Merge pull request #8246 from dstftw/initial-json-ld-metadata-support

Initial JSON-LD metadata extraction support
This commit is contained in:
Sergey M 2016-01-16 07:20:15 +06:00
commit ac25992bc7
2 changed files with 43 additions and 1 deletions

View file

@ -55,4 +55,9 @@ class AEIE(InfoExtractor):
] ]
video_url = self._search_regex(video_url_re, webpage, 'video url') video_url = self._search_regex(video_url_re, webpage, 'video url')
return self.url_result(smuggle_url(video_url, {'sig': {'key': 'crazyjava', 'secret': 's3cr3t'}})) info = self._search_json_ld(webpage, video_id, fatal=False)
info.update({
'_type': 'url_transparent',
'url': smuggle_url(video_url, {'sig': {'key': 'crazyjava', 'secret': 's3cr3t'}}),
})
return info

View file

@ -34,6 +34,7 @@ from ..utils import (
fix_xml_ampersands, fix_xml_ampersands,
float_or_none, float_or_none,
int_or_none, int_or_none,
parse_iso8601,
RegexNotFoundError, RegexNotFoundError,
sanitize_filename, sanitize_filename,
sanitized_Request, sanitized_Request,
@ -762,6 +763,42 @@ class InfoExtractor(object):
return self._html_search_meta('twitter:player', html, return self._html_search_meta('twitter:player', html,
'twitter card player') 'twitter card player')
def _search_json_ld(self, html, video_id, fatal=True):
json_ld = self._search_regex(
r'(?s)<script[^>]+type=(["\'])application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>',
html, 'JSON-LD', fatal=fatal, group='json_ld')
if not json_ld:
return {}
return self._json_ld(json_ld, video_id, fatal=fatal)
def _json_ld(self, json_ld, video_id, fatal=True):
if isinstance(json_ld, compat_str):
json_ld = self._parse_json(json_ld, video_id, fatal=fatal)
if not json_ld:
return {}
info = {}
if json_ld.get('@context') == 'http://schema.org':
item_type = json_ld.get('@type')
if item_type == 'TVEpisode':
info.update({
'episode': unescapeHTML(json_ld.get('name')),
'episode_number': int_or_none(json_ld.get('episodeNumber')),
'description': unescapeHTML(json_ld.get('description')),
})
part_of_season = json_ld.get('partOfSeason')
if isinstance(part_of_season, dict) and part_of_season.get('@type') == 'TVSeason':
info['season_number'] = int_or_none(part_of_season.get('seasonNumber'))
part_of_series = json_ld.get('partOfSeries')
if isinstance(part_of_series, dict) and part_of_series.get('@type') == 'TVSeries':
info['series'] = unescapeHTML(part_of_series.get('name'))
elif item_type == 'Article':
info.update({
'timestamp': parse_iso8601(json_ld.get('datePublished')),
'title': unescapeHTML(json_ld.get('headline')),
'description': unescapeHTML(json_ld.get('articleBody')),
})
return dict((k, v) for k, v in info.items() if v is not None)
@staticmethod @staticmethod
def _hidden_inputs(html): def _hidden_inputs(html):
html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html) html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)