[medialaan] Fix and improve extraction (closes #11912)
authorSergey M․ <dstftw@gmail.com>
Fri, 17 Mar 2017 22:58:54 +0000 (05:58 +0700)
committerSergey M․ <dstftw@gmail.com>
Fri, 17 Mar 2017 22:58:54 +0000 (05:58 +0700)
youtube_dl/extractor/extractors.py
youtube_dl/extractor/medialaan.py [new file with mode: 0644]
youtube_dl/extractor/vtm.py [deleted file]

index 64316d4a8971c66d0b0a4fa76fe6ddf38dfaf3b3..6b4742ed821dcc81d15d59fab4b1cef0377b789a 100644 (file)
@@ -1174,7 +1174,7 @@ from .voxmedia import VoxMediaIE
 from .vporn import VpornIE
 from .vrt import VRTIE
 from .vrak import VrakIE
-from .vtm import VTMIE
+from .medialaan import MedialaanIE
 from .vube import VubeIE
 from .vuclip import VuClipIE
 from .vvvvid import VVVVIDIE
diff --git a/youtube_dl/extractor/medialaan.py b/youtube_dl/extractor/medialaan.py
new file mode 100644 (file)
index 0000000..e70d467
--- /dev/null
@@ -0,0 +1,263 @@
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import (
+    ExtractorError,
+    int_or_none,
+    parse_duration,
+    try_get,
+    unified_timestamp,
+    urlencode_postdata,
+)
+
+
+class MedialaanIE(InfoExtractor):
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:www\.)?
+                        (?:
+                            (?P<site_id>vtm|q2|vtmkzoom)\.be/
+                            (?:
+                                video(?:/[^/]+/id/|/?\?.*?\baid=)|
+                                (?:[^/]+/)*
+                            )
+                        )
+                        (?P<id>[^/?#&]+)
+                    '''
+    _NETRC_MACHINE = 'medialaan'
+    _APIKEY = '3_HZ0FtkMW_gOyKlqQzW5_0FHRC7Nd5XpXJZcDdXY4pk5eES2ZWmejRW5egwVm4ug-'
+    _SITE_TO_APP_ID = {
+        'vtm': 'vtm_watch',
+        'q2': 'q2',
+        'vtmkzoom': 'vtmkzoom',
+    }
+    _TESTS = [{
+        # vod
+        'url': 'http://vtm.be/video/volledige-afleveringen/id/vtm_20170219_VM0678361_vtmwatch',
+        'info_dict': {
+            'id': 'vtm_20170219_VM0678361_vtmwatch',
+            'ext': 'mp4',
+            'title': 'Allemaal Chris afl. 6',
+            'description': 'md5:4be86427521e7b07e0adb0c9c554ddb2',
+            'timestamp': 1487533280,
+            'upload_date': '20170219',
+            'duration': 2562,
+            'series': 'Allemaal Chris',
+            'season': 'Allemaal Chris',
+            'season_number': 1,
+            'season_id': '256936078124527',
+            'episode': 'Allemaal Chris afl. 6',
+            'episode_number': 6,
+            'episode_id': '256936078591527',
+        },
+        'params': {
+            'skip_download': True,
+        },
+        'skip': 'Requires account credentials',
+    }, {
+        # clip
+        'url': 'http://vtm.be/video?aid=168332',
+        'info_dict': {
+            'id': '168332',
+            'ext': 'mp4',
+            'title': '"Veronique liegt!"',
+            'description': 'md5:1385e2b743923afe54ba4adc38476155',
+            'timestamp': 1489002029,
+            'upload_date': '20170308',
+            'duration': 96,
+        },
+    }, {
+        # vod
+        'url': 'http://vtm.be/video/volledige-afleveringen/id/257107153551000',
+        'only_matching': True,
+    }, {
+        # vod
+        'url': 'http://vtm.be/video?aid=163157',
+        'only_matching': True,
+    }, {
+        # vod
+        'url': 'http://www.q2.be/video/volledige-afleveringen/id/2be_20170301_VM0684442_q2',
+        'only_matching': True,
+    }, {
+        # clip
+        'url': 'http://vitaya.be/de-jurk/precies-je-hebt-geen-borsten',
+        'only_matching': True,
+    }, {
+        # clip
+        'url': 'http://vtmkzoom.be/k3-dansstudio/een-nieuw-seizoen-van-k3-dansstudio',
+        'only_matching': True,
+    }]
+
+    def _real_initialize(self):
+        self._logged_in = False
+
+    def _login(self):
+        username, password = self._get_login_info()
+        if username is None:
+            self.raise_login_required()
+
+        auth_data = {
+            'APIKey': self._APIKEY,
+            'sdk': 'js_6.1',
+            'format': 'json',
+            'loginID': username,
+            'password': password,
+        }
+
+        auth_info = self._download_json(
+            'https://accounts.eu1.gigya.com/accounts.login', None,
+            note='Logging in', errnote='Unable to log in',
+            data=urlencode_postdata(auth_data))
+
+        error_message = auth_info.get('errorDetails') or auth_info.get('errorMessage')
+        if error_message:
+            raise ExtractorError(
+                'Unable to login: %s' % error_message, expected=True)
+
+        self._uid = auth_info['UID']
+        self._uid_signature = auth_info['UIDSignature']
+        self._signature_timestamp = auth_info['signatureTimestamp']
+
+        self._logged_in = True
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id, site_id = mobj.group('id', 'site_id')
+
+        webpage = self._download_webpage(url, video_id)
+
+        config = self._parse_json(
+            self._search_regex(
+                r'videoJSConfig\s*=\s*JSON\.parse\(\'({.+?})\'\);',
+                webpage, 'config', default='{}'), video_id,
+            transform_source=lambda s: s.replace(
+                '\\\\', '\\').replace(r'\"', '"').replace(r"\'", "'"))
+
+        vod_id = config.get('vodId') or self._search_regex(
+            (r'\\"vodId\\"\s*:\s*\\"(.+?)\\"',
+             r'<[^>]+id=["\']vod-(\d+)'),
+            webpage, 'video_id', default=None)
+
+        # clip, no authentication required
+        if not vod_id:
+            player = self._parse_json(
+                self._search_regex(
+                    r'vmmaplayer\(({.+?})\);', webpage, 'vmma player',
