X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/4eb7d8923b3bad26edab01df0fd2650b0563ea8b..39393b81acfaf4045fb7f20454a0226f0dc9142e:/youtube_dl/extractor/canvas.py

diff --git a/youtube_dl/extractor/canvas.py b/youtube_dl/extractor/canvas.py
index 6899f84..8667a0d 100644
--- a/youtube_dl/extractor/canvas.py
+++ b/youtube_dl/extractor/canvas.py
@@ -1,53 +1,88 @@
 from __future__ import unicode_literals
 
 import re
+import json
 
 from .common import InfoExtractor
+from .gigya import GigyaBaseIE
+from ..compat import compat_HTTPError
 from ..utils import (
-    float_or_none,
+    ExtractorError,
     strip_or_none,
+    float_or_none,
+    int_or_none,
+    merge_dicts,
+    parse_iso8601,
+    str_or_none,
+    url_or_none,
 )
 
 
 class CanvasIE(InfoExtractor):
-    _VALID_URL = r'https?://mediazone\.vrt\.be/api/v1/(?P<site_id>canvas|een|ketnet)/assets/(?P<id>m[dz]-ast-[^/?#&]+)'
+    _VALID_URL = r'https?://mediazone\.vrt\.be/api/v1/(?P<site_id>canvas|een|ketnet|vrt(?:video|nieuws)|sporza)/assets/(?P<id>[^/?#&]+)'
     _TESTS = [{
         'url': 'https://mediazone.vrt.be/api/v1/ketnet/assets/md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
-        'md5': '90139b746a0a9bd7bb631283f6e2a64e',
+        'md5': '68993eda72ef62386a15ea2cf3c93107',
         'info_dict': {
             'id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
             'display_id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Nachtwacht: De Greystook',
-            'description': 'md5:1db3f5dc4c7109c821261e7512975be7',
+            'description': 'Nachtwacht: De Greystook',
             'thumbnail': r're:^https?://.*\.jpg$',
-            'duration': 1468.03,
+            'duration': 1468.04,
         },
         'expected_warnings': ['is not a supported codec', 'Unknown MIME type'],
     }, {
         'url': 'https://mediazone.vrt.be/api/v1/canvas/assets/mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
         'only_matching': True,
     }]
+    _HLS_ENTRY_PROTOCOLS_MAP = {
+        'HLS': 'm3u8_native',
+        'HLS_AES': 'm3u8',
+    }
+    _REST_API_BASE = 'https://media-services-public.vrt.be/vualto-video-aggregator-web/rest/external/v1'
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         site_id, video_id = mobj.group('site_id'), mobj.group('id')
 
+        # Old API endpoint, serves more formats but may fail for some videos
         data = self._download_json(
             'https://mediazone.vrt.be/api/v1/%s/assets/%s'
-            % (site_id, video_id), video_id)
+            % (site_id, video_id), video_id, 'Downloading asset JSON',
+            'Unable to download asset JSON', fatal=False)
+
+        # New API endpoint
+        if not data:
+            token = self._download_json(
+                '%s/tokens' % self._REST_API_BASE, video_id,
+                'Downloading token', data=b'',
+                headers={'Content-Type': 'application/json'})['vrtPlayerToken']
+            data = self._download_json(
+                '%s/videos/%s' % (self._REST_API_BASE, video_id),
+                video_id, 'Downloading video JSON', fatal=False, query={
+                    'vrtPlayerToken': token,
+                    'client': '%s@PROD' % site_id,
+                }, expected_status=400)
+            message = data.get('message')
+            if message and not data.get('title'):
+                if data.get('code') == 'AUTHENTICATION_REQUIRED':
+                    self.raise_login_required(message)
+                raise ExtractorError(message, expected=True)
 
         title = data['title']
         description = data.get('description')
 
