ytdl/youtube_dl/extractor/lynda.py

251 lines
8.8 KiB
Python
Raw Normal View History

2014-01-03 13:24:29 +01:00
from __future__ import unicode_literals
import re
import json
from .common import InfoExtractor
from ..compat import (
compat_str,
compat_urllib_parse,
compat_urllib_request,
)
from ..utils import (
ExtractorError,
clean_html,
int_or_none,
)
class LyndaBaseIE(InfoExtractor):
_LOGIN_URL = 'https://www.lynda.com/login/login.aspx'
_ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.'
_NETRC_MACHINE = 'lynda'
def _real_initialize(self):
self._login()
def _login(self):
2015-11-14 11:44:24 +01:00
username, password = self._get_login_info()
if username is None:
return
login_form = {
'username': username.encode('utf-8'),
'password': password.encode('utf-8'),
'remember': 'false',
'stayPut': 'false'
}
request = compat_urllib_request.Request(
self._LOGIN_URL, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
login_page = self._download_webpage(
request, None, 'Logging in as %s' % username)
# Not (yet) logged in
2015-08-05 16:06:48 +02:00
m = re.search(r'loginResultJson\s*=\s*\'(?P<json>[^\']+)\';', login_page)
if m is not None:
response = m.group('json')
response_json = json.loads(response)
state = response_json['state']
if state == 'notlogged':
raise ExtractorError(
'Unable to login, incorrect username and/or password',
expected=True)
# This is when we get popup:
# > You're already logged in to lynda.com on two devices.
# > If you log in here, we'll log you out of another device.
# So, we need to confirm this.
if state == 'conflicted':
confirm_form = {
'username': '',
'password': '',
'resolve': 'true',
'remember': 'false',
'stayPut': 'false',
}
request = compat_urllib_request.Request(
self._LOGIN_URL, compat_urllib_parse.urlencode(confirm_form).encode('utf-8'))
login_page = self._download_webpage(
request, None,
'Confirming log in and log out from another device')
2015-08-05 16:06:48 +02:00
if all(not re.search(p, login_page) for p in ('isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')):
if 'login error' in login_page:
mobj = re.search(
r'(?s)<h1[^>]+class="topmost">(?P<title>[^<]+)</h1>\s*<div>(?P<description>.+?)</div>',
login_page)
if mobj:
raise ExtractorError(
'lynda returned error: %s - %s'
% (mobj.group('title'), clean_html(mobj.group('description'))),
expected=True)
raise ExtractorError('Unable to log in')
def _logout(self):
username, _ = self._get_login_info()
if username is None:
return
self._download_webpage(
'http://www.lynda.com/ajax/logout.aspx', None,
'Logging out', 'Unable to log out', fatal=False)
class LyndaIE(LyndaBaseIE):
2014-01-03 13:24:29 +01:00
IE_NAME = 'lynda'
IE_DESC = 'lynda.com videos'
_VALID_URL = r'https?://www\.lynda\.com/(?:[^/]+/[^/]+/\d+|player/embed)/(?P<id>\d+)'
_NETRC_MACHINE = 'lynda'
_TIMECODE_REGEX = r'\[(?P<timecode>\d+:\d+:\d+[\.,]\d+)\]'
2015-02-27 15:56:06 +01:00
_TESTS = [{
2014-01-03 13:24:29 +01:00
'url': 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html',
'md5': 'ecfc6862da89489161fb9cd5f5a6fac1',
2014-01-07 07:14:12 +01:00
'info_dict': {
2014-03-07 16:11:01 +01:00
'id': '114408',
'ext': 'mp4',
2014-01-03 13:24:29 +01:00
'title': 'Using the exercise files',
'duration': 68
}
2015-02-27 15:56:06 +01:00
}, {
'url': 'https://www.lynda.com/player/embed/133770?tr=foo=1;bar=g;fizz=rt&fs=0',
'only_matching': True,
}]
def _real_extract(self, url):
video_id = self._match_id(url)
2015-11-06 18:24:39 +01:00
video = self._download_json(
'http://www.lynda.com/ajax/player?videoId=%s&type=video' % video_id,
video_id, 'Downloading video JSON')
2015-11-06 18:24:39 +01:00
if 'Status' in video:
raise ExtractorError(
2015-11-06 18:24:39 +01:00
'lynda returned error: %s' % video['Message'], expected=True)
2015-11-06 18:24:39 +01:00
if video.get('HasAccess') is False:
2015-08-26 17:25:53 +02:00
self.raise_login_required('Video %s is only available for members' % video_id)
