[InfoExtractor] Add _search_json()

* uses the error diagnostic to truncate the JSON string
* may be confused by non-C-Pythons
This commit is contained in:
dirkf 2024-02-21 00:03:17 +00:00
parent acc383b9e3
commit 7216fa2ac4

View File

@ -25,6 +25,7 @@ from ..compat import (
compat_getpass, compat_getpass,
compat_integer_types, compat_integer_types,
compat_http_client, compat_http_client,
compat_kwargs,
compat_map as map, compat_map as map,
compat_open as open, compat_open as open,
compat_os_name, compat_os_name,
@ -1102,6 +1103,60 @@ class InfoExtractor(object):
self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
return None return None
def _search_json(self, start_pattern, string, name, video_id, **kwargs):
"""Searches string for the JSON object specified by start_pattern"""
# self, start_pattern, string, name, video_id, *, end_pattern='',
# contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT
# NB: end_pattern is only used to reduce the size of the initial match
end_pattern = kwargs.pop('end_pattern', '')
# (?:[\s\S]) simulates (?(s):.) (eg)
contains_pattern = kwargs.pop('contains_pattern', r'{[\s\S]+}')
fatal = kwargs.pop('fatal', True)
default = kwargs.pop('default', NO_DEFAULT)
if default is NO_DEFAULT:
default, has_default = {}, False
else:
fatal, has_default = False, True
json_string = self._search_regex(
r'(?:{0})\s*(?P<json>{1})\s*(?:{2})'.format(
start_pattern, contains_pattern, end_pattern),
string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT)
if not json_string:
return default
# yt-dlp has a special JSON parser that allows trailing text.
# Until that arrives here, the diagnostic from the exception
# raised by json.loads() is used to extract the wanted text.
# Either way, it's a problem if a transform_source() can't
# handle the trailing text.
# force an exception
kwargs['fatal'] = True
# self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
for _ in range(2):
try:
# return self._parse_json(json_string, video_id, ignore_extra=True, **kwargs)
transform_source = kwargs.pop('transform_source', None)
if transform_source:
json_string = transform_source(json_string)
return self._parse_json(json_string, video_id, **compat_kwargs(kwargs))
except ExtractorError as e:
end = int_or_none(self._search_regex(r'\(char\s+(\d+)', error_to_compat_str(e), 'end', default=None))
if end is not None:
json_string = json_string[:end]
continue
msg = 'Unable to extract {0} - Failed to parse JSON'.format(name)
if fatal:
raise ExtractorError(msg, cause=e.cause, video_id=video_id)
elif not has_default:
self.report_warning(
'{0}: {1}'.format(msg, error_to_compat_str(e)), video_id=video_id)
return default
def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
""" """
Like _search_regex, but strips HTML tags and unescapes entities. Like _search_regex, but strips HTML tags and unescapes entities.