Merge tag 'upstream/2015.02.28'

[youtubedl] / test / test_subtitles.py

diff --git a/test/test_subtitles.py b/test/test_subtitles.py
index 06a304879d122b743f62703f08cdcf00301b0d39..3f2d8a2ba74e6b4f04d4159a64deb1f69f9d105b 100644 (file)
--- a/test/test_subtitles.py
+++ b/test/test_subtitles.py
@@ -1,4 +1,5 @@
 #!/usr/bin/env python
 #!/usr/bin/env python

+from __future__ import unicode_literals

 
 # Allow direct execution
 import os
 
 # Allow direct execution
 import os


@@ -6,62 +7,66 @@ import sys
 import unittest
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 import unittest
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 

-from test.helper import FakeYDL, global_setup, md5
-global_setup()
+from test.helper import FakeYDL, md5

 
 
 from youtube_dl.extractor import (
 
 
 from youtube_dl.extractor import (

+    BlipTVIE,

     YoutubeIE,
     DailymotionIE,
     TEDIE,
     YoutubeIE,
     DailymotionIE,
     TEDIE,

+    VimeoIE,
+    WallaIE,
+    CeskaTelevizeIE,
+    LyndaIE,
+    NPOIE,
+    ComedyCentralIE,
+    NRKTVIE,
+    RaiIE,
+    VikiIE,
+    ThePlatformIE,
+    RTVEALaCartaIE,

 )
 
 
 class BaseTestSubtitles(unittest.TestCase):
     url = None
     IE = None
 )
 
 
 class BaseTestSubtitles(unittest.TestCase):
     url = None
     IE = None

+

     def setUp(self):
         self.DL = FakeYDL()
     def setUp(self):
         self.DL = FakeYDL()

-        self.ie = self.IE(self.DL)
+        self.ie = self.IE()
+        self.DL.add_info_extractor(self.ie)

 
     def getInfoDict(self):
 
     def getInfoDict(self):

-        info_dict = self.ie.extract(self.url)
+        info_dict = self.DL.extract_info(self.url, download=False)

         return info_dict
 
     def getSubtitles(self):
         info_dict = self.getInfoDict()
         return info_dict
 
     def getSubtitles(self):
         info_dict = self.getInfoDict()

-        return info_dict['subtitles']
+        subtitles = info_dict['requested_subtitles']
+        if not subtitles:
+            return subtitles
+        for sub_info in subtitles.values():
+            if sub_info.get('data') is None:
+                uf = self.DL.urlopen(sub_info['url'])
+                sub_info['data'] = uf.read().decode('utf-8')
+        return dict((l, sub_info['data']) for l, sub_info in subtitles.items())

 
 
 class TestYoutubeSubtitles(BaseTestSubtitles):
     url = 'QRS8MkLhQmM'
     IE = YoutubeIE
 
 
 
 class TestYoutubeSubtitles(BaseTestSubtitles):
     url = 'QRS8MkLhQmM'
     IE = YoutubeIE
 

-    def getSubtitles(self):
-        info_dict = self.getInfoDict()
-        return info_dict[0]['subtitles']
-
-    def test_youtube_no_writesubtitles(self):
-        self.DL.params['writesubtitles'] = False
-        subtitles = self.getSubtitles()
-        self.assertEqual(subtitles, None)
-
-    def test_youtube_subtitles(self):
-        self.DL.params['writesubtitles'] = True
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['en']), '4cd9278a35ba2305f47354ee13472260')
-
-    def test_youtube_subtitles_lang(self):
-        self.DL.params['writesubtitles'] = True
-        self.DL.params['subtitleslangs'] = ['it']
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['it']), '164a51f16f260476a05b50fe4c2f161d')
-

     def test_youtube_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
         self.assertEqual(len(subtitles.keys()), 13)
     def test_youtube_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
         self.assertEqual(len(subtitles.keys()), 13)

+        self.assertEqual(md5(subtitles['en']), '4cd9278a35ba2305f47354ee13472260')
+        self.assertEqual(md5(subtitles['it']), '164a51f16f260476a05b50fe4c2f161d')
+        for lang in ['it', 'fr', 'de']:
+            self.assertTrue(subtitles.get(lang) is not None, 'Subtitles for \'%s\' not extracted' % lang)

 
     def test_youtube_subtitles_sbv_format(self):
         self.DL.params['writesubtitles'] = True
 
     def test_youtube_subtitles_sbv_format(self):
         self.DL.params['writesubtitles'] = True


