[ie/mediasite] Extract transcripts

This commit is contained in:
kclauhk 2024-10-29 22:52:24 +08:00
parent f101e5d34c
commit ed4d9a40c1

View file

@ -5,6 +5,7 @@
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import ( from ..utils import (
ExtractorError, ExtractorError,
determine_ext,
float_or_none, float_or_none,
mimetype2ext, mimetype2ext,
smuggle_url, smuggle_url,
@ -268,7 +269,29 @@ def _real_extract(self, url):
formats.extend(stream_formats) formats.extend(stream_formats)
# XXX: Presentation['Presenters'] # XXX: Presentation['Presenters']
# XXX: Presentation['Transcript'] transcripts = presentation.get('Transcripts', {})
captions, subtitles = {}, {}
for transcript in transcripts:
lang_code = traverse_obj(
transcript, (('DetailedLanguageCode', 'LanguageCode'), {str}), get_all=False)
lang_name = transcript.get('Language')
t = {
'url': transcript.get('CaptionsUrl'),
'name': lang_name,
}
if 'Auto-Generated' in lang_name:
captions.setdefault(lang_code, []).append(t)
else:
subtitles.setdefault(lang_code, []).append(t)
if transcript_url := presentation.get('TranscriptUrl'):
if determine_ext(transcript_url) != 'txt':
if len(transcripts) == 1 and captions:
captions.setdefault(lang_code, []).append({
'url': transcript_url,
'name': lang_name,
})
else:
subtitles.setdefault('und', []).append({'url': transcript_url})
return { return {
'id': resource_id, 'id': resource_id,
@ -277,6 +300,8 @@ def _real_extract(self, url):
'duration': float_or_none(presentation.get('Duration'), 1000), 'duration': float_or_none(presentation.get('Duration'), 1000),
'timestamp': float_or_none(presentation.get('UnixTime'), 1000), 'timestamp': float_or_none(presentation.get('UnixTime'), 1000),
'formats': formats, 'formats': formats,
'automatic_captions': captions,
'subtitles': subtitles,
'thumbnails': thumbnails, 'thumbnails': thumbnails,
} }