import base64 import collections import getpass import hashlib import http.client import http.cookiejar import http.cookies import inspect import itertools import json import math import netrc import os import random import re import sys import time import types import urllib.parse import urllib.request import xml.etree.ElementTree from ..compat import functools # isort: split from ..compat import compat_etree_fromstring, compat_expanduser, compat_os_name from ..cookies import LenientSimpleCookie from ..downloader.f4m import get_base_url, remove_encrypted_media from ..utils import ( IDENTITY, JSON_LD_RE, NO_DEFAULT, ExtractorError, FormatSorter, GeoRestrictedError, GeoUtils, LenientJSONDecoder, RegexNotFoundError, RetryManager, UnsupportedError, age_restricted, base_url, bug_reports_message, classproperty, clean_html, deprecation_warning, determine_ext, dict_get, encode_data_uri, error_to_compat_str, extract_attributes, filter_dict, fix_xml_ampersands, float_or_none, format_field, int_or_none, join_nonempty, js_to_json, mimetype2ext, network_exceptions, orderedSet, parse_bitrate, parse_codecs, parse_duration, parse_iso8601, parse_m3u8_attributes, parse_resolution, sanitize_filename, sanitize_url, sanitized_Request, smuggle_url, str_or_none, str_to_int, strip_or_none, traverse_obj, try_call, try_get, unescapeHTML, unified_strdate, unified_timestamp, update_Request, update_url_query, url_basename, url_or_none, urljoin, variadic, xpath_element, xpath_text, xpath_with_ns, ) class InfoExtractor: """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 passed to the YoutubeDL. The YoutubeDL processes this information possibly downloading the video to the file system, among other possible outcomes. The type field determines the type of the result. 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: id: Video identifier. 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 Additionally, it must contain either a formats entry or a url one: formats: A list of dictionaries for each format available, ordered from worst to best quality. Potential fields: * 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, 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 is parsed from a string (in case of fragmented media) for MSS - URL of the ISM manifest. * manifest_url The URL of the manifest file in case of 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. * manifest_stream_number (For internal use only) The index of the stream in the manifest file * ext Will be calculated from URL if missing * 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 ("mp4_h264_opus" or "19"). Technically optional, but strongly recommended. * 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 * aspect_ratio Aspect ratio of the video, if known Automatically calculated from width and height * resolution Textual description of width and height Automatically calculated from width and height * dynamic_range The dynamic range of the video. One of: "SDR" (None), "HDR10", "HDR10+, "HDR12", "HLG, "DV" * tbr Average bitrate of audio and video in KBit/s * abr Average audio bitrate in KBit/s * acodec Name of the audio codec in use * asr Audio sampling rate in Hertz * audio_channels Number of audio channels * vbr Average video bitrate in KBit/s * fps Frame rate * vcodec Name of the video codec in use * container Name of the container format * filesize The number of bytes, if known in advance * filesize_approx An estimate for the number of bytes * player_url SWF Player URL (used for rtmpdump). * protocol The protocol that will be used for the actual download, lower-case. One of "http", "https" or one of the protocols defined in downloader.PROTOCOL_MAP * 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 * "duration" (optional, int or float) * "filesize" (optional, int) * is_from_start Is a live format that can be downloaded from the start. Boolean * preference Order number of this format. If this field is present and not None, the formats get sorted by this field, regardless of all other values. -1 for default (order by other properties), -2 or smaller for less than default. < -1000 to hide the format (if there is another one which is strictly better) * language Language code, e.g. "de" or "en-US". * language_preference Is this in the language mentioned in the URL? 10 if it's what the URL is about, -1 for default (don't know), -10 otherwise, other values reserved for now. * 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. * 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. * http_headers A dictionary of additional HTTP headers to add to the request. * stretched_ratio If given and not 1, indicates that the 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. * has_drm The format has DRM and cannot be downloaded. Boolean * 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 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 url: Final video URL. ext: Video filename extension. format: The video format, defaults to ext (used for --get-format) player_url: SWF Player URL (used for rtmpdump). The following fields are optional: direct: True if a direct video file was given (must only be set by GenericIE) alt_title: A secondary title of the video. 