[pornhub] Extract categories and tags (closes #10499)

This commit is contained in:
Yen Chi Hsuan 2016-09-11 19:22:51 +08:00
parent fea74acad8
commit 6bb05b32a9
No known key found for this signature in database
GPG Key ID: 3FDDD575826C5C30
2 changed files with 17 additions and 0 deletions

View File

@ -1,6 +1,7 @@
version <unreleased>
Extractors
+ [pornhub] Extract categories and tags (#10499)
+ [foxnews] Support Fox News articles (#10598)
* [iwara] Fix extraction after relaunch (#10462, #3215)
* [newgrounds] Fix uploader extraction (#10584)

View File

@ -15,6 +15,7 @@ from ..compat import (
from ..utils import (
ExtractorError,
int_or_none,
js_to_json,
orderedSet,
sanitized_Request,
str_to_int,
@ -48,6 +49,8 @@ class PornHubIE(InfoExtractor):
'dislike_count': int,
'comment_count': int,
'age_limit': 18,
'tags': list,
'categories': list,
},
}, {
# non-ASCII title
@ -63,6 +66,8 @@ class PornHubIE(InfoExtractor):
'dislike_count': int,
'comment_count': int,
'age_limit': 18,
'tags': list,
'categories': list,
},
'params': {
'skip_download': True,
@ -183,6 +188,15 @@ class PornHubIE(InfoExtractor):
})
self._sort_formats(formats)
page_params = self._parse_json(self._search_regex(
r'page_params\.zoneDetails\[([\'"])[^\'"]+\1\]\s*=\s*(?P<data>{[^}]+})',
webpage, 'page parameters', group='data', default='{}'),
video_id, transform_source=js_to_json, fatal=False)
tags = categories = None
if page_params:
tags = page_params.get('tags', '').split(',')
categories = page_params.get('categories', '').split(',')
return {
'id': video_id,
'uploader': video_uploader,
@ -195,6 +209,8 @@ class PornHubIE(InfoExtractor):
'comment_count': comment_count,
'formats': formats,
'age_limit': 18,
'tags': tags,
'categories': categories,
}