[xtube] Fix metadata extraction (closes #21073, closes #22455)

This commit is contained in:
Sergey M․ 2020-02-29 23:51:34 +07:00
parent 278355bae4
commit e88b450771
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D

View file

@ -47,7 +47,7 @@ class XTubeIE(InfoExtractor):
'display_id': 'A-Super-Run-Part-1-YT', 'display_id': 'A-Super-Run-Part-1-YT',
'ext': 'flv', 'ext': 'flv',
'title': 'A Super Run - Part 1 (YT)', 'title': 'A Super Run - Part 1 (YT)',
'description': 'md5:ca0d47afff4a9b2942e4b41aa970fd93', 'description': 'md5:4cc3af1aa1b0413289babc88f0d4f616',
'uploader': 'tshirtguy59', 'uploader': 'tshirtguy59',
'duration': 579, 'duration': 579,
'view_count': int, 'view_count': int,
@ -87,6 +87,20 @@ class XTubeIE(InfoExtractor):
'Cookie': 'age_verified=1; cookiesAccepted=1', 'Cookie': 'age_verified=1; cookiesAccepted=1',
}) })
title, thumbnail, duration = [None] * 3
config = self._parse_json(self._search_regex(
r'playerConf\s*=\s*({.+?})\s*,\s*\n', webpage, 'config',
default='{}'), video_id, transform_source=js_to_json, fatal=False)
if config:
config = config.get('mainRoll')
if isinstance(config, dict):
title = config.get('title')
thumbnail = config.get('poster')
duration = int_or_none(config.get('duration'))
sources = config.get('sources')
if isinstance(sources, dict):
sources = self._parse_json(self._search_regex( sources = self._parse_json(self._search_regex(
r'(["\'])?sources\1?\s*:\s*(?P<sources>{.+?}),', r'(["\'])?sources\1?\s*:\s*(?P<sources>{.+?}),',
webpage, 'sources', group='sources'), video_id, webpage, 'sources', group='sources'), video_id,
@ -102,20 +116,25 @@ class XTubeIE(InfoExtractor):
self._remove_duplicate_formats(formats) self._remove_duplicate_formats(formats)
self._sort_formats(formats) self._sort_formats(formats)
if not title:
title = self._search_regex( title = self._search_regex(
(r'<h1>\s*(?P<title>[^<]+?)\s*</h1>', r'videoTitle\s*:\s*(["\'])(?P<title>.+?)\1'), (r'<h1>\s*(?P<title>[^<]+?)\s*</h1>', r'videoTitle\s*:\s*(["\'])(?P<title>.+?)\1'),
webpage, 'title', group='title') webpage, 'title', group='title')
description = self._search_regex( description = self._og_search_description(
webpage, default=None) or self._html_search_meta(
'twitter:description', webpage, default=None) or self._search_regex(
r'</h1>\s*<p>([^<]+)', webpage, 'description', fatal=False) r'</h1>\s*<p>([^<]+)', webpage, 'description', fatal=False)
uploader = self._search_regex( uploader = self._search_regex(
(r'<input[^>]+name="contentOwnerId"[^>]+value="([^"]+)"', (r'<input[^>]+name="contentOwnerId"[^>]+value="([^"]+)"',
r'<span[^>]+class="nickname"[^>]*>([^<]+)'), r'<span[^>]+class="nickname"[^>]*>([^<]+)'),
webpage, 'uploader', fatal=False) webpage, 'uploader', fatal=False)
if not duration:
duration = parse_duration(self._search_regex( duration = parse_duration(self._search_regex(
r'<dt>Runtime:?</dt>\s*<dd>([^<]+)</dd>', r'<dt>Runtime:?</dt>\s*<dd>([^<]+)</dd>',
webpage, 'duration', fatal=False)) webpage, 'duration', fatal=False))
view_count = str_to_int(self._search_regex( view_count = str_to_int(self._search_regex(
r'<dt>Views:?</dt>\s*<dd>([\d,\.]+)</dd>', (r'["\']viewsCount["\'][^>]*>(\d+)\s+views',
r'<dt>Views:?</dt>\s*<dd>([\d,\.]+)</dd>'),
webpage, 'view count', fatal=False)) webpage, 'view count', fatal=False))
comment_count = str_to_int(self._html_search_regex( comment_count = str_to_int(self._html_search_regex(
r'>Comments? \(([\d,\.]+)\)<', r'>Comments? \(([\d,\.]+)\)<',
@ -126,6 +145,7 @@ class XTubeIE(InfoExtractor):
'display_id': display_id, 'display_id': display_id,
'title': title, 'title': title,
'description': description, 'description': description,
'thumbnail': thumbnail,
'uploader': uploader, 'uploader': uploader,
'duration': duration, 'duration': duration,
'view_count': view_count, 'view_count': view_count,