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" thumbnails: A list of dictionaries, with the following entries: * "id" (optional, string) - Thumbnail format ID * "url" * "preference" (optional, int) - quality of the image * "width" (optional, int) * "height" (optional, int) * "resolution" (optional, string "{width}x{height}", deprecated) * "filesize" (optional, int) * "http_headers" (dict) - HTTP headers for the request thumbnail: Full URL to a video thumbnail image. description: Full video description. uploader: Full name of the video uploader. license: License name the video is licensed under. creator: The creator of the video. timestamp: UNIX timestamp of the moment the video was uploaded upload_date: Video upload date in UTC (YYYYMMDD). 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 release_date: The date (YYYYMMDD) when the video was released in UTC. If not explicitly set, calculated from release_timestamp modified_timestamp: UNIX timestamp of the moment the video was last modified. modified_date: The date (YYYYMMDD) when the video was last modified in UTC. If not explicitly set, calculated from modified_timestamp uploader_id: Nickname or id of the video uploader. uploader_url: Full URL to a personal webpage of the video uploader. channel: Full name of the channel the video is uploaded on. Note that channel fields may or may not repeat uploader fields. This depends on a particular extractor. channel_id: Id of the channel. channel_url: Full URL to a channel webpage. channel_follower_count: Number of followers of the channel. location: Physical location where the video was filmed. subtitles: The available subtitles as a dictionary in the format {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: * "data": The subtitles file contents * "url": A URL pointing to the subtitles file It can optionally also have: * "name": Name or description of the subtitles * "http_headers": A dictionary of additional HTTP headers to add to the request. "ext" will be calculated from URL if missing automatic_captions: Like 'subtitles'; contains automatically generated captions instead of normal subtitles duration: Length of the video in seconds, as an integer or float. view_count: How many users have watched the video on the platform. concurrent_view_count: How many users are currently watching the video on the platform. like_count: Number of positive ratings of the video dislike_count: Number of negative ratings of the video repost_count: Number of reposts of the video average_rating: Average rating give by users, the scale used depends on the webpage comment_count: Number of comments on the video 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 * "author_thumbnail" - The thumbnail of the comment author * "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. * "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 age_limit: Age restriction for the video, as an integer (years) webpage_url: The URL to the video webpage, if given to yt-dlp it should allow to get the same result again. (It will be set by YoutubeDL if it's missing) categories: A list of categories that the video falls in, for example ["Sports", "Berlin"] tags: A list of tags assigned to the video, e.g. ["sweden", "pop music"] cast: A list of the video cast is_live: True, False, or None (=unknown). Whether this video is a live stream that goes on instead of a fixed-length video. was_live: True, False, or None (=unknown). Whether this video was originally a live stream. live_status: None (=unknown), 'is_live', 'is_upcoming', 'was_live', 'not_live', or 'post_live' (was live, but VOD is not yet processed) If absent, automatically set from is_live, was_live start_time: Time in seconds where the reproduction should start, as specified in the URL. end_time: Time in seconds where the reproduction should end, as specified in the URL. 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) 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 specifying the criteria for embedability; e.g. 'whitelist' 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 _old_archive_ids: A list of old archive ids needed for backward compatibility __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. 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. 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. The following fields should only be used when the video is an episode of some series, programme or podcast: series: Title of the series or programme the video episode belongs to. series_id: Id of the series or programme the video episode belongs to, as a unicode string. season: Title of the season the video episode belongs to. 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. 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. episode_number: Number of the video episode within a season, as an integer. episode_id: Id of the video episode, as a unicode string. 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. composer: Composer of the piece 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 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 Unless mentioned otherwise, the fields should be Unicode strings. Unless mentioned otherwise, None is equivalent to absence of information. _type "playlist" indicates multiple videos. 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. Additionally, playlists can have "id", "title", and any other relevant attributes with the same semantics as videos (see above). 