field is None:
|
|
|
|
|
continue
|
|
|
|
|
if self._get_field_setting(field, 'type') == 'alias':
|
2021-11-29 12:46:06 -05:00
|
|
|
|
alias, field = field, self._get_field_setting(field, 'field')
|
2022-03-04 09:08:55 -05:00
|
|
|
|
if self._get_field_setting(alias, 'deprecated'):
|
2022-08-30 11:28:28 -04:00
|
|
|
|
self.ydl.deprecated_feature(f'Format sorting alias {alias} is deprecated and may '
|
2022-10-03 16:04:04 -04:00
|
|
|
|
f'be removed in a future version. Please use {field} instead')
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
reverse = match.group('reverse') is not None
|
2021-04-01 04:46:10 -04:00
|
|
|
|
closest = match.group('separator') == '~'
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
limit_text = match.group('limit')
|
|
|
|
|
|
|
|
|
|
has_limit = limit_text is not None
|
|
|
|
|
has_multiple_fields = self._get_field_setting(field, 'type') == 'combined'
|
|
|
|
|
has_multiple_limits = has_limit and has_multiple_fields and not self._get_field_setting(field, 'same_limit')
|
|
|
|
|
|
|
|
|
|
fields = self._get_field_setting(field, 'field') if has_multiple_fields else (field,)
|
2021-10-08 20:23:15 -04:00
|
|
|
|
limits = limit_text.split(':') if has_multiple_limits else (limit_text,) if has_limit else tuple()
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
limit_count = len(limits)
|
|
|
|
|
for (i, f) in enumerate(fields):
|
|
|
|
|
add_item(f, reverse, closest,
|
|
|
|
|
limits[i] if i < limit_count
|
|
|
|
|
else limits[0] if has_limit and not has_multiple_limits
|
|
|
|
|
else None)
|
|
|
|
|
|
2021-05-14 03:45:29 -04:00
|
|
|
|
def print_verbose_info(self, write_debug):
|
2021-04-10 20:22:55 -04:00
|
|
|
|
if self._sort_user:
|
2021-05-14 03:45:29 -04:00
|
|
|
|
write_debug('Sort order given by user: %s' % ', '.join(self._sort_user))
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
if self._sort_extractor:
|
2021-05-14 03:45:29 -04:00
|
|
|
|
write_debug('Sort order given by extractor: %s' % ', '.join(self._sort_extractor))
|
|
|
|
|
write_debug('Formats sorted by: %s' % ', '.join(['%s%s%s' % (
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
'+' if self._get_field_setting(field, 'reverse') else '', field,
|
|
|
|
|
'%s%s(%s)' % ('~' if self._get_field_setting(field, 'closest') else ':',
|
|
|
|
|
self._get_field_setting(field, 'limit_text'),
|
|
|
|
|
self._get_field_setting(field, 'limit'))
|
|
|
|
|
if self._get_field_setting(field, 'limit_text') is not None else '')
|
|
|
|
|
for field in self._order if self._get_field_setting(field, 'visible')]))
|
|
|
|
|
|
|
|
|
|
def _calculate_field_preference_from_value(self, format, field, type, value):
|
|
|
|
|
reverse = self._get_field_setting(field, 'reverse')
|
|
|
|
|
closest = self._get_field_setting(field, 'closest')
|
|
|
|
|
limit = self._get_field_setting(field, 'limit')
|
|
|
|
|
|
|
|
|
|
if type == 'extractor':
|
|
|
|
|
maximum = self._get_field_setting(field, 'max')
|
|
|
|
|
if value is None or (maximum is not None and value >= maximum):
|
2021-02-18 17:03:16 -05:00
|
|
|
|
value = -1
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
elif type == 'boolean':
|
|
|
|
|
in_list = self._get_field_setting(field, 'in_list')
|
|
|
|
|
not_in_list = self._get_field_setting(field, 'not_in_list')
|
|
|
|
|
value = 0 if ((in_list is None or value in in_list) and (not_in_list is None or value not in not_in_list)) else -1
|
|
|
|
|
elif type == 'ordered':
|
|
|
|
|
value = self._resolve_field_value(field, value, True)
|
|
|
|
|
|
|
|
|
|
# try to convert to number
|
2021-04-26 01:15:17 -04:00
|
|
|
|
val_num = float_or_none(value, default=self._get_field_setting(field, 'default'))
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
is_num = self._get_field_setting(field, 'convert') != 'string' and val_num is not None
|
|
|
|
|
if is_num:
|
|
|
|
|
value = val_num
|
|
|
|
|
|
|
|
|
|
return ((-10, 0) if value is None
|
|
|
|
|
else (1, value, 0) if not is_num # if a field has mixed strings and numbers, strings are sorted higher
|
|
|
|
|
else (0, -abs(value - limit), value - limit if reverse else limit - value) if closest
|
|
|
|
|
else (0, value, 0) if not reverse and (limit is None or value <= limit)
|
|
|
|
|
else (0, -value, 0) if limit is None or (reverse and value == limit) or value > limit
|
|
|
|
|
else (-1, value, 0))
|
|
|
|
|
|
|
|
|
|
def _calculate_field_preference(self, format, field):
|
|
|
|
|
type = self._get_field_setting(field, 'type') # extractor, boolean, ordered, field, multiple
|
|
|
|
|
get_value = lambda f: format.get(self._get_field_setting(f, 'field'))
|
|
|
|
|
if type == 'multiple':
|
|
|
|
|
type = 'field' # Only 'field' is allowed in multiple for now
|
|
|
|
|
actual_fields = self._get_field_setting(field, 'field')
|
|
|
|
|
|
2021-07-31 06:29:52 -04:00
|
|
|
|
value = self._get_field_setting(field, 'function')(get_value(f) for f in actual_fields)
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
else:
|
|
|
|
|
value = get_value(field)
|
|
|
|
|
return self._calculate_field_preference_from_value(format, field, type, value)
|
|
|
|
|
|
|
|
|
|
def calculate_preference(self, format):
|
|
|
|
|
# Determine missing protocol
|
|
|
|
|
if not format.get('protocol'):
|
|
|
|
|
format['protocol'] = determine_protocol(format)
|
|
|
|
|
|
|
|
|
|
# Determine missing ext
|
|
|
|
|
if not format.get('ext') and 'url' in format:
|
|
|
|
|
format['ext'] = determine_ext(format['url'])
|
|
|
|
|
if format.get('vcodec') == 'none':
|
2021-06-12 16:02:19 -04:00
|
|
|
|
format['audio_ext'] = format['ext'] if format.get('acodec') != 'none' else 'none'
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
format['video_ext'] = 'none'
|
|
|
|
|
else:
|
|
|
|
|
format['video_ext'] = format['ext']
|
|
|
|
|
format['audio_ext'] = 'none'
|
|
|
|
|
# if format.get('preference') is None and format.get('ext') in ('f4f', 'f4m'): # Not supported?
|
|
|
|
|
# format['preference'] = -1000
|
|
|
|
|
|
|
|
|
|
# Determine missing bitrates
|
|
|
|
|
if format.get('tbr') is None:
|
|
|
|
|
if format.get('vbr') is not None and format.get('abr') is not None:
|
|
|
|
|
format['tbr'] = format.get('vbr', 0) + format.get('abr', 0)
|
|
|
|
|
else:
|
2021-10-08 20:23:15 -04:00
|
|
|
|
if format.get('vcodec') != 'none' and format.get('vbr') is None:
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
format['vbr'] = format.get('tbr') - format.get('abr', 0)
|
2021-10-08 20:23:15 -04:00
|
|
|
|
if format.get('acodec') != 'none' and format.get('abr') is None:
|
Better Format Sorting (Squashed)
* Added --format-sort (-S height,filesize)
* Made fields reversible (-S +height)
* Added --format-sort-force, --no-format-sort-force
* Added limit (-S height:720)
* Added codec preference (-S vcodec,acodec)
* Correct handling of preference<-1000
* Rebased to yt-dlc
* Automatically determine missing bitrates
* aext, vext, protocol, acodec, vcodec can now takes priority as string, not number (-S vext:webm)
* Correct handling of None in codec, audio_codec (None means the codec is unknown while 'none' means it doesn't exist)
* Correctly parse filesize (-S filesize:200M)
* Generalized preference calculation
* Rewrote entire code into the class FormatSort
* Correctly handle user input errors
* Combined fields (-S +ext:webm:webm)
* Closest mode (-S filesize~50M)
* Aliases (framerate=fps, br=bitrate etc)
* Documentation
2020-10-26 11:50:09 -04:00
|
|
|
|
format['abr'] = format.get('tbr') - format.get('vbr', 0)
|
|
|
|
|
|
|
|
|
|
return tuple(self._calculate_field_preference(format, field) for field in self._order)
|
|
|
|
|
|
|
|
|
|
def _sort_formats(self, formats, field_preference=[]):
|
|
|
|
|
if not formats:
|
2021-08-22 16:08:38 -04:00
|
|
|
|
return
|
2022-04-28 21:48:36 -04:00
|
|
|
|
formats.sort(key=self.FormatSort(self, field_preference).calculate_preference)
|
2013-11-20 00:13:19 -05:00
|
|
|
|
|
2015-01-25 13:32:31 -05:00
|
|
|
|
def _check_formats(self, formats, video_id):
|
|
|
|
|
if formats:
|
|
|
|
|
formats[:] = filter(
|
|
|
|
|
lambda f: self._is_valid_url(
|
|
|
|
|
f['url'], video_id,
|
|
|
|
|
item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
|
|
|
|
|
formats)
|
|
|
|
|
|
2016-02-21 14:19:39 -05:00
|
|
|
|
@staticmethod
|
|
|
|
|
def _remove_duplicate_formats(formats):
|
|
|
|
|
format_urls = set()
|
|
|
|
|
unique_formats = []
|
|
|
|
|
for f in formats:
|
|
|
|
|
if f['url'] not in format_urls:
|
|
|
|
|
format_urls.add(f['url'])
|
|
|
|
|
unique_formats.append(f)
|
|
|
|
|
formats[:] = unique_formats
|
|
|
|
|
|
2017-02-02 23:10:13 -05:00
|
|
|
|
def _is_valid_url(self, url, video_id, item='video', headers={}):
|
2015-03-02 11:38:44 -05:00
|
|
|
|
url = self._proto_relative_url(url, scheme='http:')
|
|
|
|
|
# For now assume non HTTP(S) URLs always valid
|
|
|
|
|
if not (url.startswith('http://') or url.startswith('https://')):
|
|
|
|
|
return True
|
2015-01-25 13:32:31 -05:00
|
|
|
|
try:
|
2017-02-02 23:10:13 -05:00
|
|
|
|
self._request_webpage(url, video_id, 'Checking %s URL' % item, headers=headers)
|
2015-01-25 13:32:31 -05:00
|
|
|
|
return True
|
2020-11-19 14:22:59 -05:00
|
|
|
|
except ExtractorError as e:
|
2019-10-02 13:53:07 -04:00
|
|
|
|
self.to_screen(
|
2020-11-19 14:22:59 -05:00
|
|
|
|
'%s: %s URL is invalid, skipping: %s'
|
|
|
|
|
% (video_id, item, error_to_compat_str(e.cause)))
|
2019-10-02 13:53:07 -04:00
|
|
|
|
return False
|
2015-01-25 13:32:31 -05:00
|
|
|
|
|
2014-05-04 21:12:41 -04:00
|
|
|
|
def http_scheme(self):
|
2014-10-24 09:34:19 -04:00
|
|
|
|
""" Either "http:" or "https:", depending on the user's preferences """
|
2014-05-04 21:12:41 -04:00
|
|
|
|
return (
|
|
|
|
|
'http:'
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if self.get_param('prefer_insecure', False)
|
2014-05-04 21:12:41 -04:00
|
|
|
|
else 'https:')
|
|
|
|
|
|
2014-05-13 03:42:38 -04:00
|
|
|
|
def _proto_relative_url(self, url, scheme=None):
|
2022-07-31 21:22:03 -04:00
|
|
|
|
scheme = scheme or self.http_scheme()
|
|
|
|
|
assert scheme.endswith(':')
|
|
|
|
|
return sanitize_url(url, scheme=scheme[:-1])
|
2014-05-13 03:42:38 -04:00
|
|
|
|
|
2014-07-11 04:57:08 -04:00
|
|
|
|
def _sleep(self, timeout, video_id, msg_template=None):
|
|
|
|
|
if msg_template is None:
|
2014-08-27 19:04:43 -04:00
|
|
|
|
msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
|
2014-07-11 04:57:08 -04:00
|
|
|
|
msg = msg_template % {'video_id': video_id, 'timeout': timeout}
|
|
|
|
|
self.to_screen(msg)
|
|
|
|
|
time.sleep(timeout)
|
|
|
|
|
|
2021-02-18 17:03:16 -05:00
|
|
|
|
def _extract_f4m_formats(self, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
|
2015-10-01 13:03:31 -04:00
|
|
|
|
transform_source=lambda s: fix_xml_ampersands(s).strip(),
|
2019-11-15 17:44:14 -05:00
|
|
|
|
fatal=True, m3u8_id=None, data=None, headers={}, query={}):
|
2022-04-27 18:50:01 -04:00
|
|
|
|
res = self._download_xml_handle(
|
2014-07-28 09:42:19 -04:00
|
|
|
|
manifest_url, video_id, 'Downloading f4m manifest',
|
2015-07-15 15:14:08 -04:00
|
|
|
|
'Unable to download f4m manifest',
|
|
|
|
|
# Some manifests may be malformed, e.g. prosiebensat1 generated manifests
|
2019-03-09 07:14:41 -05:00
|
|
|
|
# (see https://github.com/ytdl-org/youtube-dl/issues/6215#issuecomment-121704244)
|
2015-10-01 13:03:31 -04:00
|
|
|
|
transform_source=transform_source,
|
2019-11-15 17:44:14 -05:00
|
|
|
|
fatal=fatal, data=data, headers=headers, query=query)
|
2022-04-27 18:50:01 -04:00
|
|
|
|
if res is False:
|
2015-12-27 09:33:39 -05:00
|
|
|
|
return []
|
2014-07-28 09:25:56 -04:00
|
|
|
|
|
2022-04-27 18:50:01 -04:00
|
|
|
|
manifest, urlh = res
|
|
|
|
|
manifest_url = urlh.geturl()
|
|
|
|
|
|
2016-03-12 16:16:08 -05:00
|
|
|
|
return self._parse_f4m_formats(
|
2021-02-18 17:03:16 -05:00
|
|
|
|
manifest, manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
|
2016-05-26 12:03:03 -04:00
|
|
|
|
transform_source=transform_source, fatal=fatal, m3u8_id=m3u8_id)
|
2016-03-12 16:16:08 -05:00
|
|
|
|
|
2021-02-18 17:03:16 -05:00
|
|
|
|
def _parse_f4m_formats(self, manifest, manifest_url, video_id, preference=None, quality=None, f4m_id=None,
|
2016-03-12 16:16:08 -05:00
|
|
|
|
transform_source=lambda s: fix_xml_ampersands(s).strip(),
|
2016-05-26 12:03:03 -04:00
|
|
|
|
fatal=True, m3u8_id=None):
|
2022-04-11 16:09:26 -04:00
|
|
|
|
if not isinstance(manifest, xml.etree.ElementTree.Element) and not fatal:
|
2019-03-05 11:45:40 -05:00
|
|
|
|
return []
|
|
|
|
|
|
2021-02-24 13:45:56 -05:00
|
|
|
|
# currently yt-dlp cannot decode the playerVerificationChallenge as Akamai uses Adobe Alchemy
|
2015-07-30 12:34:38 -04:00
|
|
|
|
akamai_pv = manifest.find('{http://ns.adobe.com/f4m/1.0}pv-2.0')
|
|
|
|
|
if akamai_pv is not None and ';' in akamai_pv.text:
|
|
|
|
|
playerVerificationChallenge = akamai_pv.text.split(';')[0]
|
|
|
|
|
if playerVerificationChallenge.strip() != '':
|
|
|
|
|
return []
|
|
|
|
|
|
2014-07-28 09:25:56 -04:00
|
|
|
|
formats = []
|
2014-10-24 00:17:39 -04:00
|
|
|
|
manifest_version = '1.0'
|
2014-08-25 07:03:08 -04:00
|
|
|
|
media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
|
2014-10-23 17:11:10 -04:00
|
|
|
|
if not media_nodes:
|
2014-10-24 00:17:39 -04:00
|
|
|
|
manifest_version = '2.0'
|
2014-10-23 17:11:10 -04:00
|
|
|
|
media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
|
2016-03-26 21:42:38 -04:00
|
|
|
|
# Remove unsupported DRM protected media from final formats
|
2019-03-09 07:14:41 -05:00
|
|
|
|
# rendition (see https://github.com/ytdl-org/youtube-dl/issues/8573).
|
2016-03-26 21:42:38 -04:00
|
|
|
|
media_nodes = remove_encrypted_media(media_nodes)
|
|
|
|
|
if not media_nodes:
|
|
|
|
|
return formats
|
2017-11-04 11:10:55 -04:00
|
|
|
|
|
|
|
|
|
manifest_base_url = get_base_url(manifest)
|
2016-05-26 09:41:47 -04:00
|
|
|
|
|
2016-06-07 12:19:33 -04:00
|
|
|
|
bootstrap_info = xpath_element(
|
2016-05-26 09:41:47 -04:00
|
|
|
|
manifest, ['{http://ns.adobe.com/f4m/1.0}bootstrapInfo', '{http://ns.adobe.com/f4m/2.0}bootstrapInfo'],
|
|
|
|
|
'bootstrap info', default=None)
|
|
|
|
|
|
2016-10-19 09:42:48 -04:00
|
|
|
|
vcodec = None
|
|
|
|
|
mime_type = xpath_text(
|
|
|
|
|
manifest, ['{http://ns.adobe.com/f4m/1.0}mimeType', '{http://ns.adobe.com/f4m/2.0}mimeType'],
|
|
|
|
|
'base URL', default=None)
|
|
|
|
|
if mime_type and mime_type.startswith('audio/'):
|
|
|
|
|
vcodec = 'none'
|
|
|
|
|
|
2014-08-25 07:03:08 -04:00
|
|
|
|
for i, media_el in enumerate(media_nodes):
|
2016-05-26 15:47:44 -04:00
|
|
|
|
tbr = int_or_none(media_el.attrib.get('bitrate'))
|
|
|
|
|
width = int_or_none(media_el.attrib.get('width'))
|
|
|
|
|
height = int_or_none(media_el.attrib.get('height'))
|
2021-11-05 21:05:24 -04:00
|
|
|
|
format_id = join_nonempty(f4m_id, tbr or i)
|
2016-05-26 12:03:03 -04:00
|
|
|
|
# If <bootstrapInfo> is present, the specified f4m is a
|
|
|
|
|
# stream-level manifest, and only set-level manifests may refer to
|
|
|
|
|
# external resources. See section 11.4 and section 4 of F4M spec
|
|
|
|
|
if bootstrap_info is None:
|
|
|
|
|
media_url = None
|
|
|
|
|
# @href is introduced in 2.0, see section 11.6 of F4M spec
|
|
|
|
|
if manifest_version == '2.0':
|
|
|
|
|
media_url = media_el.attrib.get('href')
|
|
|
|
|
if media_url is None:
|
|
|
|
|
media_url = media_el.attrib.get('url')
|
2015-07-15 15:25:33 -04:00
|
|
|
|
if not media_url:
|
|
|
|
|
continue
|
2015-07-15 15:14:52 -04:00
|
|
|
|
manifest_url = (
|
|
|
|
|
media_url if media_url.startswith('http://') or media_url.startswith('https://')
|
2017-11-04 11:10:55 -04:00
|
|
|
|
else ((manifest_base_url or '/'.join(manifest_url.split('/')[:-1])) + '/' + media_url))
|
2015-07-15 15:15:15 -04:00
|
|
|
|
# If media_url is itself a f4m manifest do the recursive extraction
|
|
|
|
|
# since bitrates in parent manifest (this one) and media_url manifest
|
|
|
|
|
# may differ leading to inability to resolve the format by requested
|
|
|
|
|
# bitrate in f4m downloader
|
2016-05-26 09:55:43 -04:00
|
|
|
|
ext = determine_ext(manifest_url)
|
|
|
|
|
if ext == 'f4m':
|
2016-05-26 15:47:44 -04:00
|
|
|
|
f4m_formats = self._extract_f4m_formats(
|
2021-02-18 17:03:16 -05:00
|
|
|
|
manifest_url, video_id, preference=preference, quality=quality, f4m_id=f4m_id,
|
2016-05-26 15:47:44 -04:00
|
|
|
|
transform_source=transform_source, fatal=fatal)
