[teamcoco] fix extraction(closes #16374)

This commit is contained in:
Remita Amine 2018-05-10 08:19:32 +01:00
parent 9e18bb4c67
commit ff8889cd4d
1 changed files with 80 additions and 95 deletions

View File

@ -1,35 +1,34 @@
# coding: utf-8
from __future__ import unicode_literals
import binascii
import re
import json
from .common import InfoExtractor
from ..compat import (
compat_b64decode,
compat_ord,
)
from ..utils import (
ExtractorError,
qualities,
determine_ext,
ExtractorError,
int_or_none,
mimetype2ext,
parse_duration,
parse_iso8601,
qualities,
)
class TeamcocoIE(InfoExtractor):
_VALID_URL = r'https?://teamcoco\.com/video/(?P<video_id>[0-9]+)?/?(?P<display_id>.*)'
_VALID_URL = r'https?://teamcoco\.com/video/(?P<id>[^/?#]+)'
_TESTS = [
{
'url': 'http://teamcoco.com/video/80187/conan-becomes-a-mary-kay-beauty-consultant',
'md5': '3f7746aa0dc86de18df7539903d399ea',
'url': 'http://teamcoco.com/video/mary-kay-remote',
'md5': '55d532f81992f5c92046ad02fec34d7d',
'info_dict': {
'id': '80187',
'ext': 'mp4',
'title': 'Conan Becomes A Mary Kay Beauty Consultant',
'description': 'Mary Kay is perhaps the most trusted name in female beauty, so of course Conan is a natural choice to sell their products.',
'duration': 504,
'age_limit': 0,
'duration': 495.0,
'upload_date': '20140402',
'timestamp': 1396407600,
}
}, {
'url': 'http://teamcoco.com/video/louis-ck-interview-george-w-bush',
@ -40,7 +39,8 @@ class TeamcocoIE(InfoExtractor):
'description': 'Louis C.K. got starstruck by George W. Bush, so what? Part one.',
'title': 'Louis C.K. Interview Pt. 1 11/3/11',
'duration': 288,
'age_limit': 0,
'upload_date': '20111104',
'timestamp': 1320405840,
}
}, {
'url': 'http://teamcoco.com/video/timothy-olyphant-drinking-whiskey',
@ -49,6 +49,8 @@ class TeamcocoIE(InfoExtractor):
'ext': 'mp4',
'title': 'Timothy Olyphant Raises A Toast To “Justified”',
'description': 'md5:15501f23f020e793aeca761205e42c24',
'upload_date': '20150415',
'timestamp': 1429088400,
},
'params': {
'skip_download': True, # m3u8 downloads
@ -63,110 +65,93 @@ class TeamcocoIE(InfoExtractor):
},
'params': {
'skip_download': True, # m3u8 downloads
}
},
'skip': 'This video is no longer available.',
}
]
_VIDEO_ID_REGEXES = (
r'"eVar42"\s*:\s*(\d+)',
r'Ginger\.TeamCoco\.openInApp\("video",\s*"([^"]+)"',
r'"id_not"\s*:\s*(\d+)'
)
def _graphql_call(self, query_template, object_type, object_id):
find_object = 'find' + object_type
return self._download_json(
'http://teamcoco.com/graphql/', object_id, data=json.dumps({
'query': query_template % (find_object, object_id)
}))['data'][find_object]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
display_id = self._match_id(url)
display_id = mobj.group('display_id')
webpage, urlh = self._download_webpage_handle(url, display_id)
if 'src=expired' in urlh.geturl():
raise ExtractorError('This video is expired.', expected=True)
response = self._graphql_call('''{
%s(slug: "video/%s") {
... on RecordSlug {
record {
id
title
teaser
publishOn
thumb {
preview
}
tags {
name
}
duration
}
}
... on NotFoundSlug {
status
}
}
}''', 'Slug', display_id)
if response.get('status'):
raise ExtractorError('This video is no longer available.', expected=True)
video_id = mobj.group('video_id')
if not video_id:
video_id = self._html_search_regex(
self._VIDEO_ID_REGEXES, webpage, 'video id')
record = response['record']
