0
0
Fork 0

[adobetv] Support embeddable videos (closes #6039)

main
Yen Chi Hsuan 2015-06-21 18:55:26 +08:00
parent 396726244a
commit 607841af64
2 changed files with 63 additions and 1 deletions

View File

@ -4,7 +4,10 @@ from .abc import ABCIE
from .abc7news import Abc7NewsIE from .abc7news import Abc7NewsIE
from .academicearth import AcademicEarthCourseIE from .academicearth import AcademicEarthCourseIE
from .addanime import AddAnimeIE from .addanime import AddAnimeIE
from .adobetv import AdobeTVIE from .adobetv import (
AdobeTVIE,
AdobeTVVideoIE,
)
from .adultswim import AdultSwimIE from .adultswim import AdultSwimIE
from .aftenposten import AftenpostenIE from .aftenposten import AftenpostenIE
from .aftonbladet import AftonbladetIE from .aftonbladet import AftonbladetIE

View File

@ -5,6 +5,8 @@ from ..utils import (
parse_duration, parse_duration,
unified_strdate, unified_strdate,
str_to_int, str_to_int,
float_or_none,
ISO639Utils,
) )
@ -69,3 +71,60 @@ class AdobeTVIE(InfoExtractor):
'view_count': view_count, 'view_count': view_count,
'formats': formats, 'formats': formats,
} }
class AdobeTVVideoIE(InfoExtractor):
_VALID_URL = r'https?://video\.tv\.adobe\.com/v/(?P<id>\d+)'
_TEST = {
'url': 'https://video.tv.adobe.com/v/2456/',
'md5': '43662b577c018ad707a63766462b1e87',
'info_dict': {
'id': '2456',
'ext': 'mp4',
'title': 'New experience with Acrobat DC',
'description': 'New experience with Acrobat DC',
'duration': 248.667,
},
}
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
player_params = self._parse_json(self._search_regex(
r'var\s+bridge\s*=\s*([^;]+);', webpage, 'player parameters'),
video_id)
formats = [{
'url': source['src'],
'width': source.get('width'),
'height': source.get('height'),
'tbr': source.get('bitrate'),
} for source in player_params['sources']]
# For both metadata and downloaded files the duration varies among
# formats. I just pick the max one
duration = max(filter(None, [
float_or_none(source.get('duration'), scale=1000)
for source in player_params['sources']]))
subtitles = {}
for translation in player_params.get('translations', []):
lang_id = translation.get('language_w3c') or ISO639Utils.long2short(translation['language_medium'])
if lang_id not in subtitles:
subtitles[lang_id] = []
subtitles[lang_id].append({
'url': translation['vttPath'],
'ext': 'vtt',
})
return {
'id': video_id,
'formats': formats,
'title': player_params['title'],
'description': self._og_search_description(webpage),
'duration': duration,
'subtitles': subtitles,
}