@@ -73,13 +78,7 @@ class TestYoutubeSubtitles(BaseTestSubtitles):
         self.DL.params['writesubtitles'] = True
         self.DL.params['subtitlesformat'] = 'vtt'
         subtitles = self.getSubtitles()
         self.DL.params['writesubtitles'] = True
         self.DL.params['subtitlesformat'] = 'vtt'
         subtitles = self.getSubtitles()

-        self.assertEqual(md5(subtitles['en']), '356cdc577fde0c6783b9b822e7206ff7')
-
-    def test_youtube_list_subtitles(self):
-        self.DL.expect_warning(u'Video doesn\'t have automatic captions')
-        self.DL.params['listsubtitles'] = True
-        info_dict = self.getInfoDict()
-        self.assertEqual(info_dict, None)
+        self.assertEqual(md5(subtitles['en']), '3cb210999d3e021bd6c7f0ea751eab06')

 
     def test_youtube_automatic_captions(self):
         self.url = '8YoUxe5ncPo'
 
     def test_youtube_automatic_captions(self):
         self.url = '8YoUxe5ncPo'


@@ -88,124 +87,238 @@ class TestYoutubeSubtitles(BaseTestSubtitles):
         subtitles = self.getSubtitles()
         self.assertTrue(subtitles['it'] is not None)
 
         subtitles = self.getSubtitles()
         self.assertTrue(subtitles['it'] is not None)
 

-    def test_youtube_nosubtitles(self):
-        self.DL.expect_warning(u'video doesn\'t have subtitles')
-        self.url = 'sAjKT8FhjI8'
-        self.DL.params['writesubtitles'] = True
-        self.DL.params['allsubtitles'] = True
+    def test_youtube_translated_subtitles(self):
+        # This video has a subtitles track, which can be translated
+        self.url = 'Ky9eprVWzlI'
+        self.DL.params['writeautomaticsub'] = True
+        self.DL.params['subtitleslangs'] = ['it']

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertEqual(len(subtitles), 0)
+        self.assertTrue(subtitles['it'] is not None)

 
 

-    def test_youtube_multiple_langs(self):
-        self.url = 'QRS8MkLhQmM'
+    def test_youtube_nosubtitles(self):
+        self.DL.expect_warning('video doesn\'t have subtitles')
+        self.url = 'n5BB19UTcdA'

         self.DL.params['writesubtitles'] = True
         self.DL.params['writesubtitles'] = True

-        langs = ['it', 'fr', 'de']
-        self.DL.params['subtitleslangs'] = langs
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        for lang in langs:
-            self.assertTrue(subtitles.get(lang) is not None, u'Subtitles for \'%s\' not extracted' % lang)
+        self.assertFalse(subtitles)

 
 
 class TestDailymotionSubtitles(BaseTestSubtitles):
     url = 'http://www.dailymotion.com/video/xczg00'
     IE = DailymotionIE
 
 
 
 class TestDailymotionSubtitles(BaseTestSubtitles):
     url = 'http://www.dailymotion.com/video/xczg00'
     IE = DailymotionIE
 

-    def test_no_writesubtitles(self):
+    def test_allsubtitles(self):
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertEqual(subtitles, None)
+        self.assertTrue(len(subtitles.keys()) >= 6)
+        self.assertEqual(md5(subtitles['en']), '976553874490cba125086bbfea3ff76f')
+        self.assertEqual(md5(subtitles['fr']), '594564ec7d588942e384e920e5341792')
+        for lang in ['es', 'fr', 'de']:
+            self.assertTrue(subtitles.get(lang) is not None, 'Subtitles for \'%s\' not extracted' % lang)

 
 

-    def test_subtitles(self):
+    def test_nosubtitles(self):
+        self.DL.expect_warning('video doesn\'t have subtitles')
+        self.url = 'http://www.dailymotion.com/video/x12u166_le-zapping-tele-star-du-08-aout-2013_tv'

         self.DL.params['writesubtitles'] = True
         self.DL.params['writesubtitles'] = True

+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertEqual(md5(subtitles['en']), '976553874490cba125086bbfea3ff76f')
+        self.assertFalse(subtitles)

 
 

-    def test_subtitles_lang(self):
+
+class TestTedSubtitles(BaseTestSubtitles):
+    url = 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html'
+    IE = TEDIE
+
+    def test_allsubtitles(self):

         self.DL.params['writesubtitles'] = True
         self.DL.params['writesubtitles'] = True