video_id = record['id']
data = None
preload_codes = self._html_search_regex(
r'(function.+)setTimeout\(function\(\)\{playlist',
webpage, 'preload codes')
base64_fragments = re.findall(r'"([a-zA-Z0-9+/=]+)"', preload_codes)
base64_fragments.remove('init')
def _check_sequence(cur_fragments):
if not cur_fragments:
return
for i in range(len(cur_fragments)):
cur_sequence = (''.join(cur_fragments[i:] + cur_fragments[:i])).encode('ascii')
try:
raw_data = compat_b64decode(cur_sequence)
if compat_ord(raw_data[0]) == compat_ord('{'):
return json.loads(raw_data.decode('utf-8'))
except (TypeError, binascii.Error, UnicodeDecodeError, ValueError):
continue
def _check_data():
for i in range(len(base64_fragments) + 1):
for j in range(i, len(base64_fragments) + 1):
data = _check_sequence(base64_fragments[:i] + base64_fragments[j:])
if data:
return data
self.to_screen('Try to compute possible data sequence. This may take some time.')
data = _check_data()
if not data:
raise ExtractorError(
'Preload information could not be extracted', expected=True)
srcs = self._graphql_call('''{
%s(id: "%s") {
src
}
}''', 'RecordVideoSource', video_id)['src']
formats = []
get_quality = qualities(['500k', '480p', '1000k', '720p', '1080p'])
for filed in data['files']:
if determine_ext(filed['url']) == 'm3u8':
# compat_urllib_parse.urljoin does not work here
if filed['url'].startswith('/'):
m3u8_url = 'http://ht.cdn.turner.com/tbs/big/teamcoco' + filed['url']
else:
m3u8_url = filed['url']
m3u8_formats = self._extract_m3u8_formats(
m3u8_url, video_id, ext='mp4')
for m3u8_format in m3u8_formats:
if m3u8_format not in formats:
formats.append(m3u8_format)
elif determine_ext(filed['url']) == 'f4m':
# TODO Correct f4m extraction
get_quality = qualities(['low', 'sd', 'hd', 'uhd'])
for format_id, src in srcs.items():
if not isinstance(src, dict):
continue
src_url = src.get('src')
if not src_url:
continue
ext = determine_ext(src_url, mimetype2ext(src.get('type')))
if format_id == 'hls' or ext == 'm3u8':
# compat_urllib_parse.urljoin does not work here
if src_url.startswith('/'):
src_url = 'http://ht.cdn.turner.com/tbs/big/teamcoco' + src_url
formats.extend(self._extract_m3u8_formats(
src_url, video_id, 'mp4', m3u8_id=format_id, fatal=False))
else:
if filed['url'].startswith('/mp4:protected/'):
if src_url.startswith('/mp4:protected/'):
# TODO Correct extraction for these files
continue
m_format = re.search(r'(\d+(k|p))\.mp4', filed['url'])
if m_format is not None:
format_id = m_format.group(1)
else:
format_id = filed['bitrate']
tbr = (
int(filed['bitrate'])
if filed['bitrate'].isdigit()
else None)
tbr = int_or_none(self._search_regex(
r'(\d+)k\.mp4', src_url, 'tbr', default=None))
formats.append({
'url': filed['url'],
'ext': 'mp4',
'url': src_url,
'ext': ext,
'tbr': tbr,
'format_id': format_id,
'quality': get_quality(format_id),
})
self._sort_formats(formats)
return {
'id': video_id,
'display_id': display_id,
'formats': formats,
'title': data['title'],
'thumbnail': data.get('thumb', {}).get('href'),
'description': data.get('teaser'),
'duration': data.get('duration'),
'age_limit': self._family_friendly_search(webpage),
'title': record['title'],
'thumbnail': record.get('thumb', {}).get('preview'),
'description': record.get('teaser'),
'duration': parse_duration(record.get('duration')),
'timestamp': parse_iso8601(record.get('publishOn')),
}