|
|
|
|
|
# Sometimes stream-level manifest contains single media entry that
|
|
|
|
|
# does not contain any quality metadata (e.g. http://matchtv.ru/#live-player).
|
|
|
|
|
# At the same time parent's media entry in set-level manifest may
|
|
|
|
|
# contain it. We will copy it from parent in such cases.
|
|
|
|
|
if len(f4m_formats) == 1:
|
|
|
|
|
f = f4m_formats[0]
|
|
|
|
|
f.update({
|
|
|
|
|
'tbr': f.get('tbr') or tbr,
|
|
|
|
|
'width': f.get('width') or width,
|
|
|
|
|
'height': f.get('height') or height,
|
|
|
|
|
'format_id': f.get('format_id') if not tbr else format_id,
|
2016-10-19 09:42:48 -04:00
|
|
|
|
'vcodec': vcodec,
|
2016-05-26 15:47:44 -04:00
|
|
|
|
})
|
|
|
|
|
formats.extend(f4m_formats)
|
2015-07-15 15:15:15 -04:00
|
|
|
|
continue
|
2016-05-26 09:55:43 -04:00
|
|
|
|
elif ext == 'm3u8':
|
|
|
|
|
formats.extend(self._extract_m3u8_formats(
|
|
|
|
|
manifest_url, video_id, 'mp4', preference=preference,
|
2021-02-18 17:03:16 -05:00
|
|
|
|
quality=quality, m3u8_id=m3u8_id, fatal=fatal))
|
2016-05-26 09:55:43 -04:00
|
|
|
|
continue
|
2014-07-28 09:25:56 -04:00
|
|
|
|
formats.append({
|
2016-05-26 15:47:44 -04:00
|
|
|
|
'format_id': format_id,
|
2014-07-28 09:25:56 -04:00
|
|
|
|
'url': manifest_url,
|
2016-09-17 10:33:38 -04:00
|
|
|
|
'manifest_url': manifest_url,
|
2016-06-07 12:19:33 -04:00
|
|
|
|
'ext': 'flv' if bootstrap_info is not None else None,
|
2017-11-04 11:11:39 -04:00
|
|
|
|
'protocol': 'f4m',
|
2014-08-25 07:03:08 -04:00
|
|
|
|
'tbr': tbr,
|
2016-05-26 15:47:44 -04:00
|
|
|
|
'width': width,
|
|
|
|
|
'height': height,
|
2016-10-19 09:42:48 -04:00
|
|
|
|
'vcodec': vcodec,
|
2015-02-05 11:16:27 -05:00
|
|
|
|
'preference': preference,
|
2021-02-18 17:03:16 -05:00
|
|
|
|
'quality': quality,
|
2014-07-28 09:25:56 -04:00
|
|
|
|
})
|
|
|
|
|
return formats
|
|
|
|
|
|
2021-02-18 17:03:16 -05:00
|
|
|
|
def _m3u8_meta_format(self, m3u8_url, ext=None, preference=None, quality=None, m3u8_id=None):
|
2016-05-21 01:15:28 -04:00
|
|
|
|
return {
|
2021-11-05 21:05:24 -04:00
|
|
|
|
'format_id': join_nonempty(m3u8_id, 'meta'),
|
2014-08-26 06:51:13 -04:00
|
|
|
|
'url': m3u8_url,
|
|
|
|
|
'ext': ext,
|
|
|
|
|
'protocol': 'm3u8',
|
2016-08-07 05:58:11 -04:00
|
|
|
|
'preference': preference - 100 if preference else -100,
|
2021-02-18 17:03:16 -05:00
|
|
|
|
'quality': quality,
|
2014-08-26 06:51:13 -04:00
|
|
|
|
'resolution': 'multiple',
|
|
|
|
|
'format_note': 'Quality selection URL',
|
2016-05-21 01:15:28 -04:00
|
|
|
|
}
|
|
|
|
|
|
2021-10-08 20:23:15 -04:00
|
|
|
|
def _report_ignoring_subs(self, name):
|
|
|
|
|
self.report_warning(bug_reports_message(
|
|
|
|
|
f'Ignoring subtitle tracks found in the {name} manifest; '
|
|
|
|
|
'if any subtitle tracks are missing,'
|
|
|
|
|
), only_once=True)
|
|
|
|
|
|
2016-11-07 09:45:42 -05:00
|
|
|
|
def _extract_m3u8_formats(self, *args, **kwargs):
|
|
|
|
|
fmts, subs = self._extract_m3u8_formats_and_subtitles(*args, **kwargs)
|
|
|
|
|
if subs:
|
2021-10-08 20:23:15 -04:00
|
|
|
|
self._report_ignoring_subs('HLS')
|
2016-11-07 09:45:42 -05:00
|
|
|
|
return fmts
|
|
|
|
|
|
|
|
|
|
def _extract_m3u8_formats_and_subtitles(
|
2021-05-22 14:28:11 -04:00
|
|
|
|
self, m3u8_url, video_id, ext=None, entry_protocol='m3u8_native',
|
2016-11-07 09:45:42 -05:00
|
|
|
|
preference=None, quality=None, m3u8_id=None, note=None,
|
|
|
|
|
errnote=None, fatal=True, live=False, data=None, headers={},
|
|
|
|
|
query={}):
|
|
|
|
|
|
2015-10-31 14:01:34 -04:00
|
|
|
|
res = self._download_webpage_handle(
|
2014-10-26 21:28:37 -04:00
|
|
|
|
m3u8_url, video_id,
|
2021-05-29 04:52:44 -04:00
|
|
|
|
note='Downloading m3u8 information' if note is None else note,
|
|
|
|
|
errnote='Failed to download m3u8 information' if errnote is None else errnote,
|
2019-11-15 17:44:14 -05:00
|
|
|
|
fatal=fatal, data=data, headers=headers, query=query)
|
2017-04-21 20:01:00 -04:00
|
|
|
|
|
2015-10-31 14:01:34 -04:00
|
|
|
|
if res is False:
|
2016-11-07 09:45:42 -05:00
|
|
|
|
return [], {}
|
2017-04-21 20:01:00 -04:00
|
|
|
|
|
2015-10-31 14:01:34 -04:00
|
|
|
|
m3u8_doc, urlh = res
|
2015-09-08 14:35:41 -04:00
|
|
|
|
m3u8_url = urlh.geturl()
|
2016-02-26 20:01:11 -05:00
|
|
|
|
|
2016-11-07 09:45:42 -05:00
|
|
|
|
return self._parse_m3u8_formats_and_subtitles(
|
2017-04-21 20:01:00 -04:00
|
|
|
|
m3u8_doc, m3u8_url, ext=ext, entry_protocol=entry_protocol,
|
2021-02-24 09:47:53 -05:00
|
|
|
|
preference=preference, quality=quality, m3u8_id=m3u8_id,
|
|
|
|
|
note=note, errnote=errnote, fatal=fatal, live=live, data=data,
|
|
|
|
|
headers=headers, query=query, video_id=video_id)
|
2017-04-21 20:01:00 -04:00
|
|
|
|
|
2016-11-07 09:45:42 -05:00
|
|
|
|
def _parse_m3u8_formats_and_subtitles(
|
2022-01-19 17:23:55 -05:00
|
|
|
|
self, m3u8_doc, m3u8_url=None, ext=None, entry_protocol='m3u8_native',
|
2016-11-07 09:45:42 -05:00
|
|
|
|
preference=None, quality=None, m3u8_id=None, live=False, note=None,
|
|
|
|
|
errnote=None, fatal=True, data=None, headers={}, query={},
|
|
|
|
|
video_id=None):
|
2021-07-06 16:54:58 -04:00
|
|
|
|
formats, subtitles = [], {}
|
2016-11-07 09:45:42 -05:00
|
|
|
|
|
2021-11-18 20:49:51 -05:00
|
|
|
|
has_drm = re.search('|'.join([
|
|
|
|
|
r'#EXT-X-FAXS-CM:', # Adobe Flash Access
|
|
|
|
|
r'#EXT-X-(?:SESSION-)?KEY:.*?URI="skd://', # Apple FairPlay
|
|
|
|
|
]), m3u8_doc)
|
2016-11-07 09:45:42 -05:00
|
|
|
|
|
2021-07-06 16:54:58 -04:00
|
|
|
|
def format_url(url):
|
2022-06-24 06:54:43 -04:00
|
|
|
|
return url if re.match(r'^https?://', url) else urllib.parse.urljoin(m3u8_url, url)
|
2021-07-06 16:54:58 -04:00
|
|
|
|
|
|
|
|
|
if self.get_param('hls_split_discontinuity', False):
|
|
|
|
|
def _extract_m3u8_playlist_indices(manifest_url=None, m3u8_doc=None):
|
|
|
|
|
if not m3u8_doc:
|
|
|
|
|
if not manifest_url:
|
|
|
|
|
return []
|
|
|
|
|
m3u8_doc = self._download_webpage(
|
|
|
|
|
manifest_url, video_id, fatal=fatal, data=data, headers=headers,
|
|
|
|
|
note=False, errnote='Failed to download m3u8 playlist information')
|
|
|
|
|
if m3u8_doc is False:
|
|
|
|
|
return []
|
|
|
|
|
return range(1 + sum(line.startswith('#EXT-X-DISCONTINUITY') for line in m3u8_doc.splitlines()))
|
2016-09-04 06:42:15 -04:00
|
|
|
|
|
2021-07-06 16:54:58 -04:00
|
|
|
|
else:
|
|
|
|
|
def _extract_m3u8_playlist_indices(*args, **kwargs):
|
|
|
|
|
return [None]
|
2021-02-24 09:47:53 -05:00
|
|
|
|
|
2017-04-21 20:01:00 -04:00
|
|
|
|
# References:
|
|
|
|
|
# 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-21
|
2019-03-09 07:14:41 -05:00
|
|
|
|
# 2. https://github.com/ytdl-org/youtube-dl/issues/12211
|
|
|
|
|
# 3. https://github.com/ytdl-org/youtube-dl/issues/18923
|
2017-04-21 20:01:00 -04:00
|
|
|
|
|
|
|
|
|
# We should try extracting formats only from master playlists [1, 4.3.4],
|
|
|
|
|
# i.e. playlists that describe available qualities. On the other hand
|
|
|
|
|
# media playlists [1, 4.3.3] should be returned as is since they contain
|
|
|
|
|
# just the media without qualities renditions.
|
2016-02-26 20:01:11 -05:00
|
|
|
|
# Fortunately, master playlist can be easily distinguished from media
|
2017-04-21 20:01:00 -04:00
|
|
|
|
# playlist based on particular tags availability. As of [1, 4.3.3, 4.3.4]
|
2020-11-21 09:50:42 -05:00
|
|
|
|
# master playlist tags MUST NOT appear in a media playlist and vice versa.
|
2017-04-21 20:01:00 -04:00
|
|
|
|
# As of [1, 4.3.3.1] #EXT-X-TARGETDURATION tag is REQUIRED for every
|
|
|
|
|
# media playlist and MUST NOT appear in master playlist thus we can
|
|
|
|
|
# clearly detect media playlist with this criterion.
|
|
|
|
|
|
2016-02-26 20:01:11 -05:00
|
|
|
|
if '#EXT-X-TARGETDURATION' in m3u8_doc: # media playlist, return as is
|
2021-07-06 16:54:58 -04:00
|
|
|
|
formats = [{
|
2021-11-05 21:05:24 -04:00
|
|
|
|
'format_id': join_nonempty(m3u8_id, idx),
|
2021-07-06 16:54:58 -04:00
|
|
|
|
'format_index': idx,
|
2022-01-19 17:23:55 -05:00
|
|
|
|
'url': m3u8_url or encode_data_uri(m3u8_doc.encode('utf-8'), 'application/x-mpegurl'),
|
2021-07-06 16:54:58 -04:00
|
|
|
|
'ext': ext,
|
|
|
|
|
'protocol': entry_protocol,
|
|
|
|
|
'preference': preference,
|
|
|
|
|
'quality': quality,
|
2021-08-22 16:08:38 -04:00
|
|
|
|
'has_drm': has_drm,
|
2021-07-06 16:54:58 -04:00
|
|
|
|
} for idx in _extract_m3u8_playlist_indices(m3u8_doc=m3u8_doc)]
|
2021-02-24 09:47:53 -05:00
|
|
|
|
|
2016-11-07 09:45:42 -05:00
|
|
|
|
return formats, subtitles
|
2017-04-21 20:01:00 -04:00
|
|
|
|
|
|
|
|
|
groups = {}
|
|
|
|
|
last_stream_inf = {}
|
|
|
|
|
|
|
|
|
|
def extract_media(x_media_line):
|
|
|
|
|
media = parse_m3u8_attributes(x_media_line)
|
|
|
|
|
# As per [1, 4.3.4.1] TYPE, GROUP-ID and NAME are REQUIRED
|
|
|
|
|
media_type, group_id, name = media.get('TYPE'), media.get('GROUP-ID'), media.get('NAME')
|
|
|
|
|
if not (media_type and group_id and name):
|
|
|
|
|
return
|
|
|
|
|
groups.setdefault(group_id, []).append(media)
|
2016-11-07 09:45:42 -05:00
|
|
|
|
# <https://tools.ietf.org/html/rfc8216#section-4.3.4.1>
|
|
|
|
|
if media_type == 'SUBTITLES':
|
2021-05-23 12:59:28 -04:00
|
|
|
|
# According to RFC 8216 §4.3.4.2.1, URI is REQUIRED in the
|
|
|
|
|
# EXT-X-MEDIA tag if the media type is SUBTITLES.
|
|
|
|
|
# However, lack of URI has been spotted in the wild.
|
|
|
|
|
# e.g. NebulaIE; see https://github.com/yt-dlp/yt-dlp/issues/339
|
|
|
|
|
if not media.get('URI'):
|
|
|
|
|
return
|
2016-11-07 09:45:42 -05:00
|
|
|
|
url = format_url(media['URI'])
|
|
|
|
|
sub_info = {
|
|
|
|
|
'url': url,
|
|
|
|
|
'ext': determine_ext(url),
|
|
|
|
|
}
|
2021-04-28 06:47:30 -04:00
|
|
|
|
if sub_info['ext'] == 'm3u8':
|
|
|
|
|
# Per RFC 8216 §3.1, the only possible subtitle format m3u8
|
|
|
|
|
# files may contain is WebVTT:
|
|
|
|
|
# <https://tools.ietf.org/html/rfc8216#section-3.1>
|
|
|
|
|
sub_info['ext'] = 'vtt'
|
|
|
|
|
sub_info['protocol'] = 'm3u8_native'
|
2021-05-29 04:52:44 -04:00
|
|
|
|
lang = media.get('LANGUAGE') or 'und'
|
2016-11-07 09:45:42 -05:00
|
|
|
|
subtitles.setdefault(lang, []).append(sub_info)
|
2017-04-21 20:01:00 -04:00
|
|
|
|
if media_type not in ('VIDEO', 'AUDIO'):
|
|
|
|
|
return
|
|
|
|
|
media_url = media.get('URI')
|
|
|
|
|
if media_url:
|
2021-02-24 09:47:53 -05:00
|
|
|
|
manifest_url = format_url(media_url)
|
2021-07-06 16:54:58 -04:00
|
|
|
|
formats.extend({
|
2021-11-05 21:05:24 -04:00
|
|
|
|
'format_id': join_nonempty(m3u8_id, group_id, name, idx),
|
2021-07-06 16:54:58 -04:00
|
|
|
|
'format_note': name,
|
|
|
|
|
'format_index': idx,
|
|
|
|
|
'url': manifest_url,
|
|
|
|
|
'manifest_url': m3u8_url,
|
|
|
|
|
'language': media.get('LANGUAGE'),
|
|
|
|
|
'ext': ext,
|
|
|
|
|
'protocol': entry_protocol,
|
|
|
|
|
'preference': preference,
|
|
|
|
|
'quality': quality,
|
|
|
|
|
'vcodec': 'none' if media_type == 'AUDIO' else None,
|
|
|
|
|
} for idx in _extract_m3u8_playlist_indices(manifest_url))
|
2017-04-21 20:01:00 -04:00
|
|
|
|
|
|
|
|
|
def build_stream_name():
|
|
|
|
|
# Despite specification does not mention NAME attribute for
|
2017-04-23 00:51:53 -04:00
|
|
|
|
# EXT-X-STREAM-INF tag it still sometimes may be present (see [1]
|
|
|
|
|
# or vidio test in TestInfoExtractor.test_parse_m3u8_formats)
|
2017-04-23 00:49:57 -04:00
|
|
|
|
# 1. http://www.vidio.com/watch/165683-dj_ambred-booyah-live-2015
|
2017-04-21 20:01:00 -04:00
|
|
|
|
stream_name = last_stream_inf.get('NAME')
|
|
|
|
|
if stream_name:
|
|
|
|
|
return stream_name
|
|
|
|
|
# If there is no NAME in EXT-X-STREAM-INF it will be obtained
|
|
|
|
|
# from corresponding rendition group
|
|
|
|
|
stream_group_id = last_stream_inf.get('VIDEO')
|
|
|
|
|
if not stream_group_id:
|
|
|
|
|
return
|
|
|
|
|
stream_group = groups.get(stream_group_id)
|
|
|
|
|
if not stream_group:
|
|
|
|
|
return stream_group_id
|
|
|
|
|
rendition = stream_group[0]
|
|
|
|
|
return rendition.get('NAME') or stream_group_id
|
|
|
|
|
|
2019-01-19 15:35:02 -05:00
|
|
|
|
# parse EXT-X-MEDIA tags before EXT-X-STREAM-INF in order to have the
|
2019-01-19 15:25:15 -05:00
|
|
|
|
# chance to detect video only formats when EXT-X-STREAM-INF tags
|
|
|
|
|
# precede EXT-X-MEDIA tags in HLS manifest such as [3].
|
|
|
|
|
for line in m3u8_doc.splitlines():
|
|
|
|
|
if line.startswith('#EXT-X-MEDIA:'):
|
|
|
|
|
extract_media(line)
|
|
|
|
|
|
2014-08-26 06:51:13 -04:00
|
|
|
|
for line in m3u8_doc.splitlines():
|
|
|
|
|
if line.startswith('#EXT-X-STREAM-INF:'):
|
2017-04-21 20:01:00 -04:00
|
|
|
|
last_stream_inf = parse_m3u8_attributes(line)
|
2014-08-26 06:51:13 -04:00
|
|
|
|
elif line.startswith('#') or not line.strip():
|
|
|
|
|
continue
|
|
|
|
|
else:
|
2017-04-23 00:33:19 -04:00
|
|
|
|
tbr = float_or_none(
|
2019-05-10 16:56:22 -04:00
|
|
|
|
last_stream_inf.get('AVERAGE-BANDWIDTH')
|
|
|
|
|
or last_stream_inf.get('BANDWIDTH'), scale=1000)
|
2016-09-17 10:33:38 -04:00
|
|
|
|
manifest_url = format_url(line.strip())
|
2019-11-26 16:01:34 -05:00
|
|
|
|
|
2021-07-06 16:54:58 -04:00
|
|
|
|
for idx in _extract_m3u8_playlist_indices(manifest_url):
|
|
|
|
|
format_id = [m3u8_id, None, idx]
