Skip to content

Commit

Permalink
[ie/ondemandkorea] Overhaul extractor (#8386)
Browse files Browse the repository at this point in the history 
Closes #8374
Authored by: seproDev
  • Loading branch information
@seproDev
seproDev committed Nov 11, 2023
1 parent 3ff494f commit 05adfd8
Show file tree
Hide file tree
Showing 2 changed files with 149 additions and 66 deletions.
5 changes: 4 additions & 1 deletion yt_dlp/extractor/_extractors.py
View file
Expand Up @@ -1387,7 +1387,10 @@
from .oktoberfesttv import OktoberfestTVIE
from .olympics import OlympicsReplayIE
from .on24 import On24IE
from .ondemandkorea import OnDemandKoreaIE
from .ondemandkorea import (
OnDemandKoreaIE,
OnDemandKoreaProgramIE,
)
from .onefootball import OneFootballIE
from .onenewsnz import OneNewsNZIE
from .oneplace import OnePlacePodcastIE
Expand Down
210 changes: 145 additions & 65 deletions yt_dlp/extractor/ondemandkorea.py
View file
@@ -1,87 +1,167 @@
import functools
import re
import uuid

from .common import InfoExtractor
from ..networking import HEADRequest
from ..utils import (
ExtractorError,
js_to_json,
OnDemandPagedList,
float_or_none,
int_or_none,
join_nonempty,
parse_age_limit,
parse_qs,
unified_strdate,
url_or_none,
)
from ..utils.traversal import traverse_obj


class OnDemandKoreaIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?ondemandkorea\.com/(?P<id>[^/]+)\.html'
_VALID_URL = r'https?://(?:www\.)?ondemandkorea\.com/(?:en/)?player/vod/[a-z0-9-]+\?(?:[^#]+&)?contentId=(?P<id>\d+)'
_GEO_COUNTRIES = ['US', 'CA']

_TESTS = [{
'url': 'https://www.ondemandkorea.com/ask-us-anything-e351.html',
'url': 'https://www.ondemandkorea.com/player/vod/ask-us-anything?contentId=686471',
'md5': 'e2ff77255d989e3135bde0c5889fbce8',
'info_dict': {
'id': 'ask-us-anything-e351',
'id': '686471',
'ext': 'mp4',
'title': 'Ask Us Anything : Jung Sung-ho, Park Seul-gi, Kim Bo-min, Yang Seung-won - 09/24/2022',
'description': 'A talk show/game show with a school theme where celebrity guests appear as “transfer students.”',
'thumbnail': r're:^https?://.*\.jpg$',
'title': 'Ask Us Anything: Jung Sung-ho, Park Seul-gi, Kim Bo-min, Yang Seung-won',
'thumbnail': r're:^https?://.*\.(jpg|jpeg|png)',
'duration': 5486.955,
'release_date': '20220924',
'series': 'Ask Us Anything',
'series_id': 11790,
'episode_number': 351,
'episode': 'Jung Sung-ho, Park Seul-gi, Kim Bo-min, Yang Seung-won',
},
'params': {
'skip_download': 'm3u8 download'
}
}, {
'url': 'https://www.ondemandkorea.com/work-later-drink-now-e1.html',
'url': 'https://www.ondemandkorea.com/player/vod/breakup-probation-a-week?contentId=1595796',
'md5': '57266c720006962be7ff415b24775caa',
'info_dict': {
'id': 'work-later-drink-now-e1',
'id': '1595796',
'ext': 'mp4',
'title': 'Work Later, Drink Now : E01',
'description': 'Work Later, Drink First follows three women who find solace in a glass of liquor at the end of the day. So-hee, who gets comfort from a cup of soju af',
'thumbnail': r're:^https?://.*\.png$',
'subtitles': {
'English': 'mincount:1',
},
'title': 'Breakup Probation, A Week: E08',
'thumbnail': r're:^https?://.*\.(jpg|jpeg|png)',
'duration': 1586.0,
'release_date': '20231001',
'series': 'Breakup Probation, A Week',
'series_id': 22912,
'episode_number': 8,
'episode': 'E08',
},
'params': {
'skip_download': 'm3u8 download'
}
}, {
'url': 'https://www.ondemandkorea.com/player/vod/the-outlaws?contentId=369531',
'md5': 'fa5523b87aa1f6d74fc622a97f2b47cd',
'info_dict': {
'id': '369531',
'ext': 'mp4',
'release_date': '20220519',
'duration': 7267.0,
'title': 'The Outlaws: Main Movie',
'thumbnail': r're:^https?://.*\.(jpg|jpeg|png)',
'age_limit': 18,
},
}, {
'url': 'https://www.ondemandkorea.com/en/player/vod/capture-the-moment-how-is-that-possible?contentId=1605006',
'only_matching': True,
}]

def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id, fatal=False)

if not webpage:
# Page sometimes returns captcha page with HTTP 403
raise ExtractorError(
'Unable to access page. You may have been blocked.',
expected=True)

if 'msg_block_01.png' in webpage:
self.raise_geo_restricted(
msg='This content is not available in your region',
countries=self._GEO_COUNTRIES)

