Merge branch 'lecture2go' of https://github.com/nichdu/youtube-dl into nichdu-lecture2go

[youtube-dl] / youtube_dl / extractor / viki.py

diff --git a/youtube_dl/extractor/viki.py b/youtube_dl/extractor/viki.py
index 071e280fb9fabf2c797f175be083b30fdce2f20c..51cdc6b65143aaf4a0d2823ffa8c859c96e25972 100644 (file)
--- a/youtube_dl/extractor/viki.py
+++ b/youtube_dl/extractor/viki.py
@@ -1,6 +1,7 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
-import re
+import json
 import time
 import hmac
 import hashlib
@@ -12,10 +13,12 @@ from ..utils import (
     parse_age_limit,
     parse_iso8601,
 )
+from ..compat import compat_urllib_request
 from .common import InfoExtractor
 
 
 class VikiBaseIE(InfoExtractor):
+    _VALID_URL_BASE = r'https?://(?:www\.)?viki\.(?:com|net|mx|jp|fr)/'
     _API_QUERY_TEMPLATE = '/v4/%sapp=%s&t=%s&site=www.viki.com'
     _API_URL_TEMPLATE = 'http://api.viki.io%s&sig=%s'
 
@@ -23,27 +26,35 @@ class VikiBaseIE(InfoExtractor):
     _APP_VERSION = '2.2.5.1428709186'
     _APP_SECRET = '-$iJ}@p7!G@SyU/je1bEyWg}upLu-6V6-Lg9VD(]siH,r.,m-r|ulZ,U4LC/SeR)'
 
-    def _prepare_call(self, path, timestamp=None):
+    _NETRC_MACHINE = 'viki'
+
+    _token = None
+
+    def _prepare_call(self, path, timestamp=None, post_data=None):
         path += '?' if '?' not in path else '&'
         if not timestamp:
             timestamp = int(time.time())
         query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp)
+        if self._token:
+            query += '&token=%s' % self._token
         sig = hmac.new(
             self._APP_SECRET.encode('ascii'),
             query.encode('ascii'),
             hashlib.sha1
         ).hexdigest()
-        return self._API_URL_TEMPLATE % (query, sig)
+        url = self._API_URL_TEMPLATE % (query, sig)
+        return compat_urllib_request.Request(
+            url, json.dumps(post_data).encode('utf-8')) if post_data else url
 
-    def _call_api(self, path, video_id, note, timestamp=None):
+    def _call_api(self, path, video_id, note, timestamp=None, post_data=None):
         resp = self._download_json(
-            self._prepare_call(path, timestamp), video_id, note)
+            self._prepare_call(path, timestamp, post_data), video_id, note)
 
         error = resp.get('error')
         if error:
             if error == 'invalid timestamp':
                 resp = self._download_json(
-                    self._prepare_call(path, int(resp['current_timestamp'])),
+                    self._prepare_call(path, int(resp['current_timestamp']), post_data),
                     video_id, '%s (retry)' % note)
                 error = resp.get('error')
             if error:
@@ -56,10 +67,31 @@ class VikiBaseIE(InfoExtractor):
             '%s returned error: %s' % (self.IE_NAME, error),
             expected=True)
 
+    def _real_initialize(self):
+        self._login()
+
+    def _login(self):
+        (username, password) = self._get_login_info()
+        if username is None:
+            return
+
+        login_form = {
+            'login_id': username,
+            'password': password,
+        }
+
+        login = self._call_api(
+            'sessions.json', None,
+            'Logging in as %s' % username, post_data=login_form)
+
+        self._token = login.get('token')
+        if not self._token:
+            self.report_warning('Unable to get session token, login has probably failed')
+
 
 class VikiIE(VikiBaseIE):
     IE_NAME = 'viki'
-    _VALID_URL = r'https?://(?:www\.)?viki\.com/(?:videos|player)/(?P<id>[0-9]+v)'
+    _VALID_URL = r'%s(?:videos|player)/(?P<id>[0-9]+v)' % VikiBaseIE._VALID_URL_BASE
     _TESTS = [{
         'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14',
         'info_dict': {
@@ -157,7 +189,7 @@ class VikiIE(VikiBaseIE):
             title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id
             container_titles = video.get('container', {}).get('titles')
             if container_titles:
-                container_title = container_titles.get('en') or container_titles[titles.keys()[0]]
+                container_title = container_titles.get('en') or container_titles[container_titles.keys()[0]]
                 title = '%s - %s' % (container_title, title)
 
         descriptions = video.get('descriptions')
@@ -230,7 +262,7 @@ class VikiIE(VikiBaseIE):
 
 class VikiChannelIE(VikiBaseIE):
     IE_NAME = 'viki:channel'
-    _VALID_URL = r'https?://(?:www\.)?viki\.com/tv/(?P<id>[0-9]+c)'
+    _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P<id>[0-9]+c)' % VikiBaseIE._VALID_URL_BASE
     _TESTS = [{
         'url': 'http://www.viki.com/tv/50c-boys-over-flowers',
         'info_dict': {
@@ -247,6 +279,15 @@ class VikiChannelIE(VikiBaseIE):
             'description': 'md5:05bf5471385aa8b21c18ad450e350525',
         },
         'playlist_count': 127,
+    }, {
+        'url': 'http://www.viki.com/news/24569c-showbiz-korea',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.viki.com/artists/2141c-shinee',
+        'only_matching': True,
     }]
 
     _PER_PAGE = 25
@@ -265,7 +306,7 @@ class VikiChannelIE(VikiBaseIE):
         description = descriptions.get('en') or descriptions[descriptions.keys()[0]]
 
         entries = []
-        for video_type in ('episodes', 'clips'):
+        for video_type in ('episodes', 'clips', 'movies'):
             for page_num in itertools.count(1):
                 page = self._call_api(
                     'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d'