|
2021-02-24 09:47:53 -05:00
|
|
|
|
# Bandwidth of live streams may differ over time thus making
|
|
|
|
|
# format_id unpredictable. So it's better to keep provided
|
|
|
|
|
# format_id intact.
|
|
|
|
|
if not live:
|
2021-07-06 16:54:58 -04:00
|
|
|
|
stream_name = build_stream_name()
|
2021-11-05 21:05:24 -04:00
|
|
|
|
format_id[1] = stream_name or '%d' % (tbr or len(formats))
|
2021-02-24 09:47:53 -05:00
|
|
|
|
f = {
|
2021-11-05 21:05:24 -04:00
|
|
|
|
'format_id': join_nonempty(*format_id),
|
2021-07-06 16:54:58 -04:00
|
|
|
|
'format_index': idx,
|
2021-02-24 09:47:53 -05:00
|
|
|
|
'url': manifest_url,
|
|
|
|
|
'manifest_url': m3u8_url,
|
|
|
|
|
'tbr': tbr,
|
|
|
|
|
'ext': ext,
|
|
|
|
|
'fps': float_or_none(last_stream_inf.get('FRAME-RATE')),
|
|
|
|
|
'protocol': entry_protocol,
|
|
|
|
|
'preference': preference,
|
|
|
|
|
'quality': quality,
|
|
|
|
|
}
|
|
|
|
|
resolution = last_stream_inf.get('RESOLUTION')
|
|
|
|
|
if resolution:
|
|
|
|
|
mobj = re.search(r'(?P<width>\d+)[xX](?P<height>\d+)', resolution)
|
|
|
|
|
if mobj:
|
|
|
|
|
f['width'] = int(mobj.group('width'))
|
|
|
|
|
f['height'] = int(mobj.group('height'))
|
|
|
|
|
# Unified Streaming Platform
|
|
|
|
|
mobj = re.search(
|
|
|
|
|
r'audio.*?(?:%3D|=)(\d+)(?:-video.*?(?:%3D|=)(\d+))?', f['url'])
|
|
|
|
|
if mobj:
|
|
|
|
|
abr, vbr = mobj.groups()
|
|
|
|
|
abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000)
|
|
|
|
|
f.update({
|
|
|
|
|
'vbr': vbr,
|
|
|
|
|
'abr': abr,
|
|
|
|
|
})
|
|
|
|
|
codecs = parse_codecs(last_stream_inf.get('CODECS'))
|
|
|
|
|
f.update(codecs)
|
|
|
|
|
audio_group_id = last_stream_inf.get('AUDIO')
|
|
|
|
|
# As per [1, 4.3.4.1.1] any EXT-X-STREAM-INF tag which
|
|
|
|
|
# references a rendition group MUST have a CODECS attribute.
|
2022-08-14 08:04:13 -04:00
|
|
|
|
# However, this is not always respected. E.g. [2]
|
2021-02-24 09:47:53 -05:00
|
|
|
|
# contains EXT-X-STREAM-INF tag which references AUDIO
|
|
|
|
|
# rendition group but does not have CODECS and despite
|
|
|
|
|
# referencing an audio group it represents a complete
|
|
|
|
|
# (with audio and video) format. So, for such cases we will
|
|
|
|
|
# ignore references to rendition groups and treat them
|
|
|
|
|
# as complete formats.
|
|
|
|
|
if audio_group_id and codecs and f.get('vcodec') != 'none':
|
|
|
|
|
audio_group = groups.get(audio_group_id)
|
|
|
|
|
if audio_group and audio_group[0].get('URI'):
|
|
|
|
|
# TODO: update acodec for audio only formats with
|
|
|
|
|
# the same GROUP-ID
|
|
|
|
|
f['acodec'] = 'none'
|
2021-03-07 12:52:12 -05:00
|
|
|
|
if not f.get('ext'):
|
|
|
|
|
f['ext'] = 'm4a' if f.get('vcodec') == 'none' else 'mp4'
|
2021-02-24 09:47:53 -05:00
|
|
|
|
formats.append(f)
|
|
|
|
|
|
|
|
|
|
# for DailyMotion
|
|
|
|
|
progressive_uri = last_stream_inf.get('PROGRESSIVE-URI')
|
|
|
|
|
if progressive_uri:
|
|
|
|
|
http_f = f.copy()
|
|
|
|
|
del http_f['manifest_url']
|
|
|
|
|
http_f.update({
|
|
|
|
|
'format_id': f['format_id'].replace('hls-', 'http-'),
|
|
|
|
|
'protocol': 'http',
|
|
|
|
|
'url': progressive_uri,
|
|
|
|
|
})
|
|
|
|
|
formats.append(http_f)
|
2019-11-26 16:01:34 -05:00
|
|
|
|
|
2017-04-21 20:01:00 -04:00
|
|
|
|
last_stream_inf = {}
|
2016-11-07 09:45:42 -05:00
|
|
|
|
return formats, subtitles
|
2014-08-26 06:51:13 -04:00
|
|
|
|
|
2021-09-27 17:00:41 -04:00
|
|
|
|
def _extract_m3u8_vod_duration(
|
|
|
|
|
self, m3u8_vod_url, video_id, note=None, errnote=None, data=None, headers={}, query={}):
|
|
|
|
|
|
|
|
|
|
m3u8_vod = self._download_webpage(
|
|
|
|
|
m3u8_vod_url, video_id,
|
|
|
|
|
note='Downloading m3u8 VOD manifest' if note is None else note,
|
|
|
|
|
errnote='Failed to download VOD manifest' if errnote is None else errnote,
|
|
|
|
|
fatal=False, data=data, headers=headers, query=query)
|
|
|
|
|
|
|
|
|
|
return self._parse_m3u8_vod_duration(m3u8_vod or '', video_id)
|
|
|
|
|
|
|
|
|
|
def _parse_m3u8_vod_duration(self, m3u8_vod, video_id):
|
|
|
|
|
if '#EXT-X-PLAYLIST-TYPE:VOD' not in m3u8_vod:
|
|
|
|
|
return None
|
|
|
|
|
|
|
|
|
|
return int(sum(
|
|
|
|
|
float(line[len('#EXTINF:'):].split(',')[0])
|
|
|
|
|
for line in m3u8_vod.splitlines() if line.startswith('#EXTINF:'))) or None
|
|
|
|
|
|
2015-08-01 15:13:21 -04:00
|
|
|
|
@staticmethod
|
|
|
|
|
def _xpath_ns(path, namespace=None):
|
|
|
|
|
if not namespace:
|
|
|
|
|
return path
|
|
|
|
|
out = []
|
|
|
|
|
for c in path.split('/'):
|
|
|
|
|
if not c or c == '.':
|
|
|
|
|
out.append(c)
|
|
|
|
|
else:
|
|
|
|
|
out.append('{%s}%s' % (namespace, c))
|
|
|
|
|
return '/'.join(out)
|
|
|
|
|
|
2021-07-16 10:22:56 -04:00
|
|
|
|
def _extract_smil_formats_and_subtitles(self, smil_url, video_id, fatal=True, f4m_params=None, transform_source=None):
|
2022-04-27 18:50:01 -04:00
|
|
|
|
res = self._download_smil(smil_url, video_id, fatal=fatal, transform_source=transform_source)
|
|
|
|
|
if res is False:
|
2015-02-02 17:38:35 -05:00
|
|
|
|
assert not fatal
|
2021-12-23 21:15:48 -05:00
|
|
|
|
return [], {}
|
2014-12-09 11:28:28 -05:00
|
|
|
|
|
2022-04-27 18:50:01 -04:00
|
|
|
|
smil, urlh = res
|
|
|
|
|
smil_url = urlh.geturl()
|
|
|
|
|
|
2015-08-01 15:31:17 -04:00
|
|
|
|
namespace = self._parse_smil_namespace(smil)
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
2021-07-16 10:22:56 -04:00
|
|
|
|
fmts = self._parse_smil_formats(
|
2015-08-01 15:13:21 -04:00
|
|
|
|
smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
|
2021-07-16 10:22:56 -04:00
|
|
|
|
subs = self._parse_smil_subtitles(
|
|
|
|
|
smil, namespace=namespace)
|
|
|
|
|
|
|
|
|
|
return fmts, subs
|
|
|
|
|
|
|
|
|
|
def _extract_smil_formats(self, *args, **kwargs):
|
|
|
|
|
fmts, subs = self._extract_smil_formats_and_subtitles(*args, **kwargs)
|
|
|
|
|
if subs:
|
2021-10-08 20:23:15 -04:00
|
|
|
|
self._report_ignoring_subs('SMIL')
|
2021-07-16 10:22:56 -04:00
|
|
|
|
return fmts
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
|
|
|
|
def _extract_smil_info(self, smil_url, video_id, fatal=True, f4m_params=None):
|
2022-04-27 18:50:01 -04:00
|
|
|
|
res = self._download_smil(smil_url, video_id, fatal=fatal)
|
|
|
|
|
if res is False:
|
2015-08-01 15:13:21 -04:00
|
|
|
|
return {}
|
2022-04-27 18:50:01 -04:00
|
|
|
|
|
|
|
|
|
smil, urlh = res
|
|
|
|
|
smil_url = urlh.geturl()
|
|
|
|
|
|
2015-08-01 15:13:21 -04:00
|
|
|
|
return self._parse_smil(smil, smil_url, video_id, f4m_params=f4m_params)
|
|
|
|
|
|
2016-03-11 16:37:07 -05:00
|
|
|
|
def _download_smil(self, smil_url, video_id, fatal=True, transform_source=None):
|
2022-04-27 18:50:01 -04:00
|
|
|
|
return self._download_xml_handle(
|
2015-08-01 15:13:21 -04:00
|
|
|
|
smil_url, video_id, 'Downloading SMIL file',
|
2016-03-11 16:37:07 -05:00
|
|
|
|
'Unable to download SMIL file', fatal=fatal, transform_source=transform_source)
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
|
|
|
|
def _parse_smil(self, smil, smil_url, video_id, f4m_params=None):
|
2015-08-01 15:31:17 -04:00
|
|
|
|
namespace = self._parse_smil_namespace(smil)
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
|
|
|
|
formats = self._parse_smil_formats(
|
|
|
|
|
smil, smil_url, video_id, namespace=namespace, f4m_params=f4m_params)
|
|
|
|
|
subtitles = self._parse_smil_subtitles(smil, namespace=namespace)
|
|
|
|
|
|
|
|
|
|
video_id = os.path.splitext(url_basename(smil_url))[0]
|
|
|
|
|
title = None
|
|
|
|
|
description = None
|
2015-10-01 12:18:59 -04:00
|
|
|
|
upload_date = None
|
2015-08-01 15:13:21 -04:00
|
|
|
|
for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
|
|
|
|
|
name = meta.attrib.get('name')
|
|
|
|
|
content = meta.attrib.get('content')
|
|
|
|
|
if not name or not content:
|
|
|
|
|
continue
|
|
|
|
|
if not title and name == 'title':
|
|
|
|
|
title = content
|
|
|
|
|
elif not description and name in ('description', 'abstract'):
|
|
|
|
|
description = content
|
2015-10-01 12:18:59 -04:00
|
|
|
|
elif not upload_date and name == 'date':
|
|
|
|
|
upload_date = unified_strdate(content)
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
2015-10-01 12:08:16 -04:00
|
|
|
|
thumbnails = [{
|
|
|
|
|
'id': image.get('type'),
|
|
|
|
|
'url': image.get('src'),
|
|
|
|
|
'width': int_or_none(image.get('width')),
|
|
|
|
|
'height': int_or_none(image.get('height')),
|
|
|
|
|
} for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')]
|
|
|
|
|
|
2015-08-01 15:13:21 -04:00
|
|
|
|
return {
|
|
|
|
|
'id': video_id,
|
|
|
|
|
'title': title or video_id,
|
|
|
|
|
'description': description,
|
2015-10-01 12:18:59 -04:00
|
|
|
|
'upload_date': upload_date,
|
2015-10-01 12:08:16 -04:00
|
|
|
|
'thumbnails': thumbnails,
|
2015-08-01 15:13:21 -04:00
|
|
|
|
'formats': formats,
|
|
|
|
|
'subtitles': subtitles,
|
|
|
|
|
}
|
|
|
|
|
|
2015-08-01 15:31:17 -04:00
|
|
|
|
def _parse_smil_namespace(self, smil):
|
|
|
|
|
return self._search_regex(
|
|
|
|
|
r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None)
|
|
|
|
|
|
2015-08-19 11:11:25 -04:00
|
|
|
|
def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
|
2015-08-01 15:13:21 -04:00
|
|
|
|
base = smil_url
|
|
|
|
|
for meta in smil.findall(self._xpath_ns('./head/meta', namespace)):
|
|
|
|
|
b = meta.get('base') or meta.get('httpBase')
|
|
|
|
|
if b:
|
|
|
|
|
base = b
|
|
|
|
|
break
|
2014-12-09 11:28:28 -05:00
|
|
|
|
|
|
|
|
|
formats = []
|
|
|
|
|
rtmp_count = 0
|
2015-08-01 15:13:21 -04:00
|
|
|
|
http_count = 0
|
2016-01-26 11:44:44 -05:00
|
|
|
|
m3u8_count = 0
|
2021-10-02 14:43:42 -04:00
|
|
|
|
imgs_count = 0
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
2021-10-02 14:43:42 -04:00
|
|
|
|
srcs = set()
|
2016-05-20 07:02:53 -04:00
|
|
|
|
media = smil.findall(self._xpath_ns('.//video', namespace)) + smil.findall(self._xpath_ns('.//audio', namespace))
|
|
|
|
|
for medium in media:
|
|
|
|
|
src = medium.get('src')
|
2016-02-11 11:58:48 -05:00
|
|
|
|
if not src or src in srcs:
|
2015-08-01 15:13:21 -04:00
|
|
|
|
continue
|
2021-10-02 14:43:42 -04:00
|
|
|
|
srcs.add(src)
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
2016-05-20 07:02:53 -04:00
|
|
|
|
bitrate = float_or_none(medium.get('system-bitrate') or medium.get('systemBitrate'), 1000)
|
|
|
|
|
filesize = int_or_none(medium.get('size') or medium.get('fileSize'))
|
|
|
|
|
width = int_or_none(medium.get('width'))
|
|
|
|
|
height = int_or_none(medium.get('height'))
|
|
|
|
|
proto = medium.get('proto')
|
|
|
|
|
ext = medium.get('ext')
|
2015-08-01 15:13:21 -04:00
|
|
|
|
src_ext = determine_ext(src)
|
2016-05-20 07:02:53 -04:00
|
|
|
|
streamer = medium.get('streamer') or base
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
|
|
|
|
if proto == 'rtmp' or streamer.startswith('rtmp'):
|
|
|
|
|
rtmp_count += 1
|
|
|
|
|
formats.append({
|
|
|
|
|
'url': streamer,
|
|
|
|
|
'play_path': src,
|
|
|
|
|
'ext': 'flv',
|
|
|
|
|
'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
|
|
|
|
|
'tbr': bitrate,
|
|
|
|
|
'filesize': filesize,
|
|
|
|
|
'width': width,
|
|
|
|
|
'height': height,
|
|
|
|
|
})
|
2015-08-19 11:11:25 -04:00
|
|
|
|
if transform_rtmp_url:
|
|
|
|
|
streamer, src = transform_rtmp_url(streamer, src)
|
|
|
|
|
formats[-1].update({
|
|
|
|
|
'url': streamer,
|
|
|
|
|
'play_path': src,
|
|
|
|
|
})
|
2015-08-01 15:13:21 -04:00
|
|
|
|
continue
|
|
|
|
|
|
2022-06-24 06:54:43 -04:00
|
|
|
|
src_url = src if src.startswith('http') else urllib.parse.urljoin(base, src)
|
2016-02-12 11:38:48 -05:00
|
|
|
|
src_url = src_url.strip()
|
2015-08-01 15:13:21 -04:00
|
|
|
|
|
|
|
|
|
if proto == 'm3u8' or src_ext == 'm3u8':
|
2016-01-26 11:44:44 -05:00
|
|
|
|
m3u8_formats = self._extract_m3u8_formats(
|
|
|
|
|
src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False)
|
|
|
|
|
if len(m3u8_formats) == 1:
|
|
|
|
|
m3u8_count += 1
|
|
|
|
|
m3u8_formats[0].update({
|
|
|
|
|
'format_id': 'hls-%d' % (m3u8_count if bitrate is None else bitrate),
|
|
|
|
|
'tbr': bitrate,
|
|
|
|
|
'width': width,
|
|
|
|
|
'height': height,
|
|
|
|
|
})
|
|
|
|
|
formats.extend(m3u8_formats)
|
2018-07-18 13:29:18 -04:00
|
|
|
|
elif src_ext == 'f4m':
|
2015-08-01 15:13:21 -04:00
|
|
|
|
f4m_url = src_url
|
|
|
|
|
if not f4m_params:
|
|
|
|
|
f4m_params = {
|
|
|
|
|
'hdcore': '3.2.0',
|
|
|
|
|
'plugin': 'flowplayer-3.2.0.1',
|
|
|
|
|
}
|
|
|
|
|
f4m_url += '&' if '?' in f4m_url else '?'
