X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/9dc487f48b50767cf540fa36c3de2c386fd74c04..f844a9ea177d7a03120f00d87c1c52e87c299af3:/youtube_dl/extractor/safari.py?ds=sidebyside diff --git a/youtube_dl/extractor/safari.py b/youtube_dl/extractor/safari.py index 6ba91f2..8d48067 100644 --- a/youtube_dl/extractor/safari.py +++ b/youtube_dl/extractor/safari.py @@ -1,25 +1,27 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals +import json import re from .common import InfoExtractor +from ..compat import ( + compat_parse_qs, + compat_str, + compat_urlparse, +) from ..utils import ( ExtractorError, - sanitized_Request, - std_headers, - urlencode_postdata, update_url_query, ) class SafariBaseIE(InfoExtractor): - _LOGIN_URL = 'https://www.safaribooksonline.com/accounts/login/' - _SUCCESSFUL_LOGIN_REGEX = r']*>Sign Out' + _LOGIN_URL = 'https://learning.oreilly.com/accounts/login/' _NETRC_MACHINE = 'safari' - _API_BASE = 'https://www.safaribooksonline.com/api/v1' + _API_BASE = 'https://learning.oreilly.com/api/v1' _API_FORMAT = 'json' LOGGED_IN = False @@ -28,54 +30,69 @@ class SafariBaseIE(InfoExtractor): self._login() def _login(self): - # We only need to log in once for courses or individual videos - if self.LOGGED_IN: - return - - (username, password) = self._get_login_info() + username, password = self._get_login_info() if username is None: return - headers = std_headers.copy() - if 'Referer' not in headers: - headers['Referer'] = self._LOGIN_URL - login_page_request = sanitized_Request(self._LOGIN_URL, headers=headers) - - login_page = self._download_webpage( - login_page_request, None, - 'Downloading login form') - - csrf = self._html_search_regex( - r"name='csrfmiddlewaretoken'\s+value='([^']+)'", - login_page, 'csrf token') - - login_form = { - 'csrfmiddlewaretoken': csrf, - 'email': username, - 'password1': password, - 'login': 'Sign In', - 'next': '', - } + _, urlh = self._download_webpage_handle( + 'https://learning.oreilly.com/accounts/login-check/', None, + 'Downloading login page') - request = sanitized_Request( - self._LOGIN_URL, urlencode_postdata(login_form), headers=headers) - login_page = self._download_webpage( - request, None, 'Logging in as %s' % username) + def is_logged(urlh): + return 'learning.oreilly.com/home/' in compat_str(urlh.geturl()) - if re.search(self._SUCCESSFUL_LOGIN_REGEX, login_page) is None: + if is_logged(urlh): + self.LOGGED_IN = True + return + + redirect_url = compat_str(urlh.geturl()) + parsed_url = compat_urlparse.urlparse(redirect_url) + qs = compat_parse_qs(parsed_url.query) + next_uri = compat_urlparse.urljoin( + 'https://api.oreilly.com', qs['next'][0]) + + auth, urlh = self._download_json_handle( + 'https://www.oreilly.com/member/auth/login/', None, 'Logging in', + data=json.dumps({ + 'email': username, + 'password': password, + 'redirect_uri': next_uri, + }).encode(), headers={ + 'Content-Type': 'application/json', + 'Referer': redirect_url, + }, expected_status=400) + + credentials = auth.get('credentials') + if (not auth.get('logged_in') and not auth.get('redirect_uri') + and credentials): raise ExtractorError( - 'Login failed; make sure your credentials are correct and try again.', - expected=True) + 'Unable to login: %s' % credentials, expected=True) - SafariBaseIE.LOGGED_IN = True + # oreilly serves two same groot_sessionid cookies in Set-Cookie header + # and expects first one to be actually set + self._apply_first_set_cookie_header(urlh, 'groot_sessionid') - self.to_screen('Login successful') + _, urlh = self._download_webpage_handle( + auth.get('redirect_uri') or next_uri, None, 'Completing login',) + + if is_logged(urlh): + self.LOGGED_IN = True + return + + raise ExtractorError('Unable to log in') class SafariIE(SafariBaseIE): IE_NAME = 'safari' IE_DESC = 'safaribooksonline.com online video' - _VALID_URL = r'https?://(?:www\.)?safaribooksonline\.com/library/view/[^/]+/(?P[^/]+)/(?Ppart\d+)\.html' + _VALID_URL = r'''(?x) + https?:// + (?:www\.)?(?:safaribooksonline|(?:learning\.)?oreilly)\.com/ + (?: + library/view/[^/]+/(?P[^/]+)/(?P[^/?\#&]+)\.html| + videos/[^/]+/[^/]+/(?P[^-]+-[^/?