From 1a34a802f44a1dab8f642c79c3cc810e21541d3b Mon Sep 17 00:00:00 2001 From: bashonly <88596187+bashonly@users.noreply.github.com> Date: Tue, 23 Jul 2024 18:08:24 -0500 Subject: [PATCH] [ie/facebook] Fix extraction (#10531) Closes #10532 Authored by: bashonly --- yt_dlp/extractor/facebook.py | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-) diff --git a/yt_dlp/extractor/facebook.py b/yt_dlp/extractor/facebook.py index a3ca291fc..6aba477a6 100644 --- a/yt_dlp/extractor/facebook.py +++ b/yt_dlp/extractor/facebook.py @@ -571,16 +571,21 @@ def process_formats(info): # Formats larger than ~500MB will return error 403 unless chunk size is regulated f.setdefault('downloader_options', {})['http_chunk_size'] = 250 << 20 - def extract_relay_data(_filter): - return self._parse_json(self._search_regex( - rf'data-sjs>({{.*?{_filter}.*?}})', - webpage, 'replay data', default='{}'), video_id, fatal=False) or {} + def yield_all_relay_data(_filter): + for relay_data in re.findall(rf'data-sjs>({{.*?{_filter}.*?}})', webpage): + yield self._parse_json(relay_data, video_id, fatal=False) or {} - def extract_relay_prefetched_data(_filter): - return traverse_obj(extract_relay_data(_filter), ( - 'require', (None, (..., ..., ..., '__bbox', 'require')), + def extract_relay_data(_filter): + return next(filter(None, yield_all_relay_data(_filter)), {}) + + def extract_relay_prefetched_data(_filter, target_keys=None): + path = 'data' + if target_keys is not None: + path = lambda k, v: k == 'data' and any(target in v for target in variadic(target_keys)) + return traverse_obj(yield_all_relay_data(_filter), ( + ..., 'require', (None, (..., ..., ..., '__bbox', 'require')), lambda _, v: any(key.startswith('RelayPrefetchedStreamCache') for key in v), - ..., ..., '__bbox', 'result', 'data', {dict}), get_all=False) or {} + ..., ..., '__bbox', 'result', path, {dict}), get_all=False) or {} if not video_data: server_js_data = self._parse_json(self._search_regex([ @@ -591,7 +596,8 @@ def extract_relay_prefetched_data(_filter): if not video_data: data = extract_relay_prefetched_data( - r'"(?:dash_manifest|playable_url(?:_quality_hd)?)') + r'"(?:dash_manifest|playable_url(?:_quality_hd)?)', + target_keys=('video', 'event', 'nodes', 'node', 'mediaset')) if data: entries = []