[extractor/generic:quoted-html] Add extractor (#5213)

Extracts embeds from escaped HTML within `data-html` attribute.
Related: https://github.com/ytdl-org/youtube-dl/issues/21294, https://github.com/yt-dlp/yt-dlp/pull/5121

Authored by: coletdjnz
Co-authored-by: pukkandan <pukkandan.ytdlp@gmail.com>
This commit is contained in:
Matthew 2022-10-14 17:32:52 +13:00 committed by GitHub
parent 6678a4f0b3
commit 6dca2aa66d
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 89 additions and 87 deletions

View File

@ -698,7 +698,10 @@
HSEShowIE, HSEShowIE,
HSEProductIE, HSEProductIE,
) )
from .genericembeds import HTML5MediaEmbedIE from .genericembeds import (
HTML5MediaEmbedIE,
QuotedHTMLIE,
)
from .huajiao import HuajiaoIE from .huajiao import HuajiaoIE
from .huya import HuyaLiveIE from .huya import HuyaLiveIE
from .huffpost import HuffPostIE from .huffpost import HuffPostIE
@ -1884,7 +1887,6 @@
) )
from .tv24ua import ( from .tv24ua import (
TV24UAVideoIE, TV24UAVideoIE,
TV24UAGenericPassthroughIE
) )
from .tv2dk import ( from .tv2dk import (
TV2DKIE, TV2DKIE,

View File

@ -1980,22 +1980,6 @@ class GenericIE(InfoExtractor):
}, },
'playlist_count': 6, 'playlist_count': 6,
}, },
{
# Squarespace video embed, 2019-08-28
'url': 'http://ootboxford.com',
'info_dict': {
'id': 'Tc7b_JGdZfw',
'title': 'Out of the Blue, at Childish Things 10',
'ext': 'mp4',
'description': 'md5:a83d0026666cf5ee970f8bd1cfd69c7f',
'uploader_id': 'helendouglashouse',
'uploader': 'Helen & Douglas House',
'upload_date': '20140328',
},
'params': {
'skip_download': True,
},
},
# { # {
# # Zype embed # # Zype embed
# 'url': 'https://www.cookscountry.com/episode/554-smoky-barbecue-favorites', # 'url': 'https://www.cookscountry.com/episode/554-smoky-barbecue-favorites',
@ -2784,12 +2768,6 @@ def _extract_embeds(self, url, webpage, *, urlh=None, info_dict={}):
# There probably should be a second run of generic extractor on unescaped webpage. # There probably should be a second run of generic extractor on unescaped webpage.
# webpage = urllib.parse.unquote(webpage) # webpage = urllib.parse.unquote(webpage)
# Unescape squarespace embeds to be detected by generic extractor,
# see https://github.com/ytdl-org/youtube-dl/issues/21294
webpage = re.sub(
r'<div[^>]+class=[^>]*?\bsqs-video-wrapper\b[^>]*>',
lambda x: unescapeHTML(x.group(0)), webpage)
# TODO: Move to respective extractors # TODO: Move to respective extractors
bc_urls = BrightcoveLegacyIE._extract_brightcove_urls(webpage) bc_urls = BrightcoveLegacyIE._extract_brightcove_urls(webpage)
if bc_urls: if bc_urls:

View File

@ -1,5 +1,8 @@
import re
import urllib.parse
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import make_archive_id from ..utils import make_archive_id, unescapeHTML
class HTML5MediaEmbedIE(InfoExtractor): class HTML5MediaEmbedIE(InfoExtractor):
@ -29,3 +32,84 @@ def _extract_from_webpage(self, url, webpage):
}) })
self._sort_formats(entry['formats']) self._sort_formats(entry['formats'])
yield entry yield entry
class QuotedHTMLIE(InfoExtractor):
"""For common cases of quoted/escaped html parts in the webpage"""
_VALID_URL = False
IE_NAME = 'generic:quoted-html'
IE_DESC = False # Do not list
_WEBPAGE_TESTS = [{
# 2 YouTube embeds in data-html
'url': 'https://24tv.ua/bronetransporteri-ozbroyenni-zsu-shho-vidomo-pro-bronovik-wolfhound_n2167966',
'info_dict': {
'id': 'bronetransporteri-ozbroyenni-zsu-shho-vidomo-pro-bronovik-wolfhound_n2167966',
'title': 'Броньовик Wolfhound: гігант, який допомагає ЗСУ знищувати окупантів на фронті',
'thumbnail': r're:^https?://.*\.jpe?g',
'timestamp': float,
'upload_date': str,
'description': 'md5:6816e1e5a65304bd7898e4c7eb1b26f7',
'age_limit': 0,
},
'playlist_count': 2
}, {
# Generic iframe embed of TV24UAPlayerIE within data-html
'url': 'https://24tv.ua/harkivyani-zgaduyut-misto-do-viyni-shhemlive-video_n1887584',
'info_dict': {
'id': '1887584',
'ext': 'mp4',
'title': 'Харків\'яни згадують місто до війни: щемливе відео',
'thumbnail': r're:^https?://.*\.jpe?g',
},
'params': {'skip_download': True}
}, {
# YouTube embeds on Squarespace (data-html): https://github.com/ytdl-org/youtube-dl/issues/21294
'url': 'https://www.harvardballetcompany.org/past-productions',
'info_dict': {
'id': 'past-productions',
'title': 'Productions — Harvard Ballet Company',
'age_limit': 0,
'description': 'Past Productions',
},
'playlist_mincount': 26
}, {
# Squarespace video embed, 2019-08-28, data-html
'url': 'http://ootboxford.com',
'info_dict': {
'id': 'Tc7b_JGdZfw',
'title': 'Out of the Blue, at Childish Things 10',
'ext': 'mp4',
'description': 'md5:a83d0026666cf5ee970f8bd1cfd69c7f',
'uploader_id': 'helendouglashouse',
'uploader': 'Helen & Douglas House',
'upload_date': '20140328',
'availability': 'public',
'view_count': int,
'channel': 'Helen & Douglas House',
'comment_count': int,
'uploader_url': 'http://www.youtube.com/user/helendouglashouse',
'duration': 253,
'channel_url': 'https://www.youtube.com/channel/UCTChGezrZVmlYlpMlkmulPA',
'playable_in_embed': True,
'age_limit': 0,
'channel_follower_count': int,
'channel_id': 'UCTChGezrZVmlYlpMlkmulPA',
'tags': 'count:6',
'categories': ['Nonprofits & Activism'],
'like_count': int,
'thumbnail': 'https://i.ytimg.com/vi/Tc7b_JGdZfw/hqdefault.jpg',
},
'params': {
'skip_download': True,
},
}]
def _extract_from_webpage(self, url, webpage):
combined = ''
for _, html in re.findall(r'(?s)\bdata-html=(["\'])((?:(?!\1).)+)\1', webpage):
# unescapeHTML can handle &quot; etc., unquote can handle percent encoding
unquoted_html = unescapeHTML(urllib.parse.unquote(html))
if unquoted_html != html:
combined += unquoted_html
if combined:
yield from self._extract_generic_embeds(url, combined)

