|
|
- from __future__ import unicode_literals
-
- import json
- import re
-
- from .common import InfoExtractor
-
-
- class NewgroundsIE(InfoExtractor):
- _VALID_URL = r'https?://(?:www\.)?newgrounds\.com/(?:audio/listen|portal/view)/(?P<id>[0-9]+)'
- _TESTS = [{
- 'url': 'http://www.newgrounds.com/audio/listen/549479',
- 'md5': 'fe6033d297591288fa1c1f780386f07a',
- 'info_dict': {
- 'id': '549479',
- 'ext': 'mp3',
- 'title': 'B7 - BusMode',
- 'uploader': 'Burn7',
- }
- }, {
- 'url': 'http://www.newgrounds.com/portal/view/673111',
- 'md5': '3394735822aab2478c31b1004fe5e5bc',
- 'info_dict': {
- 'id': '673111',
- 'ext': 'mp4',
- 'title': 'Dancin',
- 'uploader': 'Squirrelman82',
- },
- }]
-
- def _real_extract(self, url):
- mobj = re.match(self._VALID_URL, url)
- music_id = mobj.group('id')
- webpage = self._download_webpage(url, music_id)
-
- title = self._html_search_regex(
- r'<title>([^>]+)</title>', webpage, 'title')
-
- uploader = self._html_search_regex(
- [r',"artist":"([^"]+)",', r'[\'"]owner[\'"]\s*:\s*[\'"]([^\'"]+)[\'"],'],
- webpage, 'uploader')
-
- music_url_json_string = self._html_search_regex(
- r'({"url":"[^"]+"),', webpage, 'music url') + '}'
- music_url_json = json.loads(music_url_json_string)
- music_url = music_url_json['url']
-
- return {
- 'id': music_id,
- 'title': title,
- 'url': music_url,
- 'uploader': uploader,
- }
|