         formats = []
         for target in data['targetUrls']:
-            format_url, format_type = target.get('url'), target.get('type')
+            format_url, format_type = url_or_none(target.get('url')), str_or_none(target.get('type'))
             if not format_url or not format_type:
                 continue
-            if format_type == 'HLS':
+            format_type = format_type.upper()
+            if format_type in self._HLS_ENTRY_PROTOCOLS_MAP:
                 formats.extend(self._extract_m3u8_formats(
-                    format_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    format_url, video_id, 'mp4', self._HLS_ENTRY_PROTOCOLS_MAP[format_type],
                     m3u8_id=format_type, fatal=False))
             elif format_type == 'HDS':
                 formats.extend(self._extract_f4m_formats(
@@ -123,20 +158,20 @@ class CanvasEenIE(InfoExtractor):
         },
         'skip': 'Pagina niet gevonden',
     }, {
-        'url': 'https://www.een.be/sorry-voor-alles/herbekijk-sorry-voor-alles',
+        'url': 'https://www.een.be/thuis/emma-pakt-thilly-aan',
         'info_dict': {
-            'id': 'mz-ast-11a587f8-b921-4266-82e2-0bce3e80d07f',
-            'display_id': 'herbekijk-sorry-voor-alles',
+            'id': 'md-ast-3a24ced2-64d7-44fb-b4ed-ed1aafbf90b8',
+            'display_id': 'emma-pakt-thilly-aan',
             'ext': 'mp4',
-            'title': 'Herbekijk Sorry voor alles',
-            'description': 'md5:8bb2805df8164e5eb95d6a7a29dc0dd3',
+            'title': 'Emma pakt Thilly aan',
+            'description': 'md5:c5c9b572388a99b2690030afa3f3bad7',
             'thumbnail': r're:^https?://.*\.jpg$',
-            'duration': 3788.06,
+            'duration': 118.24,
         },
         'params': {
             'skip_download': True,
         },
-        'skip': 'Episode no longer available',
+        'expected_warnings': ['is not a supported codec'],
     }, {
         'url': 'https://www.canvas.be/check-point/najaar-2016/de-politie-uw-vriend',
         'only_matching': True,
@@ -166,3 +201,168 @@ class CanvasEenIE(InfoExtractor):
             'title': title,
             'description': self._og_search_description(webpage),
         }
+
+
+class VrtNUIE(GigyaBaseIE):
+    IE_DESC = 'VrtNU.be'
+    _VALID_URL = r'https?://(?:www\.)?vrt\.be/(?P<site_id>vrtnu)/(?:[^/]+/)*(?P<id>[^/?#&]+)'
+    _TESTS = [{
+        # Available via old API endpoint
+        'url': 'https://www.vrt.be/vrtnu/a-z/postbus-x/1/postbus-x-s1a1/',
+        'info_dict': {
+            'id': 'pbs-pub-2e2d8c27-df26-45c9-9dc6-90c78153044d$vid-90c932b1-e21d-4fb8-99b1-db7b49cf74de',
+            'ext': 'mp4',
+            'title': 'De zwarte weduwe',
+            'description': 'md5:db1227b0f318c849ba5eab1fef895ee4',
+            'duration': 1457.04,
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'season': 'Season 1',
+            'season_number': 1,
+            'episode_number': 1,
+        },
+        'skip': 'This video is only available for registered users',
+        'params': {
+            'username': '<snip>',
+            'password': '<snip>',
+        },
+        'expected_warnings': ['is not a supported codec'],
+    }, {
+        # Only available via new API endpoint
+        'url': 'https://www.vrt.be/vrtnu/a-z/kamp-waes/1/kamp-waes-s1a5/',
+        'info_dict': {
+            'id': 'pbs-pub-0763b56c-64fb-4d38-b95b-af60bf433c71$vid-ad36a73c-4735-4f1f-b2c0-a38e6e6aa7e1',
+            'ext': 'mp4',
+            'title': 'Aflevering 5',
+            'description': 'Wie valt door de mand tijdens een missie?',
+            'duration': 2967.06,
+            'season': 'Season 1',
+            'season_number': 1,
+            'episode_number': 5,
+        },
+        'skip': 'This video is only available for registered users',
+        'params': {
+            'username': '<snip>',
+            'password': '<snip>',
+        },
+        'expected_warnings': ['Unable to download asset JSON', 'is not a supported codec', 'Unknown MIME type'],
+    }]
+    _NETRC_MACHINE = 'vrtnu'
+    _APIKEY = '3_0Z2HujMtiWq_pkAjgnS2Md2E11a1AwZjYiBETtwNE-EoEHDINgtnvcAOpNgmrVGy'
+    _CONTEXT_ID = 'R3595707040'
+
+    def _real_initialize(self):
+        self._login()
+
+    def _login(self):
+        username, password = self._get_login_info()
+        if username is None:
+            return
+
+        auth_data = {
+            'APIKey': self._APIKEY,