2015-11-06 18:24:39 +01:00
video_id = compat_str(video.get('ID') or video_id)
duration = int_or_none(video.get('DurationInSeconds'))
title = video['Title']
formats = []
2015-11-06 18:24:39 +01:00
fmts = video.get('Formats')
if fmts:
2015-11-06 18:24:39 +01:00
formats.extend([{
'url': f['Url'],
'ext': f.get('Extension'),
'width': int_or_none(f.get('Width')),
'height': int_or_none(f.get('Height')),
'filesize': int_or_none(f.get('FileSize')),
'format_id': compat_str(f.get('Resolution')) if f.get('Resolution') else None,
} for f in fmts if f.get('Url')])
prioritized_streams = video.get('PrioritizedStreams')
if prioritized_streams:
for prioritized_stream_id, prioritized_stream in prioritized_streams.items():
2015-11-06 18:24:39 +01:00
formats.extend([{
'url': video_url,
'width': int_or_none(format_id),
'format_id': '%s-%s' % (prioritized_stream_id, format_id),
} for format_id, video_url in prioritized_stream.items()])
self._check_formats(formats, video_id)
self._sort_formats(formats)
2015-11-06 18:24:39 +01:00
subtitles = self.extract_subtitles(video_id)
return {
'id': video_id,
'title': title,
'duration': duration,
2014-01-05 17:59:33 +01:00
'subtitles': subtitles,
'formats': formats
}
def _fix_subtitles(self, subs):
srt = ''
2015-03-02 10:49:39 +01:00
seq_counter = 0
for pos in range(0, len(subs) - 1):
seq_current = subs[pos]
m_current = re.match(self._TIMECODE_REGEX, seq_current['Timecode'])
if m_current is None:
continue
seq_next = subs[pos + 1]
m_next = re.match(self._TIMECODE_REGEX, seq_next['Timecode'])
if m_next is None:
2014-01-05 17:59:33 +01:00
continue
appear_time = m_current.group('timecode')
disappear_time = m_next.group('timecode')
2015-03-02 10:49:39 +01:00
text = seq_current['Caption'].strip()
if text:
seq_counter += 1
srt += '%s\r\n%s --> %s\r\n%s\r\n\r\n' % (seq_counter, appear_time, disappear_time, text)
if srt:
return srt
2015-11-06 18:24:39 +01:00
def _get_subtitles(self, video_id):
2014-01-05 17:59:33 +01:00
url = 'http://www.lynda.com/ajax/player?videoId=%s&type=transcript' % video_id
subs = self._download_json(url, None, False)
if subs:
return {'en': [{'ext': 'srt', 'data': self._fix_subtitles(subs)}]}
else:
return {}
class LyndaCourseIE(LyndaBaseIE):
2014-01-03 13:24:29 +01:00
IE_NAME = 'lynda:course'
IE_DESC = 'lynda.com online courses'
# Course link equals to welcome/introduction video link of same course
# We will recognize it as course link
_VALID_URL = r'https?://(?:www|m)\.lynda\.com/(?P<coursepath>[^/]+/[^/]+/(?P<courseid>\d+))-\d\.html'
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
course_path = mobj.group('coursepath')
course_id = mobj.group('courseid')
2014-11-23 20:41:03 +01:00
course = self._download_json(
'http://www.lynda.com/ajax/player?courseId=%s&type=course' % course_id,
course_id, 'Downloading course JSON')
self._logout()
if course.get('Status') == 'NotFound':
raise ExtractorError(
'Course %s does not exist' % course_id, expected=True)
unaccessible_videos = 0
videos = []
# Might want to extract videos right here from video['Formats'] as it seems 'Formats' is not provided
# by single video API anymore
for chapter in course['Chapters']:
for video in chapter.get('Videos', []):
if video.get('HasAccess') is False:
unaccessible_videos += 1
continue
2015-11-06 18:27:38 +01:00
if video.get('ID'):
videos.append(video['ID'])
if unaccessible_videos > 0:
self._downloader.report_warning(
'%s videos are only available for members (or paid members) and will not be downloaded. '
% unaccessible_videos + self._ACCOUNT_CREDENTIALS_HINT)
2014-01-03 13:24:29 +01:00
entries = [
self.url_result(
'http://www.lynda.com/%s/%s-4.html' % (course_path, video_id),
'Lynda')
2014-01-03 13:24:29 +01:00
for video_id in videos]
course_title = course.get('Title')
2014-11-23 20:41:03 +01:00
return self.playlist_result(entries, course_id, course_title)