if 'This video is only available to ODK PLUS members.' in webpage:
raise ExtractorError(
'This video is only available to ODK PLUS members.',
expected=True)

if 'ODK PREMIUM Members Only' in webpage:
raise ExtractorError(
'This video is only available to ODK PREMIUM members.',
expected=True)

title = self._search_regex(
r'class=["\']episode_title["\'][^>]*>([^<]+)',
webpage, 'episode_title', fatal=False) or self._og_search_title(webpage)

jw_config = self._parse_json(
self._search_regex((
r'(?P<options>{\s*[\'"]tracks[\'"].*?})[)\];]+$',
r'playlist\s*=\s*\[(?P<options>.+)];?$',
r'odkPlayer\.init.*?(?P<options>{[^;]+}).*?;',
), webpage, 'jw config', flags=re.MULTILINE | re.DOTALL, group='options'),
video_id, transform_source=js_to_json)
info = self._parse_jwplayer_data(
jw_config, video_id, require_title=False, m3u8_id='hls',
base_url=url)

info.update({
'title': title,
'description': self._og_search_description(webpage),
'thumbnail': self._og_search_thumbnail(webpage)
})
return info

data = self._download_json(
f'https://odkmedia.io/odx/api/v3/playback/{video_id}/', video_id, fatal=False,
headers={'service-name': 'odk'}, query={'did': str(uuid.uuid4())}, expected_status=(403, 404))
if not traverse_obj(data, ('result', {dict})):
msg = traverse_obj(data, ('messages', '__default'), 'title', expected_type=str)
raise ExtractorError(msg or 'Got empty response from playback API', expected=True)

data = data['result']

def try_geo_bypass(url):
return traverse_obj(url, ({parse_qs}, 'stream_url', 0, {url_or_none})) or url

def try_upgrade_quality(url):
mod_url = re.sub(r'_720(p?)\.m3u8', r'_1080\1.m3u8', url)
return mod_url if mod_url != url and self._request_webpage(
HEADRequest(mod_url), video_id, note='Checking for higher quality format',
errnote='No higher quality format found', fatal=False) else url

formats = []
for m3u8_url in traverse_obj(data, (('sources', 'manifest'), ..., 'url', {url_or_none}, {try_geo_bypass})):
formats.extend(self._extract_m3u8_formats(try_upgrade_quality(m3u8_url), video_id, fatal=False))

subtitles = {}
for track in traverse_obj(data, ('text_tracks', lambda _, v: url_or_none(v['url']))):
subtitles.setdefault(track.get('language', 'und'), []).append({
'url': track['url'],
'ext': track.get('codec'),
'name': track.get('label'),
})

def if_series(key=None):
return lambda obj: obj[key] if key and obj['kind'] == 'series' else None

return {
'id': video_id,
'title': join_nonempty(
('episode', 'program', 'title'),
('episode', 'title'), from_dict=data, delim=': '),
**traverse_obj(data, {
'thumbnail': ('episode', 'images', 'thumbnail', {url_or_none}),
'release_date': ('episode', 'release_date', {lambda x: x.replace('-', '')}, {unified_strdate}),
'duration': ('duration', {functools.partial(float_or_none, scale=1000)}),
'age_limit': ('age_rating', 'name', {lambda x: x.replace('R', '')}, {parse_age_limit}),
'series': ('episode', {if_series(key='program')}, 'title'),
'series_id': ('episode', {if_series(key='program')}, 'id'),
'episode': ('episode', {if_series(key='title')}),
'episode_number': ('episode', {if_series(key='number')}, {int_or_none}),
}, get_all=False),
'formats': formats,
'subtitles': subtitles,
}


class OnDemandKoreaProgramIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?ondemandkorea\.com/(?:en/)?player/vod/(?P<id>[a-z0-9-]+)(?:$|#)'
_GEO_COUNTRIES = ['US', 'CA']

_TESTS = [{
'url': 'https://www.ondemandkorea.com/player/vod/uskn-news',
'info_dict': {
'id': 'uskn-news',
},
'playlist_mincount': 755,
}, {
'url': 'https://www.ondemandkorea.com/en/player/vod/the-land',
'info_dict': {
'id': 'the-land',
},
'playlist_count': 52,
}]

_PAGE_SIZE = 100

def _fetch_page(self, display_id, page):
page += 1
page_data = self._download_json(
f'https://odkmedia.io/odx/api/v3/program/{display_id}/episodes/', display_id,
headers={'service-name': 'odk'}, query={
'page': page,
'page_size': self._PAGE_SIZE,
}, note=f'Downloading page {page}', expected_status=404)
for episode in traverse_obj(page_data, ('result', 'results', ...)):
yield self.url_result(
f'https://www.ondemandkorea.com/player/vod/{display_id}?contentId={episode["id"]}',
ie=OnDemandKoreaIE, video_title=episode.get('title'))

def _real_extract(self, url):
display_id = self._match_id(url)

entries = OnDemandPagedList(functools.partial(
self._fetch_page, display_id), self._PAGE_SIZE)

return self.playlist_result(entries, display_id)

0 comments on commit 05adfd8

Please sign in to comment.