+            'targetEnv': 'jssdk',
+            'loginID': username,
+            'password': password,
+            'authMode': 'cookie',
+        }
+
+        auth_info = self._gigya_login(auth_data)
+
+        # Sometimes authentication fails for no good reason, retry
+        login_attempt = 1
+        while login_attempt <= 3:
+            try:
+                # When requesting a token, no actual token is returned, but the
+                # necessary cookies are set.
+                self._request_webpage(
+                    'https://token.vrt.be',
+                    None, note='Requesting a token', errnote='Could not get a token',
+                    headers={
+                        'Content-Type': 'application/json',
+                        'Referer': 'https://www.vrt.be/vrtnu/',
+                    },
+                    data=json.dumps({
+                        'uid': auth_info['UID'],
+                        'uidsig': auth_info['UIDSignature'],
+                        'ts': auth_info['signatureTimestamp'],
+                        'email': auth_info['profile']['email'],
+                    }).encode('utf-8'))
+            except ExtractorError as e:
+                if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
+                    login_attempt += 1
+                    self.report_warning('Authentication failed')
+                    self._sleep(1, None, msg_template='Waiting for %(timeout)s seconds before trying again')
+                else:
+                    raise e
+            else:
+                break
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+
+        webpage, urlh = self._download_webpage_handle(url, display_id)
+
+        info = self._search_json_ld(webpage, display_id, default={})
+
+        # title is optional here since it may be extracted by extractor
+        # that is delegated from here
+        title = strip_or_none(self._html_search_regex(
+            r'(?ms)<h1 class="content__heading">(.+?)</h1>',
+            webpage, 'title', default=None))
+
+        description = self._html_search_regex(
+            r'(?ms)<div class="content__description">(.+?)</div>',
+            webpage, 'description', default=None)
+
+        season = self._html_search_regex(
+            [r'''(?xms)<div\ class="tabs__tab\ tabs__tab--active">\s*
+                    <span>seizoen\ (.+?)</span>\s*
+                </div>''',
+             r'<option value="seizoen (\d{1,3})" data-href="[^"]+?" selected>'],
+            webpage, 'season', default=None)
+
+        season_number = int_or_none(season)
+
+        episode_number = int_or_none(self._html_search_regex(
+            r'''(?xms)<div\ class="content__episode">\s*
+                    <abbr\ title="aflevering">afl</abbr>\s*<span>(\d+)</span>
+                </div>''',
+            webpage, 'episode_number', default=None))
+
+        release_date = parse_iso8601(self._html_search_regex(
+            r'(?ms)<div class="content__broadcastdate">\s*<time\ datetime="(.+?)"',
+            webpage, 'release_date', default=None))
+
+        # If there's a ? or a # in the URL, remove them and everything after
+        clean_url = urlh.geturl().split('?')[0].split('#')[0].strip('/')
+        securevideo_url = clean_url + '.mssecurevideo.json'
+
+        try:
+            video = self._download_json(securevideo_url, display_id)
+        except ExtractorError as e:
+            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
+                self.raise_login_required()
+            raise
+
+        # We are dealing with a '../<show>.relevant' URL
+        redirect_url = video.get('url')
+        if redirect_url:
+            return self.url_result(self._proto_relative_url(redirect_url, 'https:'))
+
+        # There is only one entry, but with an unknown key, so just get
+        # the first one
+        video_id = list(video.values())[0].get('videoid')
+
+        return merge_dicts(info, {
+            '_type': 'url_transparent',
+            'url': 'https://mediazone.vrt.be/api/v1/vrtvideo/assets/%s' % video_id,
+            'ie_key': CanvasIE.ie_key(),
+            'id': video_id,
+            'display_id': display_id,
+            'title': title,
+            'description': description,
+            'season': season,
+            'season_number': season_number,
+            'episode_number': episode_number,
+            'release_date': release_date,
+        })