|
2022-06-24 06:54:43 -04:00
|
|
|
|
f4m_url += urllib.parse.urlencode(f4m_params)
|
2015-12-28 13:58:24 -05:00
|
|
|
|
formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
|
2018-07-18 13:29:18 -04:00
|
|
|
|
elif src_ext == 'mpd':
|
|
|
|
|
formats.extend(self._extract_mpd_formats(
|
|
|
|
|
src_url, video_id, mpd_id='dash', fatal=False))
|
|
|
|
|
elif re.search(r'\.ism/[Mm]anifest', src_url):
|
|
|
|
|
formats.extend(self._extract_ism_formats(
|
|
|
|
|
src_url, video_id, ism_id='mss', fatal=False))
|
|
|
|
|
elif src_url.startswith('http') and self._is_valid_url(src, video_id):
|
2015-08-01 15:13:21 -04:00
|
|
|
|
http_count += 1
|
|
|
|
|
formats.append({
|
|
|
|
|
'url': src_url,
|
|
|
|
|
'ext': ext or src_ext or 'flv',
|
|
|
|
|
'format_id': 'http-%d' % (bitrate or http_count),
|
|
|
|
|
'tbr': bitrate,
|
|
|
|
|
'filesize': filesize,
|
|
|
|
|
'width': width,
|
|
|
|
|
'height': height,
|
|
|
|
|
})
|
2015-02-22 03:16:51 -05:00
|
|
|
|
|
2021-10-02 14:43:42 -04:00
|
|
|
|
for medium in smil.findall(self._xpath_ns('.//imagestream', namespace)):
|
|
|
|
|
src = medium.get('src')
|
|
|
|
|
if not src or src in srcs:
|
|
|
|
|
continue
|
|
|
|
|
srcs.add(src)
|
|
|
|
|
|
|
|
|
|
imgs_count += 1
|
|
|
|
|
formats.append({
|
|
|
|
|
'format_id': 'imagestream-%d' % (imgs_count),
|
|
|
|
|
'url': src,
|
|
|
|
|
'ext': mimetype2ext(medium.get('type')),
|
|
|
|
|
'acodec': 'none',
|
|
|
|
|
'vcodec': 'none',
|
|
|
|
|
'width': int_or_none(medium.get('width')),
|
|
|
|
|
'height': int_or_none(medium.get('height')),
|
|
|
|
|
'format_note': 'SMIL storyboards',
|
|
|
|
|
})
|
|
|
|
|
|
2014-12-09 11:28:28 -05:00
|
|
|
|
return formats
|
|
|
|
|
|
2015-08-19 14:56:17 -04:00
|
|
|
|
def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
|
2016-02-09 11:15:41 -05:00
|
|
|
|
urls = []
|
2015-08-01 15:13:21 -04:00
|
|
|
|
subtitles = {}
|
|
|
|
|
for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))):
|
|
|
|
|
src = textstream.get('src')
|
2016-02-09 11:15:41 -05:00
|
|
|
|
if not src or src in urls:
|
2015-08-01 15:13:21 -04:00
|
|
|
|
continue
|
2016-02-09 11:15:41 -05:00
|
|
|
|
urls.append(src)
|
2016-04-01 14:39:02 -04:00
|
|
|
|
ext = textstream.get('ext') or mimetype2ext(textstream.get('type')) or determine_ext(src)
|
2015-08-20 11:18:58 -04:00
|
|
|
|
lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang
|
2015-08-01 15:13:21 -04:00
|
|
|
|
subtitles.setdefault(lang, []).append({
|
|
|
|
|
'url': src,
|
|
|
|
|
'ext': ext,
|
|
|
|
|
})
|
|
|
|
|
return subtitles
|
2015-02-22 03:16:51 -05:00
|
|
|
|
|
2018-03-17 15:46:50 -04:00
|
|
|
|
def _extract_xspf_playlist(self, xspf_url, playlist_id, fatal=True):
|
2022-04-27 18:50:01 -04:00
|
|
|
|
res = self._download_xml_handle(
|
2018-03-17 15:46:50 -04:00
|
|
|
|
xspf_url, playlist_id, 'Downloading xpsf playlist',
|
2015-08-09 09:41:55 -04:00
|
|
|
|
'Unable to download xspf manifest', fatal=fatal)
|
2022-04-27 18:50:01 -04:00
|
|
|
|
if res is False:
|
2015-08-09 09:41:55 -04:00
|
|
|
|
return []
|
2022-04-27 18:50:01 -04:00
|
|
|
|
|
|
|
|
|
xspf, urlh = res
|
|
|
|
|
xspf_url = urlh.geturl()
|
|
|
|
|
|
2018-03-17 15:46:50 -04:00
|
|
|
|
return self._parse_xspf(
|
|
|
|
|
xspf, playlist_id, xspf_url=xspf_url,
|
|
|
|
|
xspf_base_url=base_url(xspf_url))
|
2015-08-09 09:07:18 -04:00
|
|
|
|
|
2018-03-17 15:46:50 -04:00
|
|
|
|
def _parse_xspf(self, xspf_doc, playlist_id, xspf_url=None, xspf_base_url=None):
|
2015-08-09 09:07:18 -04:00
|
|
|
|
NS_MAP = {
|
|
|
|
|
'xspf': 'http://xspf.org/ns/0/',
|
|
|
|
|
's1': 'http://static.streamone.nl/player/ns/0',
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
entries = []
|
2018-03-17 15:46:50 -04:00
|
|
|
|
for track in xspf_doc.findall(xpath_with_ns('./xspf:trackList/xspf:track', NS_MAP)):
|
2015-08-09 09:07:18 -04:00
|
|
|
|
title = xpath_text(
|
2015-08-09 09:18:50 -04:00
|
|
|
|
track, xpath_with_ns('./xspf:title', NS_MAP), 'title', default=playlist_id)
|
2015-08-09 09:07:18 -04:00
|
|
|
|
description = xpath_text(
|
|
|
|
|
track, xpath_with_ns('./xspf:annotation', NS_MAP), 'description')
|
|
|
|
|
thumbnail = xpath_text(
|
|
|
|
|
track, xpath_with_ns('./xspf:image', NS_MAP), 'thumbnail')
|
|
|
|
|
duration = float_or_none(
|
|
|
|
|
xpath_text(track, xpath_with_ns('./xspf:duration', NS_MAP), 'duration'), 1000)
|
|
|
|
|
|
2018-03-17 15:46:50 -04:00
|
|
|
|
formats = []
|
|
|
|
|
for location in track.findall(xpath_with_ns('./xspf:location', NS_MAP)):
|
|
|
|
|
format_url = urljoin(xspf_base_url, location.text)
|
|
|
|
|
if not format_url:
|
|
|
|
|
continue
|
|
|
|
|
formats.append({
|
|
|
|
|
'url': format_url,
|
|
|
|
|
'manifest_url': xspf_url,
|
|
|
|
|
'format_id': location.get(xpath_with_ns('s1:label', NS_MAP)),
|
|
|
|
|
'width': int_or_none(location.get(xpath_with_ns('s1:width', NS_MAP))),
|
|
|
|
|
'height': int_or_none(location.get(xpath_with_ns('s1:height', NS_MAP))),
|
|
|
|
|
})
|
2015-08-09 09:07:18 -04:00
|
|
|
|
self._sort_formats(formats)
|
|
|
|
|
|
|
|
|
|
entries.append({
|
|
|
|
|
'id': playlist_id,
|
|
|
|
|
'title': title,
|
|
|
|
|
'description': description,
|
|
|
|
|
'thumbnail': thumbnail,
|
|
|
|
|
'duration': duration,
|
|
|
|
|
'formats': formats,
|
|
|
|
|
})
|
|
|
|
|
return entries
|
|
|
|
|
|
2021-04-17 19:49:22 -04:00
|
|
|
|
def _extract_mpd_formats(self, *args, **kwargs):
|
|
|
|
|
fmts, subs = self._extract_mpd_formats_and_subtitles(*args, **kwargs)
|
|
|
|
|
if subs:
|
2021-10-08 20:23:15 -04:00
|
|
|
|
self._report_ignoring_subs('DASH')
|
2021-04-17 19:49:22 -04:00
|
|
|
|
return fmts
|
|
|
|
|
|
|
|
|
|
def _extract_mpd_formats_and_subtitles(
|
|
|
|
|
self, mpd_url, video_id, mpd_id=None, note=None, errnote=None,
|
|
|
|
|
fatal=True, data=None, headers={}, query={}):
|
2018-03-17 15:46:50 -04:00
|
|
|
|
res = self._download_xml_handle(
|
2016-02-02 12:07:07 -05:00
|
|
|
|
mpd_url, video_id,
|
2021-05-29 04:52:44 -04:00
|
|
|
|
note='Downloading MPD manifest' if note is None else note,
|
|
|
|
|
errnote='Failed to download MPD manifest' if errnote is None else errnote,
|
2019-11-15 17:44:14 -05:00
|
|
|
|
fatal=fatal, data=data, headers=headers, query=query)
|
2016-02-02 12:07:07 -05:00
|
|
|
|
if res is False:
|
2021-04-17 19:49:22 -04:00
|
|
|
|
return [], {}
|
2018-03-17 15:46:50 -04:00
|
|
|
|
mpd_doc, urlh = res
|
2019-04-21 12:20:28 -04:00
|
|
|
|
if mpd_doc is None:
|
2021-04-17 19:49:22 -04:00
|
|
|
|
return [], {}
|
2022-04-27 14:01:35 -04:00
|
|
|
|
|
|
|
|
|
# We could have been redirected to a new url when we retrieved our mpd file.
|
|
|
|
|
mpd_url = urlh.geturl()
|
|
|
|
|
mpd_base_url = base_url(mpd_url)
|
2016-02-02 12:07:07 -05:00
|
|
|
|
|
2021-04-17 19:49:22 -04:00
|
|
|
|
return self._parse_mpd_formats_and_subtitles(
|
2021-02-04 09:37:17 -05:00
|
|
|
|
mpd_doc, mpd_id, mpd_base_url, mpd_url)
|
2016-01-30 09:52:23 -05:00
|
|
|
|
|
2021-04-17 19:49:22 -04:00
|
|
|
|
def _parse_mpd_formats(self, *args, **kwargs):
|
|
|
|
|
fmts, subs = self._parse_mpd_formats_and_subtitles(*args, **kwargs)
|
|
|
|
|
if subs:
|
2021-10-08 20:23:15 -04:00
|
|
|
|
self._report_ignoring_subs('DASH')
|
2021-04-17 19:49:22 -04:00
|
|
|
|
return fmts
|
|
|
|
|
|
|
|
|
|
def _parse_mpd_formats_and_subtitles(
|
|
|
|
|
self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
|
2016-07-23 23:27:16 -04:00
|
|
|
|
"""
|
|
|
|
|
Parse formats from MPD manifest.
|
|
|
|
|
References:
|
|
|
|
|
1. MPEG-DASH Standard, ISO/IEC 23009-1:2014(E),
|
|
|
|
|
http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
|
|
|
|
|
2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
|
|
|
|
|
"""
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if not self.get_param('dynamic_mpd', True):
|
2020-09-16 07:00:41 -04:00
|
|
|
|
if mpd_doc.get('type') == 'dynamic':
|
2021-04-17 19:49:22 -04:00
|
|
|
|
return [], {}
|
2016-01-30 09:52:23 -05:00
|
|
|
|
|
2016-02-06 08:03:48 -05:00
|
|
|
|
namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
|
2016-02-02 16:02:08 -05:00
|
|
|
|
|
|
|
|
|
def _add_ns(path):
|
|
|
|
|
return self._xpath_ns(path, namespace)
|
|
|
|
|
|
2016-02-03 12:44:43 -05:00
|
|
|
|
def is_drm_protected(element):
|
|
|
|
|
return element.find(_add_ns('ContentProtection')) is not None
|
|
|
|
|
|
2016-02-02 12:07:07 -05:00
|
|
|
|
def extract_multisegment_info(element, ms_parent_info):
|
|
|
|
|
ms_info = ms_parent_info.copy()
|
2016-09-05 14:21:57 -04:00
|
|
|
|
|
|
|
|
|
# As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
|
|
|
|
|
# common attributes and elements. We will only extract relevant
|
|
|
|
|
# for us.
|
|
|
|
|
def extract_common(source):
|
|
|
|
|
segment_timeline = source.find(_add_ns('SegmentTimeline'))
|
|
|
|
|
if segment_timeline is not None:
|
|
|
|
|
s_e = segment_timeline.findall(_add_ns('S'))
|
|
|
|
|
if s_e:
|
|
|
|
|
ms_info['total_number'] = 0
|
|
|
|
|
ms_info['s'] = []
|
|
|
|
|
for s in s_e:
|
|
|
|
|
r = int(s.get('r', 0))
|
|
|
|
|
ms_info['total_number'] += 1 + r
|
|
|
|
|
ms_info['s'].append({
|
|
|
|
|
't': int(s.get('t', 0)),
|
|
|
|
|
# @d is mandatory (see [1, 5.3.9.6.2, Table 17, page 60])
|
|
|
|
|
'd': int(s.attrib['d']),
|
|
|
|
|
'r': r,
|
|
|
|
|
})
|
|
|
|
|
start_number = source.get('startNumber')
|
|
|
|
|
if start_number:
|
|
|
|
|
ms_info['start_number'] = int(start_number)
|
|
|
|
|
timescale = source.get('timescale')
|
|
|
|
|
if timescale:
|
|
|
|
|
ms_info['timescale'] = int(timescale)
|
|
|
|
|
segment_duration = source.get('duration')
|
|
|
|
|
if segment_duration:
|
2017-08-15 12:58:00 -04:00
|
|
|
|
ms_info['segment_duration'] = float(segment_duration)
|
2016-09-05 14:21:57 -04:00
|
|
|
|
|
|
|
|
|
def extract_Initialization(source):
|
|
|
|
|
initialization = source.find(_add_ns('Initialization'))
|
|
|
|
|
if initialization is not None:
|
|
|
|
|
ms_info['initialization_url'] = initialization.attrib['sourceURL']
|
|
|
|
|
|
2016-02-02 16:02:08 -05:00
|
|
|
|
segment_list = element.find(_add_ns('SegmentList'))
|
2016-02-02 12:07:07 -05:00
|
|
|
|
if segment_list is not None:
|
2016-09-05 14:21:57 -04:00
|
|
|
|
extract_common(segment_list)
|
|
|
|
|
extract_Initialization(segment_list)
|
2016-02-02 16:02:08 -05:00
|
|
|
|
segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
|
2016-02-02 12:07:07 -05:00
|
|
|
|
if segment_urls_e:
|
|
|
|
|
ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
|
|
|
|
|
else:
|
2016-02-02 16:02:08 -05:00
|
|
|
|
segment_template = element.find(_add_ns('SegmentTemplate'))
|
2016-02-02 12:07:07 -05:00
|
|
|
|
if segment_template is not None:
|
2016-09-05 14:21:57 -04:00
|
|
|
|
extract_common(segment_template)
|
2017-01-28 18:57:39 -05:00
|
|
|
|
media = segment_template.get('media')
|
|
|
|
|
if media:
|
|
|
|
|
ms_info['media'] = media
|
2016-02-02 12:07:07 -05:00
|
|
|
|
initialization = segment_template.get('initialization')
|
|
|
|
|
if initialization:
|
2017-01-28 18:57:39 -05:00
|
|
|
|
ms_info['initialization'] = initialization
|
2016-02-02 12:07:07 -05:00
|
|
|
|
else:
|
2016-09-05 14:21:57 -04:00
|
|
|
|
extract_Initialization(segment_template)
|
2016-02-02 12:07:07 -05:00
|
|
|
|
return ms_info
|
2016-01-30 08:27:43 -05:00
|
|
|
|
|
2016-02-02 12:07:07 -05:00
|
|
|
|
mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
|
2021-08-23 20:12:45 -04:00
|
|
|
|
formats, subtitles = [], {}
|
2021-11-22 13:02:14 -05:00
|
|
|
|
stream_numbers = collections.defaultdict(int)
|
2016-02-02 16:02:08 -05:00
|
|
|
|
for period in mpd_doc.findall(_add_ns('Period')):
|
2016-02-02 12:07:07 -05:00
|
|
|
|
period_duration = parse_duration(period.get('duration')) or mpd_duration
|
|
|
|
|
period_ms_info = extract_multisegment_info(period, {
|
|
|
|
|
'start_number': 1,
|
|
|
|
|
'timescale': 1,
|
|
|
|
|
})
|
2016-02-02 16:02:08 -05:00
|
|
|
|
for adaptation_set in period.findall(_add_ns('AdaptationSet')):
|
2016-02-02 12:07:07 -05:00
|
|
|
|
adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
|
2016-02-02 16:02:08 -05:00
|
|
|
|
for representation in adaptation_set.findall(_add_ns('Representation')):
|
2016-02-02 12:07:07 -05:00
|
|
|
|
representation_attrib = adaptation_set.attrib.copy()
|
|
|
|
|
representation_attrib.update(representation.attrib)
|
2016-07-23 23:27:16 -04:00
|
|
|
|
# According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
|
2016-03-11 10:49:55 -05:00
|
|
|
|
mime_type = representation_attrib['mimeType']
|
2021-04-17 19:49:22 -04:00
|
|
|
|
content_type = representation_attrib.get('contentType', mime_type.split('/')[0])
|
|
|
|
|
|
2022-05-17 23:34:30 -04:00
|
|
|
|
codec_str = representation_attrib.get('codecs', '')
|
|
|
|
|
# Some kind of binary subtitle found in some youtube livestreams
|
|
|
|
|
if mime_type == 'application/x-rawcc':
|
|
|
|
|
codecs = {'scodec': codec_str}
|
|
|
|
|
else:
|
|
|
|
|
codecs = parse_codecs(codec_str)
|
2021-08-09 19:12:03 -04:00
|
|
|
|
if content_type not in ('video', 'audio', 'text'):
|
|
|
|
|
if mime_type == 'image/jpeg':
|
2021-08-11 04:12:23 -04:00
|
|
|
|
content_type = mime_type
|
2022-05-17 23:34:30 -04:00
|
|
|
|
elif codecs.get('vcodec', 'none') != 'none':
|
2021-12-31 15:06:45 -05:00
|
|
|
|
content_type = 'video'
|
2022-05-17 23:34:30 -04:00
|
|
|
|
elif codecs.get('acodec', 'none') != 'none':
|
2021-12-31 15:06:45 -05:00
|
|
|
|
content_type = 'audio'
|
2022-04-29 12:02:31 -04:00
|
|
|
|
elif codecs.get('scodec', 'none') != 'none':
|
2021-08-09 19:12:03 -04:00
|
|
|
|
content_type = 'text'
|
2021-10-12 19:33:40 -04:00
|
|
|
|
elif mimetype2ext(mime_type) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
|
|
|
|
|
content_type = 'text'
|
2021-05-23 12:34:49 -04:00
|
|
|
|
else:
|
2021-08-09 19:12:03 -04:00
|
|
|
|
self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
|
|
|
|
|
continue
|
|
|
|
|
|
|
|
|
|
base_url = ''
|
|
|
|
|
for element in (representation, adaptation_set, period, mpd_doc):
|
|
|
|
|
base_url_e = element.find(_add_ns('BaseURL'))
|
2022-06-28 20:37:21 -04:00
|
|
|
|
if try_call(lambda: base_url_e.text) is not None:
|
2021-08-09 19:12:03 -04:00
|
|
|
|
base_url = base_url_e.text + base_url
|
|
|
|
|
if re.match(r'^https?://', base_url):
|
|
|
|
|
break
|
2021-09-19 04:37:57 -04:00
|
|
|
|
if mpd_base_url and base_url.startswith('/'):
|
2022-06-24 06:54:43 -04:00
|
|
|
|
base_url = urllib.parse.urljoin(mpd_base_url, base_url)
|
2021-09-19 04:37:57 -04:00
|
|
|
|
elif mpd_base_url and not re.match(r'^https?://', base_url):
|
|
|
|
|
if not mpd_base_url.endswith('/'):
|
2021-08-09 19:12:03 -04:00
|
|
|
|
mpd_base_url += '/'
|
|
|
|
|
base_url = mpd_base_url + base_url
|
|
|
|
|
representation_id = representation_attrib.get('id')
|
|
|
|
|
lang = representation_attrib.get('lang')
|
|
|
|
|
url_el = representation.find(_add_ns('BaseURL'))
|
|
|
|
|
filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
|
|
|
|
|
bandwidth = int_or_none(representation_attrib.get('bandwidth'))
|
|
|
|
|
if representation_id is not None:
|
|
|
|
|
format_id = representation_id
|
|
|
|
|
else:
|
|
|
|
|
format_id = content_type
|
|
|
|
|
if mpd_id:
|
|
|
|
|
format_id = mpd_id + '-' + format_id
|
|
|
|
|
if content_type in ('video', 'audio'):
|
|
|
|
|
f = {
|
|
|
|
|
'format_id': format_id,
|
|
|
|
|
'manifest_url': mpd_url,
|
|
|
|
|
'ext': mimetype2ext(mime_type),
|
|
|
|
|
'width': int_or_none(representation_attrib.get('width')),
|
|
|
|
|
'height': int_or_none(representation_attrib.get('height')),
|
|
|
|
|
'tbr': float_or_none(bandwidth, 1000),
|
|
|
|
|
'asr': int_or_none(representation_attrib.get('audioSamplingRate')),
|
|
|
|
|
'fps': int_or_none(representation_attrib.get('frameRate')),
|
|
|
|
|
'language': lang if lang not in ('mul', 'und', 'zxx', 'mis') else None,
|
|
|
|
|
'format_note': 'DASH %s' % content_type,
|
|
|
|
|
'filesize': filesize,
|
|
|
|
|
'container': mimetype2ext(mime_type) + '_dash',
|
2021-12-31 15:06:45 -05:00
|
|
|
|
**codecs
|
2021-08-09 19:12:03 -04:00
|
|
|
|
}
|
|
|
|
|
elif content_type == 'text':
|
|
|
|
|
f = {
|
|
|
|
|
'ext': mimetype2ext(mime_type),
|
|
|
|
|
'manifest_url': mpd_url,
|
|
|
|
|
'filesize': filesize,
|
|
|
|
|
}
|
|
|
|
|
elif content_type == 'image/jpeg':
|
|
|
|
|
# See test case in VikiIE
|
|
|
|
|
# https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
|
|
|
|
|
f = {
|
|
|
|
|
'format_id': format_id,
|
|
|
|
|
'ext': 'mhtml',
|
|
|
|
|
'manifest_url': mpd_url,
|
|
|
|
|
'format_note': 'DASH storyboards (jpeg)',
|
|
|
|
|
'acodec': 'none',
|
|
|
|
|
'vcodec': 'none',
|
|
|
|
|
}
|
2021-08-22 16:08:38 -04:00
|
|
|
|
if is_drm_protected(adaptation_set) or is_drm_protected(representation):
|
|
|
|
|
f['has_drm'] = True
|
2021-08-09 19:12:03 -04:00
|
|
|
|
representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
|
|
|
|
|
|
|
|
|
|
def prepare_template(template_name, identifiers):
|
|
|
|
|
tmpl = representation_ms_info[template_name]
|
2022-09-09 00:28:41 -04:00
|
|
|
|
if representation_id is not None:
|
|
|
|
|
tmpl = tmpl.replace('$RepresentationID$', representation_id)
