[extractor] Make search_json able to parse lists

Now `contains_pattern` can be set to `\[.+\]`
pull/5124/head
pukkandan 2 years ago
parent a83333c432
commit 8b7fb8b60d
No known key found for this signature in database
GPG Key ID: 7EEE9E1E817D0A39

@ -1227,7 +1227,7 @@ class InfoExtractor:
return None
def _search_json(self, start_pattern, string, name, video_id, *, end_pattern='',
contains_pattern='(?s:.+)', fatal=True, default=NO_DEFAULT, **kwargs):
contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT, **kwargs):
"""Searches string for the JSON object specified by start_pattern"""
# NB: end_pattern is only used to reduce the size of the initial match
if default is NO_DEFAULT:
@ -1236,7 +1236,7 @@ class InfoExtractor:
fatal, has_default = False, True
json_string = self._search_regex(
rf'(?:{start_pattern})\s*(?P<json>{{\s*(?:{contains_pattern})\s*}})\s*(?:{end_pattern})',
rf'(?:{start_pattern})\s*(?P<json>{contains_pattern})\s*(?:{end_pattern})',
string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT)
if not json_string:
return default

@ -54,7 +54,7 @@ class DropboxIE(InfoExtractor):
raise ExtractorError('Password protected video, use --video-password <password>', expected=True)
info_json = self._search_json(r'InitReact\.mountComponent\(.*?,', webpage, 'mountComponent', video_id,
contains_pattern=r'.+?"preview".+?', end_pattern=r'\)')['props']
contains_pattern=r'{.+?"preview".+?}', end_pattern=r'\)')['props']
transcode_url = traverse_obj(info_json, ((None, 'preview'), 'file', 'preview', 'content', 'transcode_url'), get_all=False)
formats, subtitles = self._extract_m3u8_formats_and_subtitles(transcode_url, video_id)

@ -84,7 +84,7 @@ class FranceCultureIE(InfoExtractor):
webpage = self._download_webpage(url, display_id)
# _search_json_ld doesn't correctly handle this. See https://github.com/yt-dlp/yt-dlp/pull/3874#discussion_r891903846
video_data = self._search_json('', webpage, 'audio data', display_id, contains_pattern=r'\s*"@type"\s*:\s*"AudioObject"\s*.+')
video_data = self._search_json('', webpage, 'audio data', display_id, contains_pattern=r'{\s*"@type"\s*:\s*"AudioObject".+}')
return {
'id': video_id,

Loading…
Cancel
Save