mirror of
https://github.com/ytdl-org/youtube-dl
synced 2024-11-25 11:16:49 +01:00
8c25f81bee
utils is large enough without these compatibility functions. Everything that is present in newer versions of Python (i.e. with dev Python it's just an import) goes into compat.py . Everything else (i.e. youtube-dl-specific helpers) goes into utils.py .
89 lines
3.0 KiB
Python
89 lines
3.0 KiB
Python
from __future__ import unicode_literals
|
|
|
|
import re
|
|
|
|
from .common import InfoExtractor
|
|
from ..compat import (
|
|
compat_HTTPError,
|
|
compat_str,
|
|
compat_urllib_parse,
|
|
compat_urllib_parse_urlparse,
|
|
)
|
|
from ..utils import (
|
|
ExtractorError,
|
|
)
|
|
|
|
|
|
class AddAnimeIE(InfoExtractor):
|
|
|
|
_VALID_URL = r'^http://(?:\w+\.)?add-anime\.net/watch_video\.php\?(?:.*?)v=(?P<video_id>[\w_]+)(?:.*)'
|
|
_TEST = {
|
|
'url': 'http://www.add-anime.net/watch_video.php?v=24MR3YO5SAS9',
|
|
'md5': '72954ea10bc979ab5e2eb288b21425a0',
|
|
'info_dict': {
|
|
'id': '24MR3YO5SAS9',
|
|
'ext': 'mp4',
|
|
'description': 'One Piece 606',
|
|
'title': 'One Piece 606',
|
|
}
|
|
}
|
|
|
|
def _real_extract(self, url):
|
|
try:
|
|
mobj = re.match(self._VALID_URL, url)
|
|
video_id = mobj.group('video_id')
|
|
webpage = self._download_webpage(url, video_id)
|
|
except ExtractorError as ee:
|
|
if not isinstance(ee.cause, compat_HTTPError) or \
|
|
ee.cause.code != 503:
|
|
raise
|
|
|
|
redir_webpage = ee.cause.read().decode('utf-8')
|
|
action = self._search_regex(
|
|
r'<form id="challenge-form" action="([^"]+)"',
|
|
redir_webpage, 'Redirect form')
|
|
vc = self._search_regex(
|
|
r'<input type="hidden" name="jschl_vc" value="([^"]+)"/>',
|
|
redir_webpage, 'redirect vc value')
|
|
av = re.search(
|
|
r'a\.value = ([0-9]+)[+]([0-9]+)[*]([0-9]+);',
|
|
redir_webpage)
|
|
if av is None:
|
|
raise ExtractorError(u'Cannot find redirect math task')
|
|
av_res = int(av.group(1)) + int(av.group(2)) * int(av.group(3))
|
|
|
|
parsed_url = compat_urllib_parse_urlparse(url)
|
|
av_val = av_res + len(parsed_url.netloc)
|
|
confirm_url = (
|
|
parsed_url.scheme + '://' + parsed_url.netloc +
|
|
action + '?' +
|
|
compat_urllib_parse.urlencode({
|
|
'jschl_vc': vc, 'jschl_answer': compat_str(av_val)}))
|
|
self._download_webpage(
|
|
confirm_url, video_id,
|
|
note='Confirming after redirect')
|
|
webpage = self._download_webpage(url, video_id)
|
|
|
|
formats = []
|
|
for format_id in ('normal', 'hq'):
|
|
rex = r"var %s_video_file = '(.*?)';" % re.escape(format_id)
|
|
video_url = self._search_regex(rex, webpage, 'video file URLx',
|
|
fatal=False)
|
|
if not video_url:
|
|
continue
|
|
formats.append({
|
|
'format_id': format_id,
|
|
'url': video_url,
|
|
})
|
|
self._sort_formats(formats)
|
|
video_title = self._og_search_title(webpage)
|
|
video_description = self._og_search_description(webpage)
|
|
|
|
return {
|
|
'_type': 'video',
|
|
'id': video_id,
|
|
'formats': formats,
|
|
'title': video_title,
|
|
'description': video_description
|
|
}
|