|
2021-08-09 19:12:03 -04:00
|
|
|
|
# First of, % characters outside $...$ templates
|
|
|
|
|
# must be escaped by doubling for proper processing
|
|
|
|
|
# by % operator string formatting used further (see
|
|
|
|
|
# https://github.com/ytdl-org/youtube-dl/issues/16867).
|
|
|
|
|
t = ''
|
|
|
|
|
in_template = False
|
|
|
|
|
for c in tmpl:
|
|
|
|
|
t += c
|
|
|
|
|
if c == '$':
|
|
|
|
|
in_template = not in_template
|
|
|
|
|
elif c == '%' and not in_template:
|
2018-06-30 15:00:16 -04:00
|
|
|
|
t += c
|
2021-08-09 19:12:03 -04:00
|
|
|
|
# Next, $...$ templates are translated to their
|
|
|
|
|
# %(...) counterparts to be used with % operator
|
|
|
|
|
t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
|
|
|
|
|
t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
|
|
|
|
|
t.replace('$$', '$')
|
|
|
|
|
return t
|
|
|
|
|
|
|
|
|
|
# @initialization is a regular template like @media one
|
|
|
|
|
# so it should be handled just the same way (see
|
|
|
|
|
# https://github.com/ytdl-org/youtube-dl/issues/11605)
|
|
|
|
|
if 'initialization' in representation_ms_info:
|
|
|
|
|
initialization_template = prepare_template(
|
|
|
|
|
'initialization',
|
|
|
|
|
# As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
|
|
|
|
|
# $Time$ shall not be included for @initialization thus
|
|
|
|
|
# only $Bandwidth$ remains
|
|
|
|
|
('Bandwidth', ))
|
|
|
|
|
representation_ms_info['initialization_url'] = initialization_template % {
|
|
|
|
|
'Bandwidth': bandwidth,
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
def location_key(location):
|
|
|
|
|
return 'url' if re.match(r'^https?://', location) else 'path'
|
|
|
|
|
|
|
|
|
|
if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info:
|
|
|
|
|
|
|
|
|
|
media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time'))
|
|
|
|
|
media_location_key = location_key(media_template)
|
|
|
|
|
|
|
|
|
|
# As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
|
|
|
|
|
# can't be used at the same time
|
|
|
|
|
if '%(Number' in media_template and 's' not in representation_ms_info:
|
|
|
|
|
segment_duration = None
|
|
|
|
|
if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info:
|
|
|
|
|
segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
|
2022-02-17 08:36:22 -05:00
|
|
|
|
representation_ms_info['total_number'] = int(math.ceil(
|
|
|
|
|
float_or_none(period_duration, segment_duration, default=0)))
|
2021-08-09 19:12:03 -04:00
|
|
|
|
representation_ms_info['fragments'] = [{
|
|
|
|
|
media_location_key: media_template % {
|
|
|
|
|
'Number': segment_number,
|
|
|
|
|
'Bandwidth': bandwidth,
|
|
|
|
|
},
|
|
|
|
|
'duration': segment_duration,
|
|
|
|
|
} for segment_number in range(
|
|
|
|
|
representation_ms_info['start_number'],
|
|
|
|
|
representation_ms_info['total_number'] + representation_ms_info['start_number'])]
|
|
|
|
|
else:
|
|
|
|
|
# $Number*$ or $Time$ in media template with S list available
|
|
|
|
|
# Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
|
|
|
|
|
# Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
|
|
|
|
|
representation_ms_info['fragments'] = []
|
|
|
|
|
segment_time = 0
|
|
|
|
|
segment_d = None
|
|
|
|
|
segment_number = representation_ms_info['start_number']
|
|
|
|
|
|
|
|
|
|
def add_segment_url():
|
|
|
|
|
segment_url = media_template % {
|
|
|
|
|
'Time': segment_time,
|
|
|
|
|
'Bandwidth': bandwidth,
|
|
|
|
|
'Number': segment_number,
|
|
|
|
|
}
|
|
|
|
|
representation_ms_info['fragments'].append({
|
|
|
|
|
media_location_key: segment_url,
|
|
|
|
|
'duration': float_or_none(segment_d, representation_ms_info['timescale']),
|
|
|
|
|
})
|
|
|
|
|
|
|
|
|
|
for num, s in enumerate(representation_ms_info['s']):
|
|
|
|
|
segment_time = s.get('t') or segment_time
|
|
|
|
|
segment_d = s['d']
|
|
|
|
|
add_segment_url()
|
|
|
|
|
segment_number += 1
|
|
|
|
|
for r in range(s.get('r', 0)):
|
|
|
|
|
segment_time += segment_d
|
2016-07-23 23:27:16 -04:00
|
|
|
|
add_segment_url()
|
2016-09-05 14:21:57 -04:00
|
|
|
|
segment_number += 1
|
2021-08-09 19:12:03 -04:00
|
|
|
|
segment_time += segment_d
|
|
|
|
|
elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
|
2022-08-14 08:04:13 -04:00
|
|
|
|
# No media template,
|
|
|
|
|
# e.g. https://www.youtube.com/watch?v=iXZV5uAYMJI
|
2021-08-09 19:12:03 -04:00
|
|
|
|
# or any YouTube dashsegments video
|
|
|
|
|
fragments = []
|
|
|
|
|
segment_index = 0
|
|
|
|
|
timescale = representation_ms_info['timescale']
|
|
|
|
|
for s in representation_ms_info['s']:
|
|
|
|
|
duration = float_or_none(s['d'], timescale)
|
|
|
|
|
for r in range(s.get('r', 0) + 1):
|
|
|
|
|
segment_uri = representation_ms_info['segment_urls'][segment_index]
|
|
|
|
|
fragments.append({
|
|
|
|
|
location_key(segment_uri): segment_uri,
|
|
|
|
|
'duration': duration,
|
|
|
|
|
})
|
|
|
|
|
segment_index += 1
|
|
|
|
|
representation_ms_info['fragments'] = fragments
|
|
|
|
|
elif 'segment_urls' in representation_ms_info:
|
|
|
|
|
# Segment URLs with no SegmentTimeline
|
2022-08-14 08:04:13 -04:00
|
|
|
|
# E.g. https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
|
2021-08-09 19:12:03 -04:00
|
|
|
|
# https://github.com/ytdl-org/youtube-dl/pull/14844
|
|
|
|
|
fragments = []
|
|
|
|
|
segment_duration = float_or_none(
|
|
|
|
|
representation_ms_info['segment_duration'],
|
|
|
|
|
representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None
|
|
|
|
|
for segment_url in representation_ms_info['segment_urls']:
|
|
|
|
|
fragment = {
|
|
|
|
|
location_key(segment_url): segment_url,
|
|
|
|
|
}
|
|
|
|
|
if segment_duration:
|
|
|
|
|
fragment['duration'] = segment_duration
|
|
|
|
|
fragments.append(fragment)
|
|
|
|
|
representation_ms_info['fragments'] = fragments
|
|
|
|
|
# If there is a fragments key available then we correctly recognized fragmented media.
|
|
|
|
|
# Otherwise we will assume unfragmented media with direct access. Technically, such
|
|
|
|
|
# assumption is not necessarily correct since we may simply have no support for
|
|
|
|
|
# some forms of fragmented media renditions yet, but for now we'll use this fallback.
|
|
|
|
|
if 'fragments' in representation_ms_info:
|
|
|
|
|
f.update({
|
|
|
|
|
# NB: mpd_url may be empty when MPD manifest is parsed from a string
|
|
|
|
|
'url': mpd_url or base_url,
|
|
|
|
|
'fragment_base_url': base_url,
|
|
|
|
|
'fragments': [],
|
|
|
|
|
'protocol': 'http_dash_segments' if mime_type != 'image/jpeg' else 'mhtml',
|
|
|
|
|
})
|
|
|
|
|
if 'initialization_url' in representation_ms_info:
|
|
|
|
|
initialization_url = representation_ms_info['initialization_url']
|
|
|
|
|
if not f.get('url'):
|
|
|
|
|
f['url'] = initialization_url
|
|
|
|
|
f['fragments'].append({location_key(initialization_url): initialization_url})
|
|
|
|
|
f['fragments'].extend(representation_ms_info['fragments'])
|
2022-02-17 08:36:22 -05:00
|
|
|
|
if not period_duration:
|
|
|
|
|
period_duration = try_get(
|
|
|
|
|
representation_ms_info,
|
|
|
|
|
lambda r: sum(frag['duration'] for frag in r['fragments']), float)
|
2016-01-30 08:05:55 -05:00
|
|
|
|
else:
|
2021-08-09 19:12:03 -04:00
|
|
|
|
# Assuming direct URL to unfragmented media.
|
|
|
|
|
f['url'] = base_url
|
2021-11-22 13:02:14 -05:00
|
|
|
|
if content_type in ('video', 'audio', 'image/jpeg'):
|
|
|
|
|
f['manifest_stream_number'] = stream_numbers[f['url']]
|
|
|
|
|
stream_numbers[f['url']] += 1
|
2021-08-09 19:12:03 -04:00
|
|
|
|
formats.append(f)
|
|
|
|
|
elif content_type == 'text':
|
|
|
|
|
subtitles.setdefault(lang or 'und', []).append(f)
|
|
|
|
|
|
2021-04-17 19:49:22 -04:00
|
|
|
|
return formats, subtitles
|
2016-01-30 08:05:55 -05:00
|
|
|
|
|
2021-04-18 06:47:42 -04:00
|
|
|
|
def _extract_ism_formats(self, *args, **kwargs):
|
|
|
|
|
fmts, subs = self._extract_ism_formats_and_subtitles(*args, **kwargs)
|
|
|
|
|
if subs:
|
2021-10-08 20:23:15 -04:00
|
|
|
|
self._report_ignoring_subs('ISM')
|
2021-04-18 06:47:42 -04:00
|
|
|
|
return fmts
|
|
|
|
|
|
|
|
|
|
def _extract_ism_formats_and_subtitles(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
|
2018-03-17 15:46:50 -04:00
|
|
|
|
res = self._download_xml_handle(
|
2016-10-19 11:22:40 -04:00
|
|
|
|
ism_url, video_id,
|
2021-05-29 04:52:44 -04:00
|
|
|
|
note='Downloading ISM manifest' if note is None else note,
|
|
|
|
|
errnote='Failed to download ISM manifest' if errnote is None else errnote,
|
2019-11-15 17:44:14 -05:00
|
|
|
|
fatal=fatal, data=data, headers=headers, query=query)
|
2016-10-19 11:22:40 -04:00
|
|
|
|
if res is False:
|
2021-04-18 06:47:42 -04:00
|
|
|
|
return [], {}
|
2018-03-17 15:46:50 -04:00
|
|
|
|
ism_doc, urlh = res
|
2020-04-07 11:54:34 -04:00
|
|
|
|
if ism_doc is None:
|
2021-04-18 06:47:42 -04:00
|
|
|
|
return [], {}
|
2016-10-19 11:22:40 -04:00
|
|
|
|
|
2021-04-18 06:47:42 -04:00
|
|
|
|
return self._parse_ism_formats_and_subtitles(ism_doc, urlh.geturl(), ism_id)
|
2016-10-19 11:22:40 -04:00
|
|
|
|
|
2021-04-18 06:47:42 -04:00
|
|
|
|
def _parse_ism_formats_and_subtitles(self, ism_doc, ism_url, ism_id=None):
|
2017-05-13 19:11:45 -04:00
|
|
|
|
"""
|
|
|
|
|
Parse formats from ISM manifest.
|
|
|
|
|
References:
|
|
|
|
|
1. [MS-SSTR]: Smooth Streaming Protocol,
|
|
|
|
|
https://msdn.microsoft.com/en-us/library/ff469518.aspx
|
|
|
|
|
"""
|
2021-02-11 23:34:04 -05:00
|
|
|
|
if ism_doc.get('IsLive') == 'TRUE':
|
2021-04-18 06:47:42 -04:00
|
|
|
|
return [], {}
|
2016-10-19 11:22:40 -04:00
|
|
|
|
|
|
|
|
|
duration = int(ism_doc.attrib['Duration'])
|
|
|
|
|
timescale = int_or_none(ism_doc.get('TimeScale')) or 10000000
|
|
|
|
|
|
|
|
|
|
formats = []
|
2021-04-18 06:47:42 -04:00
|
|
|
|
subtitles = {}
|
2016-10-19 11:22:40 -04:00
|
|
|
|
for stream in ism_doc.findall('StreamIndex'):
|
|
|
|
|
stream_type = stream.get('Type')
|
2021-04-18 06:47:42 -04:00
|
|
|
|
if stream_type not in ('video', 'audio', 'text'):
|
2016-10-19 11:22:40 -04:00
|
|
|
|
continue
|
|
|
|
|
url_pattern = stream.attrib['Url']
|
|
|
|
|
stream_timescale = int_or_none(stream.get('TimeScale')) or timescale
|
|
|
|
|
stream_name = stream.get('Name')
|
2021-04-18 06:47:42 -04:00
|
|
|
|
stream_language = stream.get('Language', 'und')
|
2016-10-19 11:22:40 -04:00
|
|
|
|
for track in stream.findall('QualityLevel'):
|
2022-09-30 13:33:29 -04:00
|
|
|
|
KNOWN_TAGS = {'255': 'AACL', '65534': 'EC-3'}
|
|
|
|
|
fourcc = track.get('FourCC') or KNOWN_TAGS.get(track.get('AudioTag'))
|
2016-10-19 11:22:40 -04:00
|
|
|
|
# TODO: add support for WVC1 and WMAP
|
2022-09-30 13:33:29 -04:00
|
|
|
|
if fourcc not in ('H264', 'AVC1', 'AACL', 'TTML', 'EC-3'):
|
2016-10-19 11:22:40 -04:00
|
|
|
|
self.report_warning('%s is not a supported codec' % fourcc)
|
|
|
|
|
continue
|
|
|
|
|
tbr = int(track.attrib['Bitrate']) // 1000
|
2017-05-13 19:11:45 -04:00
|
|
|
|
# [1] does not mention Width and Height attributes. However,
|
|
|
|
|
# they're often present while MaxWidth and MaxHeight are
|
|
|
|
|
# missing, so should be used as fallbacks
|
|
|
|
|
width = int_or_none(track.get('MaxWidth') or track.get('Width'))
|
|
|
|
|
height = int_or_none(track.get('MaxHeight') or track.get('Height'))
|
2016-10-19 11:22:40 -04:00
|
|
|
|
sampling_rate = int_or_none(track.get('SamplingRate'))
|
|
|
|
|
|
|
|
|
|
track_url_pattern = re.sub(r'{[Bb]itrate}', track.attrib['Bitrate'], url_pattern)
|
2022-06-24 06:54:43 -04:00
|
|
|
|
track_url_pattern = urllib.parse.urljoin(ism_url, track_url_pattern)
|
2016-10-19 11:22:40 -04:00
|
|
|
|
|
|
|
|
|
fragments = []
|
|
|
|
|
fragment_ctx = {
|
|
|
|
|
'time': 0,
|
|
|
|
|
}
|
|
|
|
|
stream_fragments = stream.findall('c')
|
|
|
|
|
for stream_fragment_index, stream_fragment in enumerate(stream_fragments):
|
|
|
|
|
fragment_ctx['time'] = int_or_none(stream_fragment.get('t')) or fragment_ctx['time']
|
|
|
|
|
fragment_repeat = int_or_none(stream_fragment.get('r')) or 1
|
|
|
|
|
fragment_ctx['duration'] = int_or_none(stream_fragment.get('d'))
|
|
|
|
|
if not fragment_ctx['duration']:
|
|
|
|
|
try:
|
|
|
|
|
next_fragment_time = int(stream_fragment[stream_fragment_index + 1].attrib['t'])
|
|
|
|
|
except IndexError:
|
|
|
|
|
next_fragment_time = duration
|
2016-11-01 15:21:43 -04:00
|
|
|
|
fragment_ctx['duration'] = (next_fragment_time - fragment_ctx['time']) / fragment_repeat
|
2016-10-19 11:22:40 -04:00
|
|
|
|
for _ in range(fragment_repeat):
|
|
|
|
|
fragments.append({
|
2022-06-24 06:54:43 -04:00
|
|
|
|
'url': re.sub(r'{start[ _]time}', str(fragment_ctx['time']), track_url_pattern),
|
2016-10-19 11:22:40 -04:00
|
|
|
|
'duration': fragment_ctx['duration'] / stream_timescale,
|
|
|
|
|
})
|
|
|
|
|
fragment_ctx['time'] += fragment_ctx['duration']
|
|
|
|
|
|
2021-04-18 06:47:42 -04:00
|
|
|
|
if stream_type == 'text':
|
|
|
|
|
subtitles.setdefault(stream_language, []).append({
|
|
|
|
|
'ext': 'ismt',
|
|
|
|
|
'protocol': 'ism',
|
|
|
|
|
'url': ism_url,
|
|
|
|
|
'manifest_url': ism_url,
|
|
|
|
|
'fragments': fragments,
|
|
|
|
|
'_download_params': {
|
|
|
|
|
'stream_type': stream_type,
|
|
|
|
|
'duration': duration,
|
|
|
|
|
'timescale': stream_timescale,
|
|
|
|
|
'fourcc': fourcc,
|
|
|
|
|
'language': stream_language,
|
|
|
|
|
'codec_private_data': track.get('CodecPrivateData'),
|
|
|
|
|
}
|
|
|
|
|
})
|
|
|
|
|
elif stream_type in ('video', 'audio'):
|
|
|
|
|
formats.append({
|
2021-11-05 21:05:24 -04:00
|
|
|
|
'format_id': join_nonempty(ism_id, stream_name, tbr),
|
2021-04-18 06:47:42 -04:00
|
|
|
|
'url': ism_url,
|
|
|
|
|
'manifest_url': ism_url,
|
|
|
|
|
'ext': 'ismv' if stream_type == 'video' else 'isma',
|
|
|
|
|
'width': width,
|
|
|
|
|
'height': height,
|
|
|
|
|
'tbr': tbr,
|
|
|
|
|
'asr': sampling_rate,
|
|
|
|
|
'vcodec': 'none' if stream_type == 'audio' else fourcc,
|
|
|
|
|
'acodec': 'none' if stream_type == 'video' else fourcc,
|
|
|
|
|
'protocol': 'ism',
|
|
|
|
|
'fragments': fragments,
|
2021-08-22 16:08:38 -04:00
|
|
|
|
'has_drm': ism_doc.find('Protection') is not None,
|
2021-04-18 06:47:42 -04:00
|
|
|
|
'_download_params': {
|
|
|
|
|
'stream_type': stream_type,
|
|
|
|
|
'duration': duration,
|
|
|
|
|
'timescale': stream_timescale,
|
|
|
|
|
'width': width or 0,
|
|
|
|
|
'height': height or 0,
|
|
|
|
|
'fourcc': fourcc,
|
|
|
|
|
'language': stream_language,
|
|
|
|
|
'codec_private_data': track.get('CodecPrivateData'),
|
|
|
|
|
'sampling_rate': sampling_rate,
|
|
|
|
|
'channels': int_or_none(track.get('Channels', 2)),
|
|
|
|
|
'bits_per_sample': int_or_none(track.get('BitsPerSample', 16)),
|
|
|
|
|
'nal_unit_length_field': int_or_none(track.get('NALUnitLengthField', 4)),
|
|
|
|
|
},
|
|
|
|
|
})
|
|
|
|
|
return formats, subtitles
|
2016-10-19 11:22:40 -04:00
|
|
|
|
|
2022-02-11 01:39:03 -05:00
|
|
|
|
def _parse_html5_media_entries(self, base_url, webpage, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8_native', mpd_id=None, preference=None, quality=None):
|
2018-03-19 12:43:53 -04:00
|
|
|
|
def absolute_url(item_url):
|
|
|
|
|
return urljoin(base_url, item_url)
|
2016-03-16 13:50:45 -04:00
|
|
|
|
|
|
|
|
|
def parse_content_type(content_type):
|
|
|
|
|
if not content_type:
|
|
|
|
|
return {}
|
|
|
|
|
ctr = re.search(r'(?P<mimetype>[^/]+/[^;]+)(?:;\s*codecs="?(?P<codecs>[^"]+))?', content_type)
|
|
|
|
|
if ctr:
|
|
|
|
|
mimetype, codecs = ctr.groups()
|
|
|
|
|
f = parse_codecs(codecs)
|
|
|
|
|
f['ext'] = mimetype2ext(mimetype)
|
|
|
|
|
return f
|
|
|
|
|
return {}
|
|
|
|
|
|
2022-05-29 09:48:04 -04:00
|
|
|
|
def _media_formats(src, cur_media_type, type_info=None):
|
|
|
|
|
type_info = type_info or {}
|
2016-08-19 11:53:47 -04:00
|
|
|
|
full_url = absolute_url(src)
|
2017-08-12 05:48:11 -04:00
|
|
|
|
ext = type_info.get('ext') or determine_ext(full_url)
|
2016-12-17 11:03:13 -05:00
|
|
|
|
if ext == 'm3u8':
|
2016-08-19 11:53:47 -04:00
|
|
|
|
is_plain_url = False
|
|
|
|
|
formats = self._extract_m3u8_formats(
|
2016-08-21 14:18:46 -04:00
|
|
|
|
full_url, video_id, ext='mp4',
|
2017-02-25 05:40:05 -05:00
|
|
|
|
entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id,
|
2021-02-18 17:03:16 -05:00
|
|
|
|
preference=preference, quality=quality, fatal=False)
|
2016-12-17 11:03:13 -05:00
|
|
|
|
elif ext == 'mpd':
|
|
|
|
|
is_plain_url = False
|
|
|
|
|
formats = self._extract_mpd_formats(
|
2017-08-20 03:16:58 -04:00
|
|
|
|
full_url, video_id, mpd_id=mpd_id, fatal=False)
|
2016-08-19 11:53:47 -04:00
|
|
|
|
else:
|
|
|
|
|
is_plain_url = True
|
|
|
|
|
formats = [{
|
|
|
|
|
'url': full_url,
|
|
|
|
|
'vcodec': 'none' if cur_media_type == 'audio' else None,
|
2022-05-29 09:48:04 -04:00
|
|
|
|
'ext': ext,
|
2016-08-19 11:53:47 -04:00
|
|
|
|
}]
|
|
|
|
|
return is_plain_url, formats
|
|
|
|
|
|
2016-03-16 13:50:45 -04:00
|
|
|
|
entries = []
|
2017-07-09 05:29:52 -04:00
|
|
|
|
# amp-video and amp-audio are very similar to their HTML5 counterparts
|
2022-06-26 20:50:06 -04:00
|
|
|
|
# so we will include them right here (see
|
2017-07-09 05:29:52 -04:00
|
|
|
|
# https://www.ampproject.org/docs/reference/components/amp-video)
|
2021-01-01 07:26:37 -05:00
|
|
|
|
# For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
|
|
|
|
|
_MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
|
|
|
|
|
media_tags = [(media_tag, media_tag_name, media_type, '')
|
|
|
|
|
for media_tag, media_tag_name, media_type
|
|
|
|
|
in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)]