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" _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. 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 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. 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. Subclasses may also override suitable() if necessary, but ensure the function signature is preserved and that this function imports everything it needs (except other extractors), so that lazy_extractors works correctly. 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 when the extractor cannot reliably be matched using just the URL, e.g. invidious/peertube instances Embed-only extractors can be defined by setting _VALID_URL = False. 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. _GEO_BYPASS attribute may be set to False in order to disable geo restriction bypass mechanisms for a particular extractor. Though it won't disable explicit geo restriction bypass based on country code provided with geo_bypass_country. _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 geo restriction, of course, if the mechanism is not disabled. _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 to _GEO_COUNTRIES. The _ENABLED attribute should be set to False for IEs that are disabled by default and must be explicitly enabled. The _WORKING attribute should be set to False for broken IEs in order to warn the users and skip the tests. """ _ready = False _downloader = None _x_forwarded_for_ip = None _GEO_BYPASS = True _GEO_COUNTRIES = None _GEO_IP_BLOCKS = None _WORKING = True _ENABLED = True _NETRC_MACHINE = None IE_DESC = None SEARCH_KEY = None _VALID_URL = None _EMBED_REGEX = [] 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. ' 'See https://github.com/yt-dlp/yt-dlp/wiki/FAQ#how-do-i-pass-cookies-to-yt-dlp for how to manually pass cookies'), }[method if method is not NO_DEFAULT else 'any' if self.supports_login() else 'cookies'] def __init__(self, downloader=None): """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""" self._ready = False self._x_forwarded_for_ip = None self._printed_messages = set() self.set_downloader(downloader) @classmethod def _match_valid_url(cls, url): if cls._VALID_URL is False: return None # 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) return cls._VALID_URL_RE.match(url) @classmethod def suitable(cls, url): """Receives a URL and returns True if suitable for this IE.""" # This function must import everything it needs (except other extractors), # so that lazy_extractors works correctly return cls._match_valid_url(url) is not None @classmethod def _match_id(cls, url): return cls._match_valid_url(url).group('id') @classmethod def get_temp_id(cls, url): try: return cls._match_id(url) except (IndexError, AttributeError): return None @classmethod def working(cls): """Getter method for _WORKING.""" return cls._WORKING @classmethod def supports_login(cls): return bool(cls._NETRC_MACHINE) def initialize(self): """Initializes an instance (authentication, etc).""" self._printed_messages = set() self._initialize_geo_bypass({ 'countries': self._GEO_COUNTRIES, 'ip_blocks': self._GEO_IP_BLOCKS, }) if not self._ready: 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): self.report_warning(f'Login with password is not supported for this website. {self._login_hint("cookies")}') self._real_initialize() self._ready = True def _initialize_geo_bypass(self, geo_bypass_context): """ 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 is selected and a random IP belonging to this country is generated. This 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 during the instance initialization with _GEO_COUNTRIES and _GEO_IP_BLOCKS. You may also manually call it from extractor's code if geo bypass information is not available beforehand (e.g. obtained during 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) """ if not self._x_forwarded_for_ip: # Geo bypass mechanism is explicitly disabled by user if not self.get_param('geo_bypass', True): 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 ip_block = self.get_param('geo_bypass_ip_block', None) # 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) self.write_debug(f'Using fake IP {self._x_forwarded_for_ip} as X-Forwarded-For') 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 country = self.get_param('geo_bypass_country', None) # 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) self._downloader.write_debug( f'Using fake IP {self._x_forwarded_for_ip} ({country.upper()}) as X-Forwarded-For') def extract(self, url): """Extracts URL information and returns it in list of dicts.""" try: for _ in range(2): try: self.initialize() self.write_debug('Extracting URL: %s' % url) ie_result = self._real_extract(url) if ie_result is None: return None if self._x_forwarded_for_ip: ie_result['__x_forwarded_for_ip'] = self._x_forwarded_for_ip 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) return ie_result except GeoRestrictedError as e: if self.