[screenjunkies] Add new extractor

This doesn't handle the plus only videos yet

Closes #8492
This commit is contained in:
Patrick Griffis 2016-02-10 15:33:19 -05:00 committed by Sergey M․
parent 6eff2605d6
commit b9e8d7140a
2 changed files with 75 additions and 0 deletions

View file

@ -621,6 +621,7 @@
from .scivee import SciVeeIE from .scivee import SciVeeIE
from .screencast import ScreencastIE from .screencast import ScreencastIE
from .screencastomatic import ScreencastOMaticIE from .screencastomatic import ScreencastOMaticIE
from .screenjunkies import ScreenJunkiesIE
from .screenwavemedia import ScreenwaveMediaIE, TeamFourIE from .screenwavemedia import ScreenwaveMediaIE, TeamFourIE
from .senateisvp import SenateISVPIE from .senateisvp import SenateISVPIE
from .servingsys import ServingSysIE from .servingsys import ServingSysIE

View file

@ -0,0 +1,74 @@
from __future__ import unicode_literals
import re
import json
from .common import (
InfoExtractor,
ExtractorError,
)
from ..utils import (
int_or_none,
)
class ScreenJunkiesIE(InfoExtractor):
_VALID_URL = r'http://www.screenjunkies.com/video/(.+-(?P<id>\d+)|.+)'
_TESTS = [{
'url': 'http://www.screenjunkies.com/video/best-quentin-tarantino-movie-2841915',
'info_dict': {
'id': '2841915',
'ext': 'mp4',
'title': 'Best Quentin Tarantino Movie',
},
}, {
'url': 'http://www.screenjunkies.com/video/honest-trailers-the-dark-knight',
'info_dict': {
'id': '2348808',
'ext': 'mp4',
'title': "Honest Trailers - 'The Dark Knight'",
},
}, {
'url': 'http://www.screenjunkies.com/video/knocking-dead-ep-1-the-show-so-far-3003285',
'only_matching': True,
}]
def _real_extract(self, url):
video_id = self._match_id(url)
if not video_id: # Older urls didn't have the id in them, but we can grab it manually
webpage = self._download_webpage(url, url)
video_id = self._html_search_regex(r'src="/embed/(\d+)"', webpage, 'video id')
webpage = self._download_webpage('http://www.screenjunkies.com/embed/%s' %video_id, video_id)
video_vars_str = self._html_search_regex(r'embedVars = (\{.+\})\s*</script>', webpage, 'video variables', flags=re.S)
video_vars = self._parse_json(video_vars_str, video_id)
# TODO: Figure out required cookies
if video_vars['subscriptionLevel'] > 0:
raise ExtractorError('This video requires ScreenJunkiesPlus', expected=True)
formats = []
for f in video_vars['media']:
if not f['mediaPurpose'] == 'play':
continue
formats.append({
'url': f['uri'],
'width': int_or_none(f.get('width')),
'height': int_or_none(f.get('height')),
'tbr': int_or_none(f.get('bitRate')),
'format': 'mp4',
})
self._sort_formats(formats)
return {
'id': video_id,
'title': video_vars['contentName'],
'formats': formats,
'duration': int_or_none(video_vars.get('videoLengthInSeconds')),
'thumbnail': video_vars.get('thumbUri'),
'tags': video_vars['tags'].split(',') if 'tags' in video_vars else [],
}