|
2017-02-05 12:20:30 -05:00
|
|
|
|
media_tags.extend(re.findall(
|
|
|
|
|
# We only allow video|audio followed by a whitespace or '>'.
|
|
|
|
|
# Allowing more characters may end up in significant slow down (see
|
2022-08-14 08:04:13 -04:00
|
|
|
|
# https://github.com/ytdl-org/youtube-dl/issues/11979,
|
|
|
|
|
# e.g. http://www.porntrex.com/maps/videositemap.xml).
|
2021-01-01 07:26:37 -05:00
|
|
|
|
r'(?s)(<(?P<tag>%s)(?:\s+[^>]*)?>)(.*?)</(?P=tag)>' % _MEDIA_TAG_NAME_RE, webpage))
|
|
|
|
|
for media_tag, _, media_type, media_content in media_tags:
|
2016-03-16 13:50:45 -04:00
|
|
|
|
media_info = {
|
|
|
|
|
'formats': [],
|
|
|
|
|
'subtitles': {},
|
|
|
|
|
}
|
|
|
|
|
media_attributes = extract_attributes(media_tag)
|
2022-08-30 13:07:55 -04:00
|
|
|
|
src = strip_or_none(dict_get(media_attributes, ('src', 'data-video-src', 'data-src', 'data-source')))
|
2016-03-16 13:50:45 -04:00
|
|
|
|
if src:
|
2022-05-29 09:48:04 -04:00
|
|
|
|
f = parse_content_type(media_attributes.get('type'))
|
|
|
|
|
_, formats = _media_formats(src, media_type, f)
|
2016-08-19 11:53:47 -04:00
|
|
|
|
media_info['formats'].extend(formats)
|
2018-03-19 12:43:53 -04:00
|
|
|
|
media_info['thumbnail'] = absolute_url(media_attributes.get('poster'))
|
2016-03-16 13:50:45 -04:00
|
|
|
|
if media_content:
|
|
|
|
|
for source_tag in re.findall(r'<source[^>]+>', media_content):
|
2019-03-16 22:09:32 -04:00
|
|
|
|
s_attr = extract_attributes(source_tag)
|
|
|
|
|
# data-video-src and data-src are non standard but seen
|
|
|
|
|
# several times in the wild
|
2022-08-30 13:07:55 -04:00
|
|
|
|
src = strip_or_none(dict_get(s_attr, ('src', 'data-video-src', 'data-src', 'data-source')))
|
2016-03-16 13:50:45 -04:00
|
|
|
|
if not src:
|
|
|
|
|
continue
|
2019-03-16 22:09:32 -04:00
|
|
|
|
f = parse_content_type(s_attr.get('type'))
|
2017-08-12 08:24:26 -04:00
|
|
|
|
is_plain_url, formats = _media_formats(src, media_type, f)
|
2016-08-19 11:53:47 -04:00
|
|
|
|
if is_plain_url:
|
2019-03-16 22:09:32 -04:00
|
|
|
|
# width, height, res, label and title attributes are
|
|
|
|
|
# all not standard but seen several times in the wild
|
|
|
|
|
labels = [
|
|
|
|
|
s_attr.get(lbl)
|
|
|
|
|
for lbl in ('label', 'title')
|
|
|
|
|
if str_or_none(s_attr.get(lbl))
|
|
|
|
|
]
|
|
|
|
|
width = int_or_none(s_attr.get('width'))
|
2019-05-10 16:56:22 -04:00
|
|
|
|
height = (int_or_none(s_attr.get('height'))
|
|
|
|
|
or int_or_none(s_attr.get('res')))
|
2019-03-16 22:09:32 -04:00
|
|
|
|
if not width or not height:
|
|
|
|
|
for lbl in labels:
|
|
|
|
|
resolution = parse_resolution(lbl)
|
|
|
|
|
if not resolution:
|
|
|
|
|
continue
|
|
|
|
|
width = width or resolution.get('width')
|
|
|
|
|
height = height or resolution.get('height')
|
|
|
|
|
for lbl in labels:
|
|
|
|
|
tbr = parse_bitrate(lbl)
|
|
|
|
|
if tbr:
|
|
|
|
|
break
|
|
|
|
|
else:
|
|
|
|
|
tbr = None
|
2017-08-26 16:27:05 -04:00
|
|
|
|
f.update({
|
2019-03-16 22:09:32 -04:00
|
|
|
|
'width': width,
|
|
|
|
|
'height': height,
|
|
|
|
|
'tbr': tbr,
|
|
|
|
|
'format_id': s_attr.get('label') or s_attr.get('title'),
|
2017-08-26 16:27:05 -04:00
|
|
|
|
})
|
2016-08-19 11:53:47 -04:00
|
|
|
|
f.update(formats[0])
|
|
|
|
|
media_info['formats'].append(f)
|
|
|
|
|
else:
|
|
|
|
|
media_info['formats'].extend(formats)
|
2016-03-16 13:50:45 -04:00
|
|
|
|
for track_tag in re.findall(r'<track[^>]+>', media_content):
|
|
|
|
|
track_attributes = extract_attributes(track_tag)
|
|
|
|
|
kind = track_attributes.get('kind')
|
2016-09-24 02:20:42 -04:00
|
|
|
|
if not kind or kind in ('subtitles', 'captions'):
|
2019-05-23 12:52:11 -04:00
|
|
|
|
src = strip_or_none(track_attributes.get('src'))
|
2016-03-16 13:50:45 -04:00
|
|
|
|
if not src:
|
|
|
|
|
continue
|
|
|
|
|
lang = track_attributes.get('srclang') or track_attributes.get('lang') or track_attributes.get('label')
|
|
|
|
|
media_info['subtitles'].setdefault(lang, []).append({
|
|
|
|
|
'url': absolute_url(src),
|
|
|
|
|
})
|
2018-06-28 14:25:05 -04:00
|
|
|
|
for f in media_info['formats']:
|
|
|
|
|
f.setdefault('http_headers', {})['Referer'] = base_url
|
2016-09-24 02:20:42 -04:00
|
|
|
|
if media_info['formats'] or media_info['subtitles']:
|
2016-03-16 13:50:45 -04:00
|
|
|
|
entries.append(media_info)
|
|
|
|
|
return entries
|
|
|
|
|
|
2021-04-22 11:19:26 -04:00
|
|
|
|
def _extract_akamai_formats(self, *args, **kwargs):
|
|
|
|
|
fmts, subs = self._extract_akamai_formats_and_subtitles(*args, **kwargs)
|
|
|
|
|
if subs:
|
2021-10-08 20:23:15 -04:00
|
|
|
|
self._report_ignoring_subs('akamai')
|
2021-04-22 11:19:26 -04:00
|
|
|
|
return fmts
|
|
|
|
|
|
|
|
|
|
def _extract_akamai_formats_and_subtitles(self, manifest_url, video_id, hosts={}):
|
2021-01-01 07:26:37 -05:00
|
|
|
|
signed = 'hdnea=' in manifest_url
|
|
|
|
|
if not signed:
|
|
|
|
|
# https://learn.akamai.com/en-us/webhelp/media-services-on-demand/stream-packaging-user-guide/GUID-BE6C0F73-1E06-483B-B0EA-57984B91B7F9.html
|
|
|
|
|
manifest_url = re.sub(
|
|
|
|
|
r'(?:b=[\d,-]+|(?:__a__|attributes)=off|__b__=\d+)&?',
|
|
|
|
|
'', manifest_url).strip('?')
|
|
|
|
|
|
2016-08-22 02:47:25 -04:00
|
|
|
|
formats = []
|
2021-04-22 11:19:26 -04:00
|
|
|
|
subtitles = {}
|
2020-11-23 16:03:08 -05:00
|
|
|
|
|
2016-09-24 16:55:53 -04:00
|
|
|
|
hdcore_sign = 'hdcore=3.7.0'
|
2017-05-04 11:04:25 -04:00
|
|
|
|
f4m_url = re.sub(r'(https?://[^/]+)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m')
|
2017-01-13 04:08:51 -05:00
|
|
|
|
hds_host = hosts.get('hds')
|
|
|
|
|
if hds_host:
|
|
|
|
|
f4m_url = re.sub(r'(https?://)[^/]+', r'\1' + hds_host, f4m_url)
|
2016-09-24 16:55:53 -04:00
|
|
|
|
if 'hdcore=' not in f4m_url:
|
|
|
|
|
f4m_url += ('&' if '?' in f4m_url else '?') + hdcore_sign
|
|
|
|
|
f4m_formats = self._extract_f4m_formats(
|
|
|
|
|
f4m_url, video_id, f4m_id='hds', fatal=False)
|
|
|
|
|
for entry in f4m_formats:
|
|
|
|
|
entry.update({'extra_param_to_segment_url': hdcore_sign})
|
|
|
|
|
formats.extend(f4m_formats)
|
2020-11-23 16:03:08 -05:00
|
|
|
|
|
2017-01-13 04:08:51 -05:00
|
|
|
|
m3u8_url = re.sub(r'(https?://[^/]+)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8')
|
|
|
|
|
hls_host = hosts.get('hls')
|
|
|
|
|
if hls_host:
|
|
|
|
|
m3u8_url = re.sub(r'(https?://)[^/]+', r'\1' + hls_host, m3u8_url)
|
2021-04-22 11:19:26 -04:00
|
|
|
|
m3u8_formats, m3u8_subtitles = self._extract_m3u8_formats_and_subtitles(
|
2016-08-22 02:47:25 -04:00
|
|
|
|
m3u8_url, video_id, 'mp4', 'm3u8_native',
|
2021-01-01 07:26:37 -05:00
|
|
|
|
m3u8_id='hls', fatal=False)
|
|
|
|
|
formats.extend(m3u8_formats)
|
2021-04-22 11:19:26 -04:00
|
|
|
|
subtitles = self._merge_subtitles(subtitles, m3u8_subtitles)
|
2020-11-23 16:03:08 -05:00
|
|
|
|
|
|
|
|
|
http_host = hosts.get('http')
|
2021-01-01 07:26:37 -05:00
|
|
|
|
if http_host and m3u8_formats and not signed:
|
|
|
|
|
REPL_REGEX = r'https?://[^/]+/i/([^,]+),([^/]+),([^/]+)\.csmil/.+'
|
2020-11-23 16:03:08 -05:00
|
|
|
|
qualities = re.match(REPL_REGEX, m3u8_url).group(2).split(',')
|
|
|
|
|
qualities_length = len(qualities)
|
2021-01-01 07:26:37 -05:00
|
|
|
|
if len(m3u8_formats) in (qualities_length, qualities_length + 1):
|
2020-11-23 16:03:08 -05:00
|
|
|
|
i = 0
|
2021-01-01 07:26:37 -05:00
|
|
|
|
for f in m3u8_formats:
|
|
|
|
|
if f['vcodec'] != 'none':
|
2020-11-23 16:03:08 -05:00
|
|
|
|
for protocol in ('http', 'https'):
|
|
|
|
|
http_f = f.copy()
|
|
|
|
|
del http_f['manifest_url']
|
|
|
|
|
http_url = re.sub(
|
2022-04-11 11:10:28 -04:00
|
|
|
|
REPL_REGEX, protocol + fr'://{http_host}/\g<1>{qualities[i]}\3', f['url'])
|
2020-11-23 16:03:08 -05:00
|
|
|
|
http_f.update({
|
|
|
|
|
'format_id': http_f['format_id'].replace('hls-', protocol + '-'),
|
|
|
|
|
'url': http_url,
|
|
|
|
|
'protocol': protocol,
|
|
|
|
|
})
|
2021-01-01 07:26:37 -05:00
|
|
|
|
formats.append(http_f)
|
2020-11-23 16:03:08 -05:00
|
|
|
|
i += 1
|
|
|
|
|
|
2021-04-22 11:19:26 -04:00
|
|
|
|
return formats, subtitles
|
2016-08-22 02:47:25 -04:00
|
|
|
|
|
2016-09-16 14:30:38 -04:00
|
|
|
|
def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]):
|
2022-06-24 06:54:43 -04:00
|
|
|
|
query = urllib.parse.urlparse(url).query
|
2016-09-16 14:30:38 -04:00
|
|
|
|
url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url)
|
2018-02-05 11:41:55 -05:00
|
|
|
|
mobj = re.search(
|
|
|
|
|
r'(?:(?:http|rtmp|rtsp)(?P<s>s)?:)?(?P<url>//[^?]+)', url)
|
|
|
|
|
url_base = mobj.group('url')
|
|
|
|
|
http_base_url = '%s%s:%s' % ('http', mobj.group('s') or '', url_base)
|
2016-09-16 14:30:38 -04:00
|
|
|
|
formats = []
|
2017-11-01 12:39:26 -04:00
|
|
|
|
|
|
|
|
|
def manifest_url(manifest):
|
2022-04-11 11:10:28 -04:00
|
|
|
|
m_url = f'{http_base_url}/{manifest}'
|
2017-11-01 12:39:26 -04:00
|
|
|
|
if query:
|
|
|
|
|
m_url += '?%s' % query
|
|
|
|
|
return m_url
|
|
|
|
|
|
2016-09-16 14:30:38 -04:00
|
|
|
|
if 'm3u8' not in skip_protocols:
|
|
|
|
|
formats.extend(self._extract_m3u8_formats(
|
2017-11-01 12:39:26 -04:00
|
|
|
|
manifest_url('playlist.m3u8'), video_id, 'mp4',
|
2016-09-16 14:30:38 -04:00
|
|
|
|
m3u8_entry_protocol, m3u8_id='hls', fatal=False))
|
|
|
|
|
if 'f4m' not in skip_protocols:
|
|
|
|
|
formats.extend(self._extract_f4m_formats(
|
2017-11-01 12:39:26 -04:00
|
|
|
|
manifest_url('manifest.f4m'),
|
2016-09-16 14:30:38 -04:00
|
|
|
|
video_id, f4m_id='hds', fatal=False))
|
2016-10-19 09:57:12 -04:00
|
|
|
|
if 'dash' not in skip_protocols:
|
|
|
|
|
formats.extend(self._extract_mpd_formats(
|
2017-11-01 12:39:26 -04:00
|
|
|
|
manifest_url('manifest.mpd'),
|
2016-10-19 09:57:12 -04:00
|
|
|
|
video_id, mpd_id='dash', fatal=False))
|
2016-09-16 14:30:38 -04:00
|
|
|
|
if re.search(r'(?:/smil:|\.smil)', url_base):
|
|
|
|
|
if 'smil' not in skip_protocols:
|
|
|
|
|
rtmp_formats = self._extract_smil_formats(
|
2017-11-01 12:39:26 -04:00
|
|
|
|
manifest_url('jwplayer.smil'),
|
2016-09-16 14:30:38 -04:00
|
|
|
|
video_id, fatal=False)
|
|
|
|
|
for rtmp_format in rtmp_formats:
|
|
|
|
|
rtsp_format = rtmp_format.copy()
|
|
|
|
|
rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path'])
|
|
|
|
|
del rtsp_format['play_path']
|
|
|
|
|
del rtsp_format['ext']
|
|
|
|
|
rtsp_format.update({
|
|
|
|
|
'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'),
|
|
|
|
|
'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'),
|
|
|
|
|
'protocol': 'rtsp',
|
|
|
|
|
})
|
|
|
|
|
formats.extend([rtmp_format, rtsp_format])
|
|
|
|
|
else:
|
|
|
|
|
for protocol in ('rtmp', 'rtsp'):
|
|
|
|
|
if protocol not in skip_protocols:
|
|
|
|
|
formats.append({
|
2022-04-11 11:10:28 -04:00
|
|
|
|
'url': f'{protocol}:{url_base}',
|