+                    default=''),
+                video_id, transform_source=lambda s: '[%s]' % s, fatal=False)
+            if player:
+                video = player[-1]
+                info = {
+                    'id': video_id,
+                    'url': video['videoUrl'],
+                    'title': video['title'],
+                    'thumbnail': video.get('imageUrl'),
+                    'timestamp': int_or_none(video.get('createdDate')),
+                    'duration': int_or_none(video.get('duration')),
+                }
+            else:
+                info = self._parse_html5_media_entries(
+                    url, webpage, video_id, m3u8_id='hls')[0]
+                info.update({
+                    'id': video_id,
+                    'title': self._html_search_meta('description', webpage),
+                    'duration': parse_duration(self._html_search_meta('duration', webpage)),
+                })
+        # vod, authentication required
+        else:
+            if not self._logged_in:
+                self._login()
+
+            settings = self._parse_json(
+                self._search_regex(
+                    r'jQuery\.extend\(Drupal\.settings\s*,\s*({.+?})\);',
+                    webpage, 'drupal settings', default='{}'),
+                video_id)
+
+            def get(container, item):
+                return try_get(
+                    settings, lambda x: x[container][item],
+                    compat_str) or self._search_regex(
+                    r'"%s"\s*:\s*"([^"]+)' % item, webpage, item,
+                    default=None)
+
+            app_id = get('vod', 'app_id') or self._SITE_TO_APP_ID.get(site_id, 'vtm_watch')
+            sso = get('vod', 'gigyaDatabase') or 'vtm-sso'
+
+            data = self._download_json(
+                'http://vod.medialaan.io/api/1.0/item/%s/video' % vod_id,
+                video_id, query={
+                    'app_id': app_id,
+                    'user_network': sso,
+                    'UID': self._uid,
+                    'UIDSignature': self._uid_signature,
+                    'signatureTimestamp': self._signature_timestamp,
+                })
+
+            formats = self._extract_m3u8_formats(
+                data['response']['uri'], video_id, entry_protocol='m3u8_native',
+                ext='mp4', m3u8_id='hls')
+
+            self._sort_formats(formats)
+
+            info = {
+                'id': vod_id,
+                'formats': formats,
+            }
+
+            api_key = get('vod', 'apiKey')
+            channel = get('medialaanGigya', 'channel')
+
+            if api_key:
+                videos = self._download_json(
+                    'http://vod.medialaan.io/vod/v2/videos', video_id, fatal=False,
+                    query={
+                        'channels': channel,
+                        'ids': vod_id,
+                        'limit': 1,
+                        'apikey': api_key,
+                    })
+                if videos:
+                    video = try_get(
+                        videos, lambda x: x['response']['videos'][0], dict)
+                    if video:
+                        def get(container, item, expected_type=None):
+                            return try_get(
+                                video, lambda x: x[container][item], expected_type)
+
+                        def get_string(container, item):
+                            return get(container, item, compat_str)
+
+                        info.update({
+                            'series': get_string('program', 'title'),
+                            'season': get_string('season', 'title'),
+                            'season_number': int_or_none(get('season', 'number')),
+                            'season_id': get_string('season', 'id'),
+                            'episode': get_string('episode', 'title'),
+                            'episode_number': int_or_none(get('episode', 'number')),
+                            'episode_id': get_string('episode', 'id'),
+                            'duration': int_or_none(
+                                video.get('duration')) or int_or_none(
+                                video.get('durationMillis'), scale=1000),
+                            'title': get_string('episode', 'title'),
+                            'description': get_string('episode', 'text'),
+                            'timestamp': unified_timestamp(get_string(
+                                'publication', 'begin')),
+                        })
+
+            if not info.get('title'):
+                info['title'] = try_get(
+                    config, lambda x: x['videoConfig']['title'],
+                    compat_str) or self._html_search_regex(
+                    r'\\"title\\"\s*:\s*\\"(.+?)\\"', webpage, 'title',
+                    default=None) or self._og_search_title(webpage)
+
+        if not info.get('description'):
+            info['description'] = self._html_search_regex(
+                r'<div[^>]+class="field-item\s+even">\s*<p>(.+?)</p>',
+                webpage, 'description', default=None)
+
+        return info
diff --git a/youtube_dl/extractor/vtm.py b/youtube_dl/extractor/vtm.py
deleted file mode 100644 (file)
index f0a7004..0000000
+++ /dev/null
@@ -1,136 +0,0 @@
-from __future__ import unicode_literals
-
-import re
-
-from .generic import GenericIE
-from .common import InfoExtractor
-from ..utils import (
-    urlencode_postdata,
-    compat_urllib_parse_urlencode,
-    ExtractorError,
-    remove_end,
-)
-
-
-class VTMIE(InfoExtractor):
-    """Download full episodes that require an account from vtm.be or q2.be.
