diff options
Diffstat (limited to 'yt_dlp/extractor/mofosex.py')
-rw-r--r-- | yt_dlp/extractor/mofosex.py | 70 |
1 files changed, 70 insertions, 0 deletions
diff --git a/yt_dlp/extractor/mofosex.py b/yt_dlp/extractor/mofosex.py new file mode 100644 index 0000000..9cb6980 --- /dev/null +++ b/yt_dlp/extractor/mofosex.py @@ -0,0 +1,70 @@ +from .common import InfoExtractor +from ..utils import ( + int_or_none, + str_to_int, + unified_strdate, +) +from .keezmovies import KeezMoviesIE + + +class MofosexIE(KeezMoviesIE): # XXX: Do not subclass from concrete IE + _VALID_URL = r'https?://(?:www\.)?mofosex\.com/videos/(?P<id>\d+)/(?P<display_id>[^/?#&.]+)\.html' + _TESTS = [{ + 'url': 'http://www.mofosex.com/videos/318131/amateur-teen-playing-and-masturbating-318131.html', + 'md5': '558fcdafbb63a87c019218d6e49daf8a', + 'info_dict': { + 'id': '318131', + 'display_id': 'amateur-teen-playing-and-masturbating-318131', + 'ext': 'mp4', + 'title': 'amateur teen playing and masturbating', + 'thumbnail': r're:^https?://.*\.jpg$', + 'upload_date': '20121114', + 'view_count': int, + 'like_count': int, + 'dislike_count': int, + 'age_limit': 18, + } + }, { + # This video is no longer available + 'url': 'http://www.mofosex.com/videos/5018/japanese-teen-music-video.html', + 'only_matching': True, + }] + + def _real_extract(self, url): + webpage, info = self._extract_info(url) + + view_count = str_to_int(self._search_regex( + r'VIEWS:</span>\s*([\d,.]+)', webpage, 'view count', fatal=False)) + like_count = int_or_none(self._search_regex( + r'id=["\']amountLikes["\'][^>]*>(\d+)', webpage, + 'like count', fatal=False)) + dislike_count = int_or_none(self._search_regex( + r'id=["\']amountDislikes["\'][^>]*>(\d+)', webpage, + 'like count', fatal=False)) + upload_date = unified_strdate(self._html_search_regex( + r'Added:</span>([^<]+)', webpage, 'upload date', fatal=False)) + + info.update({ + 'view_count': view_count, + 'like_count': like_count, + 'dislike_count': dislike_count, + 'upload_date': upload_date, + 'thumbnail': self._og_search_thumbnail(webpage), + }) + + return info + + +class MofosexEmbedIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)?mofosex\.com/embed/?\?.*?\bvideoid=(?P<id>\d+)' + _EMBED_REGEX = [r'<iframe[^>]+\bsrc=["\'](?P<url>(?:https?:)?//(?:www\.)?mofosex\.com/embed/?\?.*?\bvideoid=\d+)'] + _TESTS = [{ + 'url': 'https://www.mofosex.com/embed/?videoid=318131&referrer=KM', + 'only_matching': True, + }] + + def _real_extract(self, url): + video_id = self._match_id(url) + return self.url_result( + 'http://www.mofosex.com/videos/{0}/{0}.html'.format(video_id), + ie=MofosexIE.ie_key(), video_id=video_id) |