__maybe_fake_ip_and_retry(e.countries): continue raise except UnsupportedError: raise except ExtractorError as e: kwargs = { 'video_id': e.video_id or self.get_temp_id(url), 'ie': self.IE_NAME, 'tb': e.traceback or sys.exc_info()[2], 'expected': e.expected, 'cause': e.cause } if hasattr(e, 'countries'): kwargs['countries'] = e.countries raise type(e)(e.orig_msg, **kwargs) except http.client.IncompleteRead as e: raise ExtractorError('A network error has occurred.', cause=e, expected=True, video_id=self.get_temp_id(url)) except (KeyError, StopIteration) as e: raise ExtractorError('An extractor error has occurred.', cause=e, video_id=self.get_temp_id(url)) def __maybe_fake_ip_and_retry(self, countries): if (not self.get_param('geo_bypass_country', None) and self._GEO_BYPASS and self.get_param('geo_bypass', True) and not self._x_forwarded_for_ip and countries): country_code = random.choice(countries) self._x_forwarded_for_ip = GeoUtils.random_ipv4(country_code) if self._x_forwarded_for_ip: self.report_warning( 'Video is geo restricted. Retrying extraction with fake IP %s (%s) as X-Forwarded-For.' % (self._x_forwarded_for_ip, country_code.upper())) return True return False def set_downloader(self, downloader): """Sets a YoutubeDL instance as the downloader for this IE.""" self._downloader = downloader @property def cache(self): return self._downloader.cache @property def cookiejar(self): return self._downloader.cookiejar def _initialize_pre_login(self): """ Initialization before login. Redefine in subclasses.""" pass def _perform_login(self, username, password): """ Login with username and password. Redefine in subclasses.""" pass def _real_initialize(self): """Real initialization process. Redefine in subclasses.""" pass def _real_extract(self, url): """Real extraction process. Redefine in subclasses.""" raise NotImplementedError('This method must be implemented by subclasses') @classmethod def ie_key(cls): """A string for getting the InfoExtractor with get_info_extractor""" return cls.__name__[:-2] @classproperty def IE_NAME(cls): return cls.__name__[:-2] @staticmethod def __can_accept_status_code(err, expected_status): assert isinstance(err, urllib.error.HTTPError) if expected_status is None: return False elif callable(expected_status): return expected_status(err.code) is True else: return err.code in variadic(expected_status) def _create_request(self, url_or_request, data=None, headers=None, query=None): if isinstance(url_or_request, urllib.request.Request): return update_Request(url_or_request, data=data, headers=headers, query=query) if query: url_or_request = update_url_query(url_or_request, query) return sanitized_Request(url_or_request, data, headers or {}) def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, data=None, headers=None, query=None, expected_status=None): """ Return the response handle. See _download_webpage docstring for arguments specification. """ if not self._downloader._first_webpage_request: sleep_interval = self.get_param('sleep_interval_requests') or 0 if sleep_interval > 0: self.to_screen('Sleeping %s seconds ...' % sleep_interval) time.sleep(sleep_interval) else: self._downloader._first_webpage_request = False if note is None: self.report_download_webpage(video_id) elif note is not False: if video_id is None: self.to_screen(str(note)) else: self.to_screen(f'{video_id}: {note}') # 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: headers = (headers or {}).copy() headers.setdefault('X-Forwarded-For', self._x_forwarded_for_ip) try: return self._downloader.urlopen(self._create_request(url_or_request, data, headers, query)) except network_exceptions as err: if isinstance(err, urllib.error.HTTPError): if self.__can_accept_status_code(err, expected_status): # Retain reference to error to prevent file object from # being closed before it can be read. Works around the # effects of # introduced in Python 3.4.1. err.fp._error = err return err.fp if errnote is False: return False if errnote is None: errnote = 'Unable to download webpage' errmsg = f'{errnote}: {error_to_compat_str(err)}' if fatal: raise ExtractorError(errmsg, cause=err) else: self.report_warning(errmsg) return False 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): """ Return a tuple (page content as string, URL handle). Arguments: url_or_request -- plain text URL as a string or a urllib.request.Request object 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. """ # Strip hashes from the URL (#1038) if isinstance(url_or_request, str): url_or_request = url_or_request.partition('#')[0] urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal, data=data, headers=headers, query=query, expected_status=expected_status) if urlh is False: assert not fatal return False content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding) return (content, urlh) @staticmethod def _guess_encoding_from_content(content_type, webpage_bytes): m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type) if m: encoding = m.group(1) else: m = re.search(br']+charset=[\'"]?([^\'")]+)[ /\'">]', webpage_bytes[:1024]) if m: encoding = m.group(1).decode('ascii') elif webpage_bytes.startswith(b'\xff\xfe'): encoding = 'utf-16' else: encoding = 'utf-8' return encoding def __check_blocked(self, content): first_block = content[:512] if ('Access to this site is blocked' in content and 'Websense' in first_block): msg = 'Access to this webpage has been blocked by Websense filtering software in your network.' blocked_iframe = self._html_search_regex( r'