aboutsummaryrefslogtreecommitdiff
path: root/youtube_dl/extractor
diff options
context:
space:
mode:
authorSergey M․ <dstftw@gmail.com>2019-01-01 20:31:48 +0700
committerSergey M․ <dstftw@gmail.com>2019-01-01 20:31:48 +0700
commit38d15ba7f981cc82db91edcbf05ea3844082d045 (patch)
treea22de819de7341c49f4d69f9c104cb044e684a1d /youtube_dl/extractor
parent6b688b8942a0bd18b2a3835e69a903f1eeffee2a (diff)
[manyvids] Fix extraction (closes #18604, closes #18614)
Diffstat (limited to 'youtube_dl/extractor')
-rw-r--r--youtube_dl/extractor/manyvids.py62
1 files changed, 53 insertions, 9 deletions
diff --git a/youtube_dl/extractor/manyvids.py b/youtube_dl/extractor/manyvids.py
index b94b3c2ab..e8d7163e4 100644
--- a/youtube_dl/extractor/manyvids.py
+++ b/youtube_dl/extractor/manyvids.py
@@ -2,12 +2,18 @@
from __future__ import unicode_literals
from .common import InfoExtractor
-from ..utils import int_or_none
+from ..utils import (
+ determine_ext,
+ int_or_none,
+ str_to_int,
+ urlencode_postdata,
+)
class ManyVidsIE(InfoExtractor):
_VALID_URL = r'(?i)https?://(?:www\.)?manyvids\.com/video/(?P<id>\d+)'
- _TEST = {
+ _TESTS = [{
+ # preview video
'url': 'https://www.manyvids.com/Video/133957/everthing-about-me/',
'md5': '03f11bb21c52dd12a05be21a5c7dcc97',
'info_dict': {
@@ -17,7 +23,18 @@ class ManyVidsIE(InfoExtractor):
'view_count': int,
'like_count': int,
},
- }
+ }, {
+ # full video
+ 'url': 'https://www.manyvids.com/Video/935718/MY-FACE-REVEAL/',
+ 'md5': 'f3e8f7086409e9b470e2643edb96bdcc',
+ 'info_dict': {
+ 'id': '935718',
+ 'ext': 'mp4',
+ 'title': 'MY FACE REVEAL',
+ 'view_count': int,
+ 'like_count': int,
+ },
+ }]
def _real_extract(self, url):
video_id = self._match_id(url)
@@ -28,12 +45,41 @@ class ManyVidsIE(InfoExtractor):
r'data-(?:video-filepath|meta-video)\s*=s*(["\'])(?P<url>(?:(?!\1).)+)\1',
webpage, 'video URL', group='url')
- title = '%s (Preview)' % self._html_search_regex(
- r'<h2[^>]+class="m-a-0"[^>]*>([^<]+)', webpage, 'title')
+ title = self._html_search_regex(
+ (r'<span[^>]+class=["\']item-title[^>]+>([^<]+)',
+ r'<h2[^>]+class=["\']h2 m-0["\'][^>]*>([^<]+)'),
+ webpage, 'title', default=None) or self._html_search_meta(
+ 'twitter:title', webpage, 'title', fatal=True)
+
+ if any(p in webpage for p in ('preview_videos', '_preview.mp4')):
+ title += ' (Preview)'
+
+ mv_token = self._search_regex(
+ r'data-mvtoken=(["\'])(?P<value>(?:(?!\1).)+)\1', webpage,
+ 'mv token', default=None, group='value')
+
+ if mv_token:
+ # Sets some cookies
+ self._download_webpage(
+ 'https://www.manyvids.com/includes/ajax_repository/you_had_me_at_hello.php',
+ video_id, fatal=False, data=urlencode_postdata({
+ 'mvtoken': mv_token,
+ 'vid': video_id,
+ }), headers={
+ 'Referer': url,
+ 'X-Requested-With': 'XMLHttpRequest'
+ })
+
+ if determine_ext(video_url) == 'm3u8':
+ formats = self._extract_m3u8_formats(
+ video_url, video_id, 'mp4', entry_protocol='m3u8_native',
+ m3u8_id='hls')
+ else:
+ formats = [{'url': video_url}]
like_count = int_or_none(self._search_regex(
r'data-likes=["\'](\d+)', webpage, 'like count', default=None))
- view_count = int_or_none(self._html_search_regex(
+ view_count = str_to_int(self._html_search_regex(
r'(?s)<span[^>]+class="views-wrapper"[^>]*>(.+?)</span', webpage,
'view count', default=None))
@@ -42,7 +88,5 @@ class ManyVidsIE(InfoExtractor):
'title': title,
'view_count': view_count,
'like_count': like_count,
- 'formats': [{
- 'url': video_url,
- }],
+ 'formats': formats,
}