View File

@ -1,15 +1,10 @@
import base64
import re import re
import urllib.parse
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import ( from ..utils import (
determine_ext, determine_ext,
extract_attributes,
get_elements_html_by_class,
js_to_json, js_to_json,
mimetype2ext, mimetype2ext,
smuggle_url,
traverse_obj, traverse_obj,
) )
@ -87,60 +82,3 @@ def _real_extract(self, url):
'title': self._html_extract_title(webpage) or self._og_search_title(webpage), 'title': self._html_extract_title(webpage) or self._og_search_title(webpage),
'description': self._og_search_description(webpage, default=None), 'description': self._og_search_description(webpage, default=None),
} }
class TV24UAGenericPassthroughIE(InfoExtractor):
_VALID_URL = r'https?://(?:[a-zA-Z0-9]+?\.)?24tv\.ua/(?P<id>[^/]+?_n\d+)'
_TESTS = [{
# Generic iframe, not within media_embed
'url': 'https://24tv.ua/vipalyuyut-nashi-mista-sela-dsns-pokazali-motoroshni-naslidki_n1883966',
'info_dict': {
'id': '1883966',
'ext': 'mp4',
'title': 'Випалюють наші міста та села, моторошні наслідки обстрілів на Чернігівщині',
'thumbnail': r're:^https?://.*\.jpe?g',
}
}, {
# Generic iframe embed of TV24UAPlayerIE, within media_embed
'url': 'https://24tv.ua/harkivyani-zgaduyut-misto-do-viyni-shhemlive-video_n1887584',
'info_dict': {
'id': 'harkivyani-zgaduyut-misto-do-viyni-shhemlive-video_n1887584',
'title': 'Харків\'яни згадують місто до війни: щемливе відео'
},
'playlist': [{
'info_dict': {
'id': '1887584',
'ext': 'mp4',
'title': 'Харків\'яни згадують місто до війни: щемливе відео',
'thumbnail': r're:^https?://.*\.jpe?g',
},
}]
}, {
# 2 media_embeds with YouTube iframes
'url': 'https://24tv.ua/bronetransporteri-ozbroyenni-zsu-shho-vidomo-pro-bronovik-wolfhound_n2167966',
'info_dict': {
'id': 'bronetransporteri-ozbroyenni-zsu-shho-vidomo-pro-bronovik-wolfhound_n2167966',
'title': 'Броньовик Wolfhound: гігант, який допомагає ЗСУ знищувати окупантів на фронті',
},
'playlist_count': 2
}, {
'url': 'https://men.24tv.ua/fitnes-bloger-sprobuvav-vikonati-trenuvannya-naysilnishoyi-lyudini_n2164538',
'only_matching': True,
}]
def _real_extract(self, url):
display_id = self._match_id(url)
webpage = self._download_webpage(url, display_id)
data_urls = []
# The site contains escaped iframe embeds within an attribute.
# Once escaped, generic can handle them, so we use a data url to pass the escaped html back.
for html in get_elements_html_by_class('media_embed', webpage):
data = urllib.parse.unquote(extract_attributes(html).get('data-html'))
data_urls.append(f'data:text/html;base64,{base64.b64encode(data.encode("utf-8")).decode("utf-8")}')
if not data_urls:
return self.url_result(url, 'Generic')
return self.playlist_from_matches(
[smuggle_url(url, {'to_generic': True}) for url in data_urls], display_id, ie='Generic',
playlist_title=self._og_search_title(webpage) or self._html_extract_title(webpage))