\#&]+) + ) + ''' _TESTS = [{ 'url': 'https://www.safaribooksonline.com/library/view/hadoop-fundamentals-livelessons/9780133392838/part00.html', @@ -92,22 +109,50 @@ class SafariIE(SafariBaseIE): # non-digits in course id 'url': 'https://www.safaribooksonline.com/library/view/create-a-nodejs/100000006A0210/part00.html', 'only_matching': True, + }, { + 'url': 'https://www.safaribooksonline.com/library/view/learning-path-red/9780134664057/RHCE_Introduction.html', + 'only_matching': True, + }, { + 'url': 'https://www.safaribooksonline.com/videos/python-programming-language/9780134217314/9780134217314-PYMC_13_00', + 'only_matching': True, + }, { + 'url': 'https://learning.oreilly.com/videos/hadoop-fundamentals-livelessons/9780133392838/9780133392838-00_SeriesIntro', + 'only_matching': True, + }, { + 'url': 'https://www.oreilly.com/library/view/hadoop-fundamentals-livelessons/9780133392838/00_SeriesIntro.html', + 'only_matching': True, }] + _PARTNER_ID = '1926081' + _UICONF_ID = '29375172' + def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - video_id = '%s/%s' % (mobj.group('course_id'), mobj.group('part')) - - webpage = self._download_webpage(url, video_id) - reference_id = self._search_regex( - r'data-reference-id=(["\'])(?P.+?)\1', - webpage, 'kaltura reference id', group='id') - partner_id = self._search_regex( - r'data-partner-id=(["\'])(?P.+?)\1', - webpage, 'kaltura widget id', group='id') - ui_id = self._search_regex( - r'data-ui-id=(["\'])(?P.+?)\1', - webpage, 'kaltura uiconf id', group='id') + + reference_id = mobj.group('reference_id') + if reference_id: + video_id = reference_id + partner_id = self._PARTNER_ID + ui_id = self._UICONF_ID + else: + video_id = '%s-%s' % (mobj.group('course_id'), mobj.group('part')) + + webpage, urlh = self._download_webpage_handle(url, video_id) + + mobj = re.match(self._VALID_URL, urlh.geturl()) + reference_id = mobj.group('reference_id') + if not reference_id: + reference_id = self._search_regex( + r'data-reference-id=(["\'])(?P(?:(?!\1).)+)\1', + webpage, 'kaltura reference id', group='id') + partner_id = self._search_regex( + r'data-partner-id=(["\'])(?P(?:(?!\1).)+)\1', + webpage, 'kaltura widget id', default=self._PARTNER_ID, + group='id') + ui_id = self._search_regex( + r'data-ui-id=(["\'])(?P(?:(?!\1).)+)\1', + webpage, 'kaltura uiconf id', default=self._UICONF_ID, + group='id') query = { 'wid': '_%s' % partner_id, @@ -132,12 +177,15 @@ class SafariIE(SafariBaseIE): class SafariApiIE(SafariBaseIE): IE_NAME = 'safari:api' - _VALID_URL = r'https?://(?:www\.)?safaribooksonline\.com/api/v1/book/(?P[^/]+)/chapter(?:-content)?/(?Ppart\d+)\.html' + _VALID_URL = r'https?://(?:www\.)?(?:safaribooksonline|(?:learning\.)?oreilly)\.com/api/v1/book/(?P[^/]+)/chapter(?:-content)?/(?P[^/?#&]+)\.html' - _TEST = { + _TESTS = [{ 'url': 'https://www.safaribooksonline.com/api/v1/book/9780133392838/chapter/part00.html', 'only_matching': True, - } + }, { + 'url': 'https://www.safaribooksonline.com/api/v1/book/9780134664057/chapter/RHCE_Introduction.html', + 'only_matching': True, + }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -151,7 +199,19 @@ class SafariCourseIE(SafariBaseIE): IE_NAME = 'safari:course' IE_DESC = 'safaribooksonline.com online courses' - _VALID_URL = r'https?://(?:www\.)?safaribooksonline\.com/(?:library/view/[^/]+|api/v1/book)/(?P[^/]+)/?(?:[#?]|$)' + _VALID_URL = r'''(?x) + https?:// + (?: + (?:www\.)?(?:safaribooksonline|(?:learning\.)?oreilly)\.com/ + (?: + library/view/[^/]+| + api/v1/book| + videos/[^/]+ + )| + techbus\.safaribooksonline\.com + ) + /(?P[^/]+) + ''' _TESTS = [{ 'url': 'https://www.safaribooksonline.com/library/view/hadoop-fundamentals-livelessons/9780133392838/', @@ -164,8 +224,25 @@ class SafariCourseIE(SafariBaseIE): }, { 'url': 'https://www.safaribooksonline.com/api/v1/book/9781449396459/?override_format=json', 'only_matching': True, + }, { + 'url': 'http://techbus.safaribooksonline.com/9780134426365', + 'only_matching': True, + }, { + 'url': 'https://www.safaribooksonline.com/videos/python-programming-language/9780134217314', + 'only_matching': True, + }, { + 'url': 'https://learning.oreilly.com/videos/hadoop-fundamentals-livelessons/9780133392838', + 'only_matching': True, + }, { + 'url': 'https://www.oreilly.com/library/view/hadoop-fundamentals-livelessons/9780133392838/', + 'only_matching': True, }] + @classmethod + def suitable(cls, url): + return (False if SafariIE.suitable(url) or SafariApiIE.suitable(url) + else super(SafariCourseIE, cls).suitable(url)) + def _real_extract(self, url): course_id = self._match_id(url)