2016-09-16 14:30:38 -04:00
|
|
|
|
'format_id': protocol,
|
|
|
|
|
'protocol': protocol,
|
|
|
|
|
})
|
|
|
|
|
return formats
|
|
|
|
|
|
2017-03-25 14:38:30 -04:00
|
|
|
|
def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json):
|
2017-02-16 10:42:36 -05:00
|
|
|
|
mobj = re.search(
|
2017-04-25 12:46:05 -04:00
|
|
|
|
r'(?s)jwplayer\((?P<quote>[\'"])[^\'" ]+(?P=quote)\)(?!</script>).*?\.setup\s*\((?P<options>[^)]+)\)',
|
2017-02-16 10:42:36 -05:00
|
|
|
|
webpage)
|
|
|
|
|
if mobj:
|
2017-03-25 14:38:30 -04:00
|
|
|
|
try:
|
|
|
|
|
jwplayer_data = self._parse_json(mobj.group('options'),
|
|
|
|
|
video_id=video_id,
|
|
|
|
|
transform_source=transform_source)
|
|
|
|
|
except ExtractorError:
|
|
|
|
|
pass
|
|
|
|
|
else:
|
|
|
|
|
if isinstance(jwplayer_data, dict):
|
|
|
|
|
return jwplayer_data
|
2017-02-16 10:42:36 -05:00
|
|
|
|
|
|
|
|
|
def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
|
2017-03-25 14:38:30 -04:00
|
|
|
|
jwplayer_data = self._find_jwplayer_data(
|
|
|
|
|
webpage, video_id, transform_source=js_to_json)
|
2017-02-16 10:42:36 -05:00
|
|
|
|
return self._parse_jwplayer_data(
|
|
|
|
|
jwplayer_data, video_id, *args, **kwargs)
|
|
|
|
|
|
|
|
|
|
def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
|
|
|
|
|
m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
|
|
|
|
|
# JWPlayer backward compatibility: flattened playlists
|
|
|
|
|
# https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/api/config.js#L81-L96
|
|
|
|
|
if 'playlist' not in jwplayer_data:
|
|
|
|
|
jwplayer_data = {'playlist': [jwplayer_data]}
|
|
|
|
|
|
|
|
|
|
entries = []
|
|
|
|
|
|
|
|
|
|
# JWPlayer backward compatibility: single playlist item
|
|
|
|
|
# https://github.com/jwplayer/jwplayer/blob/v7.7.0/src/js/playlist/playlist.js#L10
|
|
|
|
|
if not isinstance(jwplayer_data['playlist'], list):
|
|
|
|
|
jwplayer_data['playlist'] = [jwplayer_data['playlist']]
|
|
|
|
|
|
|
|
|
|
for video_data in jwplayer_data['playlist']:
|
|
|
|
|
# JWPlayer backward compatibility: flattened sources
|
|
|
|
|
# https://github.com/jwplayer/jwplayer/blob/v7.4.3/src/js/playlist/item.js#L29-L35
|
|
|
|
|
if 'sources' not in video_data:
|
|
|
|
|
video_data['sources'] = [video_data]
|
|
|
|
|
|
|
|
|
|
this_video_id = video_id or video_data['mediaid']
|
|
|
|
|
|
2017-03-05 11:28:32 -05:00
|
|
|
|
formats = self._parse_jwplayer_formats(
|
|
|
|
|
video_data['sources'], video_id=this_video_id, m3u8_id=m3u8_id,
|
|
|
|
|
mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
|
2017-02-16 10:42:36 -05:00
|
|
|
|
|
|
|
|
|
subtitles = {}
|
|
|
|
|
tracks = video_data.get('tracks')
|
|
|
|
|
if tracks and isinstance(tracks, list):
|
|
|
|
|
for track in tracks:
|
2017-06-15 12:40:39 -04:00
|
|
|
|
if not isinstance(track, dict):
|
|
|
|
|
continue
|
2018-02-24 12:59:29 -05:00
|
|
|
|
track_kind = track.get('kind')
|
2022-06-24 06:54:43 -04:00
|
|
|
|
if not track_kind or not isinstance(track_kind, str):
|
2018-02-24 12:59:29 -05:00
|
|
|
|
continue
|
|
|
|
|
if track_kind.lower() not in ('captions', 'subtitles'):
|
2017-02-16 10:42:36 -05:00
|
|
|
|
continue
|
|
|
|
|
track_url = urljoin(base_url, track.get('file'))
|
|
|
|
|
if not track_url:
|
|
|
|
|
continue
|
|
|
|
|
subtitles.setdefault(track.get('label') or 'en', []).append({
|
|
|
|
|
'url': self._proto_relative_url(track_url)
|
|
|
|
|
})
|
|
|
|
|
|
2017-10-12 12:12:47 -04:00
|
|
|
|
entry = {
|
2017-02-16 10:42:36 -05:00
|
|
|
|
'id': this_video_id,
|
2017-10-12 12:12:47 -04:00
|
|
|
|
'title': unescapeHTML(video_data['title'] if require_title else video_data.get('title')),
|
2019-11-09 07:11:59 -05:00
|
|
|
|
'description': clean_html(video_data.get('description')),
|
2019-01-20 07:31:41 -05:00
|
|
|
|
'thumbnail': urljoin(base_url, self._proto_relative_url(video_data.get('image'))),
|
2017-02-16 10:42:36 -05:00
|
|
|
|
'timestamp': int_or_none(video_data.get('pubdate')),
|
|
|
|
|
'duration': float_or_none(jwplayer_data.get('duration') or video_data.get('duration')),
|
|
|
|
|
'subtitles': subtitles,
|
2017-10-12 12:12:47 -04:00
|
|
|
|
}
|
|
|
|
|
# https://github.com/jwplayer/jwplayer/blob/master/src/js/utils/validator.js#L32
|
|
|
|
|
if len(formats) == 1 and re.search(r'^(?:http|//).*(?:youtube\.com|youtu\.be)/.+', formats[0]['url']):
|
|
|
|
|
entry.update({
|
|
|
|
|
'_type': 'url_transparent',
|
|
|
|
|
'url': formats[0]['url'],
|
|
|
|
|
})
|
|
|
|
|
else:
|
|
|
|
|
self._sort_formats(formats)
|
|
|
|
|
entry['formats'] = formats
|
|
|
|
|
entries.append(entry)
|
2017-02-16 10:42:36 -05:00
|
|
|
|
if len(entries) == 1:
|
|
|
|
|
return entries[0]
|
|
|
|
|
else:
|
|
|
|
|
return self.playlist_result(entries)
|
|
|
|
|
|
2017-03-05 11:22:27 -05:00
|
|
|
|
def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
|
|
|
|
|
m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
|
2017-04-17 03:48:24 -04:00
|
|
|
|
urls = []
|
2017-03-05 11:22:27 -05:00
|
|
|
|
formats = []
|
2017-03-05 11:28:32 -05:00
|
|
|
|
for source in jwplayer_sources_data:
|
2017-06-14 11:02:15 -04:00
|
|
|
|
if not isinstance(source, dict):
|
|
|
|
|
continue
|
2019-01-20 07:31:41 -05:00
|
|
|
|
source_url = urljoin(
|
|
|
|
|
base_url, self._proto_relative_url(source.get('file')))
|
|
|
|
|
if not source_url or source_url in urls:
|
2017-04-17 03:48:24 -04:00
|
|
|
|
continue
|
|
|
|
|
urls.append(source_url)
|
2017-03-05 11:22:27 -05:00
|
|
|
|
source_type = source.get('type') or ''
|
|
|
|
|
ext = mimetype2ext(source_type) or determine_ext(source_url)
|
|
|
|
|
if source_type == 'hls' or ext == 'm3u8':
|
|
|
|
|
formats.extend(self._extract_m3u8_formats(
|
2017-03-05 11:25:03 -05:00
|
|
|
|
source_url, video_id, 'mp4', entry_protocol='m3u8_native',
|
|
|
|
|
m3u8_id=m3u8_id, fatal=False))
|
2018-01-21 05:42:48 -05:00
|
|
|
|
elif source_type == 'dash' or ext == 'mpd':
|
2017-03-05 11:22:27 -05:00
|
|
|
|
formats.extend(self._extract_mpd_formats(
|
|
|
|
|
source_url, video_id, mpd_id=mpd_id, fatal=False))
|
2017-03-15 16:30:53 -04:00
|
|
|
|
elif ext == 'smil':
|
|
|
|
|
formats.extend(self._extract_smil_formats(
|
|
|
|
|
source_url, video_id, fatal=False))
|
2017-03-05 11:22:27 -05:00
|
|
|
|
# https://github.com/jwplayer/jwplayer/blob/master/src/js/providers/default.js#L67
|
2017-03-05 11:25:03 -05:00
|
|
|
|
elif source_type.startswith('audio') or ext in (
|
|
|
|
|
'oga', 'aac', 'mp3', 'mpeg', 'vorbis'):
|
2017-03-05 11:22:27 -05:00
|
|
|
|
formats.append({
|
|
|
|
|
'url': source_url,
|
|
|
|
|
'vcodec': 'none',
|
|
|
|
|
'ext': ext,
|
|
|
|
|
})
|
|
|
|
|
else:
|
|
|
|
|
height = int_or_none(source.get('height'))
|
|
|
|
|
if height is None:
|
|
|
|
|
# Often no height is provided but there is a label in
|
2017-03-05 11:25:03 -05:00
|
|
|
|
# format like "1080p", "720p SD", or 1080.
|
2017-03-05 11:22:27 -05:00
|
|
|
|
height = int_or_none(self._search_regex(
|
2022-06-24 06:54:43 -04:00
|
|
|
|
r'^(\d{3,4})[pP]?(?:\b|$)', str(source.get('label') or ''),
|
2017-03-05 11:22:27 -05:00
|
|
|
|
'height', default=None))
|
|
|
|
|
a_format = {
|
|
|
|
|
'url': source_url,
|
|
|
|
|
'width': int_or_none(source.get('width')),
|
|
|
|
|
'height': height,
|
2022-10-03 15:37:48 -04:00
|
|
|
|
'tbr': int_or_none(source.get('bitrate'), scale=1000),
|
|
|
|
|
'filesize': int_or_none(source.get('filesize')),
|
2017-03-05 11:22:27 -05:00
|
|
|
|
'ext': ext,
|
|
|
|
|
}
|
|
|
|
|
if source_url.startswith('rtmp'):
|
|
|
|
|
a_format['ext'] = 'flv'
|
|
|
|
|
# See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
|
|
|
|
|
# of jwplayer.flash.swf
|
|
|
|
|
rtmp_url_parts = re.split(
|
|
|
|
|
r'((?:mp4|mp3|flv):)', source_url, 1)
|
|
|
|
|
if len(rtmp_url_parts) == 3:
|
|
|
|
|
rtmp_url, prefix, play_path = rtmp_url_parts
|
|
|
|
|
a_format.update({
|
|
|
|
|
'url': rtmp_url,
|
|
|
|
|
'play_path': prefix + play_path,
|
|
|
|
|
})
|
|
|
|
|
if rtmp_params:
|
|
|
|
|
a_format.update(rtmp_params)
|
|
|
|
|
formats.append(a_format)
|
|
|
|
|
return formats
|
|
|
|
|
|
2014-09-28 02:53:52 -04:00
|
|
|
|
def _live_title(self, name):
|
2021-12-15 11:00:46 -05:00
|
|
|
|
self._downloader.deprecation_warning('yt_dlp.InfoExtractor._live_title is deprecated and does not work as expected')
|
|
|
|
|
return name
|
2014-09-28 02:53:52 -04:00
|
|
|
|
|
2014-09-28 04:34:55 -04:00
|
|
|
|
def _int(self, v, name, fatal=False, **kwargs):
|
|
|
|
|
res = int_or_none(v, **kwargs)
|
|
|
|
|
if res is None:
|
2022-04-11 11:10:28 -04:00
|
|
|
|
msg = f'Failed to extract {name}: Could not parse value {v!r}'
|
2014-09-28 04:34:55 -04:00
|
|
|
|
if fatal:
|
|
|
|
|
raise ExtractorError(msg)
|
|
|
|
|
else:
|
2021-04-16 06:01:10 -04:00
|
|
|
|
self.report_warning(msg)
|
2014-09-28 04:34:55 -04:00
|
|
|
|
return res
|
|
|
|
|
|
|
|
|
|
def _float(self, v, name, fatal=False, **kwargs):
|
|
|
|
|
res = float_or_none(v, **kwargs)
|
|
|
|
|
if res is None:
|
2022-04-11 11:10:28 -04:00
|
|
|
|
msg = f'Failed to extract {name}: Could not parse value {v!r}'
|
2014-09-28 04:34:55 -04:00
|
|
|
|
if fatal:
|
|
|
|
|
raise ExtractorError(msg)
|
|
|
|
|
else:
|
2021-04-16 06:01:10 -04:00
|
|
|
|
self.report_warning(msg)
|
2014-09-28 04:34:55 -04:00
|
|
|
|
return res
|
|
|
|
|
|
2017-04-25 09:12:54 -04:00
|
|
|
|
def _set_cookie(self, domain, name, value, expire_time=None, port=None,
|
|
|
|
|
path='/', secure=False, discard=False, rest={}, **kwargs):
|
2022-06-24 04:10:17 -04:00
|
|
|
|
cookie = http.cookiejar.Cookie(
|
2017-09-17 01:53:04 -04:00
|
|
|
|
0, name, value, port, port is not None, domain, True,
|
2017-04-25 09:12:54 -04:00
|
|
|
|
domain.startswith('.'), path, True, secure, expire_time,
|
|
|
|
|
discard, None, None, rest)
|
2022-06-23 00:14:22 -04:00
|
|
|
|
self.cookiejar.set_cookie(cookie)
|
2014-11-29 18:03:59 -05:00
|
|
|
|
|
2015-07-29 18:20:37 -04:00
|
|
|
|
def _get_cookies(self, url):
|
2022-06-24 04:10:17 -04:00
|
|
|
|
""" Return a http.cookies.SimpleCookie with the cookies for the url """
|
2022-09-16 13:02:00 -04:00
|
|
|
|
return LenientSimpleCookie(self._downloader._calc_cookies(url))
|
2015-07-29 18:20:37 -04:00
|
|
|
|
|
2019-05-17 16:17:15 -04:00
|
|
|
|
def _apply_first_set_cookie_header(self, url_handle, cookie):
|
2019-05-20 12:23:18 -04:00
|
|
|
|
"""
|
|
|
|
|
Apply first Set-Cookie header instead of the last. Experimental.
|
|
|
|
|
|
|
|
|
|
Some sites (e.g. [1-3]) may serve two cookies under the same name
|
|
|
|
|
in Set-Cookie header and expect the first (old) one to be set rather
|
|
|
|
|
than second (new). However, as of RFC6265 the newer one cookie
|
|
|
|
|
should be set into cookie store what actually happens.
|
|
|
|
|
We will workaround this issue by resetting the cookie to
|
|
|
|
|
the first one manually.
