[extractor/youtube] Improve chapter parsing from description

Closes #5448
This commit is contained in:
pukkandan 2022-11-05 15:34:53 +05:30
parent 68a9a450d4
commit 2e30b46fe4
No known key found for this signature in database
GPG key ID: 7EEE9E1E817D0A39

View file

@ -3027,9 +3027,14 @@ def _extract_chapters_from_engagement_panel(self, data, duration):
for contents in content_list)), []) for contents in content_list)), [])
def _extract_chapters_from_description(self, description, duration): def _extract_chapters_from_description(self, description, duration):
duration_re = r'(?:\d+:)?\d{1,2}:\d{2}'
sep_re = r'(?m)^\s*(%s)\b\W*\s(%s)\s*$'
return self._extract_chapters( return self._extract_chapters(
re.findall(r'(?m)^((?:\d+:)?\d{1,2}:\d{2})\b\W*\s(.+?)\s*$', description or ''), re.findall(sep_re % (duration_re, r'.+?'), description or ''),
chapter_time=lambda x: parse_duration(x[0]), chapter_title=lambda x: x[1], chapter_time=lambda x: parse_duration(x[0]), chapter_title=lambda x: x[1],
duration=duration, strict=False) or self._extract_chapters(
re.findall(sep_re % (r'.+?', duration_re), description or ''),
chapter_time=lambda x: parse_duration(x[1]), chapter_title=lambda x: x[0],
duration=duration, strict=False) duration=duration, strict=False)
def _extract_chapters(self, chapter_list, chapter_time, chapter_title, duration, strict=True): def _extract_chapters(self, chapter_list, chapter_time, chapter_title, duration, strict=True):