1
0
mirror of https://github.com/l1ving/youtube-dl synced 2024-12-22 12:52:58 +08:00
youtube-dl/youtube_dl/extractor/eporner.py
2014-08-31 23:56:14 +02:00

56 lines
1.8 KiB
Python

# coding: utf-8
from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..utils import (
parse_duration,
str_to_int,
)
class EpornerIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?eporner\.com/hd-porn/(?P<id>\d+)/(?P<title_dash>[\w-]+)/?'
_TEST = {
'url': 'http://www.eporner.com/hd-porn/95008/Infamous-Tiffany-Teen-Strip-Tease-Video/',
'md5': '3b427ae4b9d60619106de3185c2987cd',
'info_dict': {
'id': '95008',
'ext': 'flv',
'title': 'Infamous Tiffany Teen Strip Tease Video',
'duration': 194,
'view_count': int,
}
}
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
video_id = mobj.group('id')
webpage = self._download_webpage(url, video_id)
title = self._html_search_regex(
r'<title>(.*?) - EPORNER', webpage, 'title')
redirect_code = self._html_search_regex(
r'<script type="text/javascript" src="/config5/%s/([a-f\d]+)/">' % video_id,
webpage, 'redirect_code')
redirect_url = 'http://www.eporner.com/config5/%s/%s' % (video_id, redirect_code)
webpage2 = self._download_webpage(redirect_url, video_id)
video_url = self._html_search_regex(
r'file: "(.*?)",', webpage2, 'video_url')
duration = parse_duration(self._search_regex(
r'class="mbtim">([0-9:]+)</div>', webpage, 'duration',
fatal=False))
view_count = str_to_int(self._search_regex(
r'id="cinemaviews">\s*([0-9,]+)\s*<small>views',
webpage, 'view count', fatal=False))
return {
'id': video_id,
'url': video_url,
'title': title,
'duration': duration,
'view_count': view_count,
}