|
|
|
|
|
1. https://new.vk.com/
|
|
|
|
|
2. https://github.com/ytdl-org/youtube-dl/issues/9841#issuecomment-227871201
|
|
|
|
|
3. https://learning.oreilly.com/
|
|
|
|
|
"""
|
2019-05-17 16:17:15 -04:00
|
|
|
|
for header, cookies in url_handle.headers.items():
|
|
|
|
|
if header.lower() != 'set-cookie':
|
|
|
|
|
continue
|
2021-12-30 07:23:36 -05:00
|
|
|
|
cookies = cookies.encode('iso-8859-1').decode('utf-8')
|
2019-05-17 16:17:15 -04:00
|
|
|
|
cookie_value = re.search(
|
|
|
|
|
r'%s=(.+?);.*?\b[Dd]omain=(.+?)(?:[,;]|$)' % cookie, cookies)
|
|
|
|
|
if cookie_value:
|
|
|
|
|
value, domain = cookie_value.groups()
|
|
|
|
|
self._set_cookie(domain, cookie, value)
|
|
|
|
|
break
|
|
|
|
|
|
2022-05-11 11:54:44 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def get_testcases(cls, include_onlymatching=False):
|
|
|
|
|
t = getattr(cls, '_TEST', None)
|
2015-01-07 01:20:20 -05:00
|
|
|
|
if t:
|
2022-05-11 11:54:44 -04:00
|
|
|
|
assert not hasattr(cls, '_TESTS'), f'{cls.ie_key()}IE has _TEST and _TESTS'
|
2015-01-07 01:20:20 -05:00
|
|
|
|
tests = [t]
|
|
|
|
|
else:
|
2022-05-11 11:54:44 -04:00
|
|
|
|
tests = getattr(cls, '_TESTS', [])
|
2015-01-07 01:20:20 -05:00
|
|
|
|
for t in tests:
|
|
|
|
|
if not include_onlymatching and t.get('only_matching', False):
|
|
|
|
|
continue
|
2022-05-11 11:54:44 -04:00
|
|
|
|
t['name'] = cls.ie_key()
|
2015-01-07 01:20:20 -05:00
|
|
|
|
yield t
|
|
|
|
|
|
2022-07-08 07:23:05 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def get_webpage_testcases(cls):
|
|
|
|
|
tests = getattr(cls, '_WEBPAGE_TESTS', [])
|
|
|
|
|
for t in tests:
|
|
|
|
|
t['name'] = cls.ie_key()
|
|
|
|
|
return tests
|
|
|
|
|
|
2022-05-16 10:06:36 -04:00
|
|
|
|
@classproperty
|
|
|
|
|
def age_limit(cls):
|
|
|
|
|
"""Get age limit from the testcases"""
|
|
|
|
|
return max(traverse_obj(
|
2022-07-08 07:23:05 -04:00
|
|
|
|
(*cls.get_testcases(include_onlymatching=False), *cls.get_webpage_testcases()),
|
2022-05-16 10:06:36 -04:00
|
|
|
|
(..., (('playlist', 0), None), 'info_dict', 'age_limit')) or [0])
|
|
|
|
|
|
2022-05-11 11:54:44 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def is_suitable(cls, age_limit):
|
2022-05-16 10:06:36 -04:00
|
|
|
|
"""Test whether the extractor is generally suitable for the given age limit"""
|
|
|
|
|
return not age_restricted(cls.age_limit, age_limit)
|
2015-01-07 01:20:20 -05:00
|
|
|
|
|
2022-05-11 11:54:44 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def description(cls, *, markdown=True, search_examples=None):
|
2022-05-09 00:32:17 -04:00
|
|
|
|
"""Description of the extractor"""
|
|
|
|
|
desc = ''
|
2022-05-11 11:54:44 -04:00
|
|
|
|
if cls._NETRC_MACHINE:
|
2022-05-09 00:32:17 -04:00
|
|
|
|
if markdown:
|
2022-05-11 11:54:44 -04:00
|
|
|
|
desc += f' [<abbr title="netrc machine"><em>{cls._NETRC_MACHINE}</em></abbr>]'
|
2022-05-09 00:32:17 -04:00
|
|
|
|
else:
|
2022-05-11 11:54:44 -04:00
|
|
|
|
desc += f' [{cls._NETRC_MACHINE}]'
|
|
|
|
|
if cls.IE_DESC is False:
|
2022-05-09 00:32:17 -04:00
|
|
|
|
desc += ' [HIDDEN]'
|
2022-05-11 11:54:44 -04:00
|
|
|
|
elif cls.IE_DESC:
|
|
|
|
|
desc += f' {cls.IE_DESC}'
|
|
|
|
|
if cls.SEARCH_KEY:
|
|
|
|
|
desc += f'; "{cls.SEARCH_KEY}:" prefix'
|
2022-05-09 00:32:17 -04:00
|
|
|
|
if search_examples:
|
|
|
|
|
_COUNTS = ('', '5', '10', 'all')
|
2022-08-14 08:04:13 -04:00
|
|
|
|
desc += f' (e.g. "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
|
2022-05-11 11:54:44 -04:00
|
|
|
|
if not cls.working():
|
2022-05-09 00:32:17 -04:00
|
|
|
|
desc += ' (**Currently broken**)' if markdown else ' (Currently broken)'
|
|
|
|
|
|
2022-05-11 11:54:44 -04:00
|
|
|
|
name = f' - **{cls.IE_NAME}**' if markdown else cls.IE_NAME
|
2022-05-09 00:32:17 -04:00
|
|
|
|
return f'{name}:{desc}' if desc else name
|
|
|
|
|
|
2015-02-15 12:03:41 -05:00
|
|
|
|
def extract_subtitles(self, *args, **kwargs):
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if (self.get_param('writesubtitles', False)
|
|
|
|
|
or self.get_param('listsubtitles')):
|
2015-02-17 16:16:29 -05:00
|
|
|
|
return self._get_subtitles(*args, **kwargs)
|
|
|
|
|
return {}
|
2015-02-15 12:03:41 -05:00
|
|
|
|
|
|
|
|
|
def _get_subtitles(self, *args, **kwargs):
|
2016-02-14 04:37:17 -05:00
|
|
|
|
raise NotImplementedError('This method must be implemented by subclasses')
|
2015-02-15 12:03:41 -05:00
|
|
|
|
|
2021-10-12 05:50:50 -04:00
|
|
|
|
def extract_comments(self, *args, **kwargs):
|
|
|
|
|
if not self.get_param('getcomments'):
|
|
|
|
|
return None
|
|
|
|
|
generator = self._get_comments(*args, **kwargs)
|
|
|
|
|
|
|
|
|
|
def extractor():
|
|
|
|
|
comments = []
|
2021-12-02 17:00:37 -05:00
|
|
|
|
interrupted = True
|
2021-10-12 05:50:50 -04:00
|
|
|
|
try:
|
|
|
|
|
while True:
|
|
|
|
|
comments.append(next(generator))
|
|
|
|
|
except StopIteration:
|
|
|
|
|
interrupted = False
|
2021-12-02 17:00:37 -05:00
|
|
|
|
except KeyboardInterrupt:
|
|
|
|
|
self.to_screen('Interrupted by user')
|
|
|
|
|
except Exception as e:
|
|
|
|
|
if self.get_param('ignoreerrors') is not True:
|
|
|
|
|
raise
|
|
|
|
|
self._downloader.report_error(e)
|
2021-10-12 05:50:50 -04:00
|
|
|
|
comment_count = len(comments)
|
|
|
|
|
self.to_screen(f'Extracted {comment_count} comments')
|
|
|
|
|
return {
|
|
|
|
|
'comments': comments,
|
|
|
|
|
'comment_count': None if interrupted else comment_count
|
|
|
|
|
}
|
|
|
|
|
return extractor
|
|
|
|
|
|
|
|
|
|
def _get_comments(self, *args, **kwargs):
|
|
|
|
|
raise NotImplementedError('This method must be implemented by subclasses')
|
|
|
|
|
|
2015-08-20 13:37:07 -04:00
|
|
|
|
@staticmethod
|
|
|
|
|
def _merge_subtitle_items(subtitle_list1, subtitle_list2):
|
2022-03-11 17:20:58 -05:00
|
|
|
|
""" Merge subtitle items for one language. Items with duplicated URLs/data
|
2015-08-20 13:37:07 -04:00
|
|
|
|
will be dropped. """
|
2022-04-11 11:10:28 -04:00
|
|
|
|
list1_data = {(item.get('url'), item.get('data')) for item in subtitle_list1}
|
2015-08-20 13:37:07 -04:00
|
|
|
|
ret = list(subtitle_list1)
|
2022-03-26 22:20:43 -04:00
|
|
|
|
ret.extend(item for item in subtitle_list2 if (item.get('url'), item.get('data')) not in list1_data)
|
2015-08-20 13:37:07 -04:00
|
|
|
|
return ret
|
|
|
|
|
|
|
|
|
|
@classmethod
|
2021-07-06 16:57:53 -04:00
|
|
|
|
def _merge_subtitles(cls, *dicts, target=None):
|
2021-04-19 13:25:54 -04:00
|
|
|
|
""" Merge subtitle dictionaries, language by language. """
|
|
|
|
|
if target is None:
|
|
|
|
|
target = {}
|
|
|
|
|
for d in dicts:
|
|
|
|
|
for lang, subs in d.items():
|
|
|
|
|
target[lang] = cls._merge_subtitle_items(target.get(lang, []), subs)
|
|
|
|
|
return target
|
2015-08-20 13:37:07 -04:00
|
|
|
|
|
2015-02-16 15:44:17 -05:00
|
|
|
|
def extract_automatic_captions(self, *args, **kwargs):
|
2021-05-17 08:23:08 -04:00
|
|
|
|
if (self.get_param('writeautomaticsub', False)
|
|
|
|
|
or self.get_param('listsubtitles')):
|
2015-02-17 16:16:29 -05:00
|
|
|
|
return self._get_automatic_captions(*args, **kwargs)
|
|
|
|
|
return {}
|
2015-02-16 15:44:17 -05:00
|
|
|
|
|
|
|
|
|
def _get_automatic_captions(self, *args, **kwargs):
|
2016-02-14 04:37:17 -05:00
|
|
|
|
raise NotImplementedError('This method must be implemented by subclasses')
|
2015-02-16 15:44:17 -05:00
|
|
|
|
|
2022-05-20 11:25:21 -04:00
|
|
|
|
@functools.cached_property
|
2022-05-16 10:06:36 -04:00
|
|
|
|
def _cookies_passed(self):
|
|
|
|
|
"""Whether cookies have been passed to YoutubeDL"""
|
|
|
|
|
return self.get_param('cookiefile') is not None or self.get_param('cookiesfrombrowser') is not None
|
|
|
|
|
|
2016-02-29 14:01:33 -05:00
|
|
|
|
def mark_watched(self, *args, **kwargs):
|
2021-09-23 16:46:03 -04:00
|
|
|
|
if not self.get_param('mark_watched', False):
|
|
|
|
|
return
|
2022-05-16 10:06:36 -04:00
|
|
|
|
if self.supports_login() and self._get_login_info()[0] is not None or self._cookies_passed:
|
2016-02-29 14:01:33 -05:00
|
|
|
|
self._mark_watched(*args, **kwargs)
|
|
|
|
|
|
|
|
|
|
def _mark_watched(self, *args, **kwargs):
|
|
|
|
|
raise NotImplementedError('This method must be implemented by subclasses')
|
|
|
|
|
|
2016-07-03 11:23:48 -04:00
|
|
|
|
def geo_verification_headers(self):
|
|
|
|
|
headers = {}
|
2021-05-17 08:23:08 -04:00
|
|
|
|
geo_verification_proxy = self.get_param('geo_verification_proxy')
|
2016-07-03 11:23:48 -04:00
|
|
|
|
if geo_verification_proxy:
|
|
|
|
|
headers['Ytdl-request-proxy'] = geo_verification_proxy
|
|
|
|
|
return headers
|
|
|
|
|
|
2022-07-31 21:22:03 -04:00
|
|
|
|
@staticmethod
|
|
|
|
|
def _generic_id(url):
|
2022-06-24 06:54:43 -04:00
|
|
|
|
return urllib.parse.unquote(os.path.splitext(url.rstrip('/').split('/')[-1])[0])
|
2016-10-07 07:20:53 -04:00
|
|
|
|
|
2022-10-31 08:05:20 -04:00
|
|
|
|
def _generic_title(self, url='', webpage='', *, default=None):
|
|
|
|
|
return (self._og_search_title(webpage, default=None)
|
|
|
|
|
or self._html_extract_title(webpage, default=None)
|
|
|
|
|
or urllib.parse.unquote(os.path.splitext(url_basename(url))[0])
|
|
|
|
|
or default)
|
2016-10-07 07:20:53 -04:00
|
|
|
|
|
2021-03-21 17:23:34 -04:00
|
|
|
|
@staticmethod
|
2021-05-30 09:59:00 -04:00
|
|
|
|
def _availability(is_private=None, needs_premium=None, needs_subscription=None, needs_auth=None, is_unlisted=None):
|
2021-03-21 17:23:34 -04:00
|
|
|
|
all_known = all(map(
|
|
|
|
|
lambda x: x is not None,
|
|
|
|
|
(is_private, needs_premium, needs_subscription, needs_auth, is_unlisted)))
|
|
|
|
|
return (
|
|
|
|
|
'private' if is_private
|
|
|
|
|
else 'premium_only' if needs_premium
|
|
|
|
|
else 'subscriber_only' if needs_subscription
|
|
|
|
|
else 'needs_auth' if needs_auth
|
|
|
|
|
else 'unlisted' if is_unlisted
|
|
|
|
|
else 'public' if all_known
|
|
|
|
|
else None)
|
|
|
|
|
|
2021-12-24 17:28:57 -05:00
|
|
|
|
def _configuration_arg(self, key, default=NO_DEFAULT, *, ie_key=None, casesense=False):
|
2021-07-08 11:33:13 -04:00
|
|
|
|
'''
|
|
|
|
|
@returns A list of values for the extractor argument given by "key"
|
|
|
|
|
or "default" if no such key is present
|
|
|
|
|
@param default The default value to return when the key is not present (default: [])
|
|
|
|
|
@param casesense When false, the values are converted to lower case
|
|
|
|
|
'''
|
2022-10-12 18:53:39 -04:00
|
|
|
|
ie_key = ie_key if isinstance(ie_key, str) else (ie_key or self).ie_key()
|
|
|
|
|
val = traverse_obj(self._downloader.params, ('extractor_args', ie_key.lower(), key))
|
2021-07-08 11:33:13 -04:00
|
|
|
|
if val is None:
|
|
|
|
|
return [] if default is NO_DEFAULT else default
|
|
|
|
|
return list(val) if casesense else [x.lower() for x in val]
|
2021-06-25 10:05:41 -04:00
|
|
|
|
|
2022-01-23 12:54:24 -05:00
|
|
|
|
def _yes_playlist(self, playlist_id, video_id, smuggled_data=None, *, playlist_label='playlist', video_label='video'):
|
|
|
|
|
if not playlist_id or not video_id:
|
|
|
|
|
return not video_id
|
|
|
|
|
|
|
|
|
|
no_playlist = (smuggled_data or {}).get('force_noplaylist')
|
|
|
|
|
if no_playlist is not None:
|
|
|
|
|
return not no_playlist
|
|
|
|
|
|
|
|
|
|
video_id = '' if video_id is True else f' {video_id}'
|
|
|
|
|
playlist_id = '' if playlist_id is True else f' {playlist_id}'
|
|
|
|
|
if self.get_param('noplaylist'):
|
|
|
|
|
self.to_screen(f'Downloading just the {video_label}{video_id} because of --no-playlist')
|
|
|
|
|
return False
|
|
|
|
|
self.to_screen(f'Downloading {playlist_label}{playlist_id} - add --no-playlist to download just the {video_label}{video_id}')
|
|
|
|
|
return True
|
|
|
|
|
|
2022-08-01 16:13:18 -04:00
|
|
|
|
def _error_or_warning(self, err, _count=None, _retries=0, *, fatal=True):
|
2022-09-21 16:23:37 -04:00
|
|
|
|
RetryManager.report_retry(
|
|
|
|
|
err, _count or int(fatal), _retries,
|
|
|
|
|
info=self.to_screen, warn=self.report_warning, error=None if fatal else self.report_warning,
|
|
|
|
|
sleep_func=self.get_param('retry_sleep_functions', {}).get('extractor'))
|
2022-08-01 16:13:18 -04:00
|
|
|
|
|
|
|
|
|
def RetryManager(self, **kwargs):
|
|
|
|
|
return RetryManager(self.get_param('extractor_retries', 3), self._error_or_warning, **kwargs)
|
|
|
|
|
|
2022-10-09 06:39:36 -04:00
|
|
|
|
def _extract_generic_embeds(self, url, *args, info_dict={}, note='Extracting generic embeds', **kwargs):
|
|
|
|
|
display_id = traverse_obj(info_dict, 'display_id', 'id')
|
|
|
|
|
self.to_screen(f'{format_field(display_id, None, "%s: ")}{note}')
|
|
|
|
|
return self._downloader.get_info_extractor('Generic')._extract_embeds(
|
|
|
|
|
smuggle_url(url, {'block_ies': [self.ie_key()]}), *args, **kwargs)
|
|
|
|
|
|
2022-07-31 21:22:03 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def extract_from_webpage(cls, ydl, url, webpage):
|
|
|
|
|
ie = (cls if isinstance(cls._extract_from_webpage, types.MethodType)
|
|
|
|
|
else ydl.get_info_extractor(cls.ie_key()))
|
2022-07-08 07:23:05 -04:00
|
|
|
|
for info in ie._extract_from_webpage(url, webpage) or []:
|
|
|
|
|
# url = None since we do not want to set (webpage/original)_url
|
|
|
|
|
ydl.add_default_extra_info(info, ie, None)
|
|
|
|
|
yield info
|
2022-07-31 21:22:03 -04:00
|
|
|
|
|
|
|
|
|
@classmethod
|
|
|
|
|
def _extract_from_webpage(cls, url, webpage):
|
|
|
|
|
for embed_url in orderedSet(
|
|
|
|
|
cls._extract_embed_urls(url, webpage) or [], lazy=True):
|
2022-09-01 07:19:03 -04:00
|
|
|
|
yield cls.url_result(embed_url, None if cls._VALID_URL is False else cls)
|
2022-07-31 21:22:03 -04:00
|
|
|
|
|
|
|
|
|
@classmethod
|
|
|
|
|
def _extract_embed_urls(cls, url, webpage):
|
|
|
|
|
"""@returns all the embed urls on the webpage"""
|
|
|
|
|
if '_EMBED_URL_RE' not in cls.__dict__:
|
|
|
|
|
assert isinstance(cls._EMBED_REGEX, (list, tuple))
|
|
|
|
|
for idx, regex in enumerate(cls._EMBED_REGEX):
|
|
|
|
|
assert regex.count('(?P<url>') == 1, \
|
|
|
|
|
f'{cls.__name__}._EMBED_REGEX[{idx}] must have exactly 1 url group\n\t{regex}'
|
|
|
|
|
cls._EMBED_URL_RE = tuple(map(re.compile, cls._EMBED_REGEX))
|
|
|
|
|
|
|
|
|
|
for regex in cls._EMBED_URL_RE:
|
|
|
|
|
for mobj in regex.finditer(webpage):
|
|
|
|
|
embed_url = urllib.parse.urljoin(url, unescapeHTML(mobj.group('url')))
|
|
|
|
|
if cls._VALID_URL is False or cls.suitable(embed_url):
|
|
|
|
|
yield embed_url
|
|
|
|
|
|
|
|
|
|
class StopExtraction(Exception):
|
|
|
|
|
pass
|
|
|
|
|
|
2022-07-31 21:23:25 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def _extract_url(cls, webpage): # TODO: Remove
|
|
|
|
|
"""Only for compatibility with some older extractors"""
|
|
|
|
|
return next(iter(cls._extract_embed_urls(None, webpage) or []), None)
|
|
|
|
|
|
2022-09-16 07:07:38 -04:00
|
|
|
|
@classmethod
|
|
|
|
|
def __init_subclass__(cls, *, plugin_name=None, **kwargs):
|
|
|
|
|
if plugin_name:
|
|
|
|
|
mro = inspect.getmro(cls)
|
|
|
|
|
super_class = cls.__wrapped__ = mro[mro.index(cls) + 1]
|
|
|
|
|
cls.IE_NAME, cls.ie_key = f'{super_class.IE_NAME}+{plugin_name}', super_class.ie_key
|
|
|
|
|
while getattr(super_class, '__wrapped__', None):
|
|
|
|
|
super_class = super_class.__wrapped__
|
|
|
|
|
setattr(sys.modules[super_class.__module__], super_class.__name__, cls)
|
|
|
|
|
|
|
|
|
|
return super().__init_subclass__(**kwargs)
|
|
|
|
|
|
2013-10-06 00:06:30 -04:00
|
|
|
|
|
2013-06-23 13:57:38 -04:00
|
|
|
|
class SearchInfoExtractor(InfoExtractor):
|
|
|
|
|
"""
|
|
|
|
|
Base class for paged search queries extractors.
|
2015-07-23 13:37:45 -04:00
|
|
|
|
They accept URLs in the format _SEARCH_KEY(|all|[0-9]):{query}
|
2021-10-23 10:29:52 -04:00
|
|
|
|
Instances should define _SEARCH_KEY and optionally _MAX_RESULTS
|
2013-06-23 13:57:38 -04:00
|
|
|
|
"""
|
|
|
|
|
|
2021-10-23 10:29:52 -04:00
|
|
|
|
_MAX_RESULTS = float('inf')
|
|
|
|
|
|
2022-07-31 21:22:03 -04:00
|
|
|
|
@classproperty
|
|
|
|
|
def _VALID_URL(cls):
|
2013-06-23 13:57:38 -04:00
|
|
|
|
return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
|
|
|
|
|
|
|
|
|
|
def _real_extract(self, query):
|
2021-12-01 12:51:19 -05:00
|
|
|
|
prefix, query = self._match_valid_url(query).group('prefix', 'query')
|
2013-06-23 13:57:38 -04:00
|
|
|
|
if prefix == '':
|
|
|
|
|
return self._get_n_results(query, 1)
|
|
|
|
|
elif prefix == 'all':
|
|
|
|
|
return self._get_n_results(query, self._MAX_RESULTS)
|
|
|
|
|
else:
|
|
|
|
|
n = int(prefix)
|
|
|
|
|
if n <= 0:
|
2022-04-11 11:10:28 -04:00
|
|
|
|
raise ExtractorError(f'invalid download number {n} for query "{query}"')
|
2013-06-23 13:57:38 -04:00
|
|
|
|
elif n > self._MAX_RESULTS:
|
2021-04-16 06:01:10 -04:00
|
|
|
|
self.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
|
2013-06-23 13:57:38 -04:00
|
|
|
|
n = self._MAX_RESULTS
|
|
|
|
|
return self._get_n_results(query, n)
|
|
|
|
|
|
|
|
|
|
def _get_n_results(self, query, n):
|
2021-10-08 16:39:55 -04:00
|
|
|
|
"""Get a specified number of results for a query.
|
|
|
|
|
Either this function or _search_results must be overridden by subclasses """
|
|
|
|
|
return self.playlist_result(
|
|
|
|
|
itertools.islice(self._search_results(query), 0, None if n == float('inf') else n),
|
|
|
|
|
query, query)
|
|
|
|
|
|
|
|
|
|
def _search_results(self, query):
|
|
|
|
|
"""Returns an iterator of search results"""
|
2016-02-14 04:37:17 -05:00
|
|
|
|
raise NotImplementedError('This method must be implemented by subclasses')
|
2013-07-01 12:52:19 -04:00
|
|
|
|
|
2022-05-11 11:54:44 -04:00
|
|
|
|
@classproperty
|
|
|
|
|
def SEARCH_KEY(cls):
|
|
|
|
|
return cls._SEARCH_KEY
|
2022-08-23 20:12:16 -04:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class UnsupportedURLIE(InfoExtractor):
|
|
|
|
|
_VALID_URL = '.*'
|
|
|
|
|
_ENABLED = False
|
|
|
|
|
IE_DESC = False
|
|
|
|
|
|
|
|
|
|
def _real_extract(self, url):
|
|
|
|
|
raise UnsupportedError(url)
|