-
-    The generic extractor can be used to download clips that do no require an
-    account.
-    """
-    _VALID_URL = r'https?://(?:www\.)?(?P<site_id>vtm|q2)\.be/video[/?].+?'
-    _NETRC_MACHINE = 'vtm'
-    _APIKEY = '3_HZ0FtkMW_gOyKlqQzW5_0FHRC7Nd5XpXJZcDdXY4pk5eES2ZWmejRW5egwVm4ug-'
-    _TESTS = [
-        {
-            'url': 'http://vtm.be/video/volledige-afleveringen/id/vtm_20170219_VM0678361_vtmwatch',
-            'info_dict': {
-                'id': 'vtm_20170219_VM0678361_vtmwatch',
-                'ext': 'mp4',
-                'title': 'Allemaal Chris afl. 6',
-                'description': 'md5:4be86427521e7b07e0adb0c9c554ddb2',
-            },
-            'skip_download': True,
-        },
-        {
-            'url': 'http://vtm.be/video/volledige-afleveringen/id/257107153551000',
-            'only_matching': True,
-        },
-        {
-            'url': 'http://vtm.be/video?aid=163157',
-            'only_matching': True,
-        },
-        {
-            'url': 'http://www.q2.be/video/volledige-afleveringen/id/2be_20170301_VM0684442_q2',
-            'only_matching': True,
-        },
-        {
-            'url': 'http://vtm.be/video?aid=168332',
-            'info_dict': {
-                'id': 'video?aid=168332',
-                'ext': 'mp4',
-                'title': 'Videozone',
-            },
-        },
-    ]
-
-    def _real_initialize(self):
-        self._logged_in = False
-
-    def _login(self):
-        (username, password) = self._get_login_info()
-        if username is None or password is None:
-            self.raise_login_required()
-
-        auth_data = {
-            'APIKey': self._APIKEY,
-            'sdk': 'js_6.1',
-            'format': 'json',
-            'loginID': username,
-            'password': password,
-        }
-
-        auth_info = self._download_json(
-            'https://accounts.eu1.gigya.com/accounts.login', None,
-            note='Logging in', errnote='Unable to log in',
-            data=urlencode_postdata(auth_data), fatal=True)
-
-        error_message = auth_info.get('errorDetails')
-        if error_message:
-            raise ExtractorError(
-                'Unable to login: %s' % error_message, expected=True)
-
-        self._uid = auth_info['UID']
-        self._uid_signature = auth_info['UIDSignature']
-        self._signature_timestamp = auth_info['signatureTimestamp']
-
-        self._logged_in = True
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        site_id = mobj.group('site_id')
-
-        webpage = self._download_webpage(url, None, "Downloading webpage")
-
-        # The URL sometimes contains the video id, but not always, e.g., test
-        # case 3. Fortunately, all webpages of videos requiring authentication
-        # contain the video id.
-        video_id = self._search_regex(
-            r'\\"vodId\\":\\"(.+?)\\"', webpage, 'video_id', default=None)
-
-        # It was most likely a video not requiring authentication.
-        if not video_id:
-            return self.url_result(url, 'Generic')
-
-        if not self._logged_in:
-            self._login()
-
-        title = self._html_search_regex(
-            r'\\"title\\":\\"(.+?)\\"', webpage, 'title', default=None)
-
-        description = self._html_search_regex(
-            r'<div[^>]+class="field-item\s+even">\s*<p>(.+?)</p>',
-            webpage, 'description', default=None)
-
-        data_url = 'http://vod.medialaan.io/api/1.0/item/%s/video' % video_id
-        m3u8_data = {
-            'app_id': 'vtm_watch' if site_id == 'vtm' else 'q2',
-            'user_network': 'vtm-sso',
-            'UID': self._uid,
-            'UIDSignature': self._uid_signature,
-            'signatureTimestamp': self._signature_timestamp,
-        }
-        data = self._download_json(data_url, video_id, query=m3u8_data)
-
-        formats = self._extract_m3u8_formats(
-            data['response']['uri'], video_id, entry_protocol='m3u8_native',
-            ext='mp4', m3u8_id='hls')
-
-        self._sort_formats(formats)
-
-        return {
-            'id': video_id,
-            'title': title,
-            'description': description,
-            'formats': formats,
-        }