-        self.DL.params['subtitleslangs'] = ['fr']
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertEqual(md5(subtitles['fr']), '594564ec7d588942e384e920e5341792')
+        self.assertTrue(len(subtitles.keys()) >= 28)
+        self.assertEqual(md5(subtitles['en']), '4262c1665ff928a2dada178f62cb8d14')
+        self.assertEqual(md5(subtitles['fr']), '66a63f7f42c97a50f8c0e90bc7797bb5')
+        for lang in ['es', 'fr', 'de']:
+            self.assertTrue(subtitles.get(lang) is not None, 'Subtitles for \'%s\' not extracted' % lang)
+
+
+class TestBlipTVSubtitles(BaseTestSubtitles):
+    url = 'http://blip.tv/a/a-6603250'
+    IE = BlipTVIE

 
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
 
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()

-        self.assertEqual(len(subtitles.keys()), 5)
+        self.assertEqual(set(subtitles.keys()), set(['en']))
+        self.assertEqual(md5(subtitles['en']), '5b75c300af65fe4476dff79478bb93e4')

 
 

-    def test_list_subtitles(self):
-        self.DL.expect_warning(u'Automatic Captions not supported by this server')
-        self.DL.params['listsubtitles'] = True
-        info_dict = self.getInfoDict()
-        self.assertEqual(info_dict, None)

 
 

-    def test_automatic_captions(self):
-        self.DL.expect_warning(u'Automatic Captions not supported by this server')
-        self.DL.params['writeautomaticsub'] = True
-        self.DL.params['subtitleslang'] = ['en']
+class TestVimeoSubtitles(BaseTestSubtitles):
+    url = 'http://vimeo.com/76979871'
+    IE = VimeoIE
+
+    def test_allsubtitles(self):
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertTrue(len(subtitles.keys()) == 0)
+        self.assertEqual(set(subtitles.keys()), set(['de', 'en', 'es', 'fr']))
+        self.assertEqual(md5(subtitles['en']), '8062383cf4dec168fc40a088aa6d5888')
+        self.assertEqual(md5(subtitles['fr']), 'b6191146a6c5d3a452244d853fde6dc8')

 
     def test_nosubtitles(self):
 
     def test_nosubtitles(self):

-        self.DL.expect_warning(u'video doesn\'t have subtitles')
-        self.url = 'http://www.dailymotion.com/video/x12u166_le-zapping-tele-star-du-08-aout-2013_tv'
+        self.DL.expect_warning('video doesn\'t have subtitles')
+        self.url = 'http://vimeo.com/56015672'

         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()

-        self.assertEqual(len(subtitles), 0)
+        self.assertFalse(subtitles)
+

 
 

-    def test_multiple_langs(self):
+class TestWallaSubtitles(BaseTestSubtitles):
+    url = 'http://vod.walla.co.il/movie/2705958/the-yes-men'
+    IE = WallaIE
+
+    def test_allsubtitles(self):
+        self.DL.expect_warning('Automatic Captions not supported by this server')

         self.DL.params['writesubtitles'] = True
         self.DL.params['writesubtitles'] = True

-        langs = ['es', 'fr', 'de']
-        self.DL.params['subtitleslangs'] = langs
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        for lang in langs:
-            self.assertTrue(subtitles.get(lang) is not None, u'Subtitles for \'%s\' not extracted' % lang)
+        self.assertEqual(set(subtitles.keys()), set(['heb']))
+        self.assertEqual(md5(subtitles['heb']), 'e758c5d7cb982f6bef14f377ec7a3920')

 
 

+    def test_nosubtitles(self):
+        self.DL.expect_warning('video doesn\'t have subtitles')
+        self.url = 'http://vod.walla.co.il/movie/2642630/one-direction-all-for-one'
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        self.assertFalse(subtitles)

 
 

-class TestTedSubtitles(BaseTestSubtitles):
-    url = 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html'
-    IE = TEDIE

 
 

-    def test_no_writesubtitles(self):
+class TestCeskaTelevizeSubtitles(BaseTestSubtitles):
+    url = 'http://www.ceskatelevize.cz/ivysilani/10600540290-u6-uzasny-svet-techniky'
+    IE = CeskaTelevizeIE
+
+    def test_allsubtitles(self):
+        self.DL.expect_warning('Automatic Captions not supported by this server')
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertEqual(subtitles, None)
+        self.assertEqual(set(subtitles.keys()), set(['cs']))
+        self.assertTrue(len(subtitles['cs']) > 20000)

 
 

-    def test_subtitles(self):
+    def test_nosubtitles(self):
+        self.DL.expect_warning('video doesn\'t have subtitles')
+        self.url = 'http://www.ceskatelevize.cz/ivysilani/ivysilani/10441294653-hyde-park-civilizace/214411058091220'

         self.DL.params['writesubtitles'] = True
         self.DL.params['writesubtitles'] = True

+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertEqual(md5(subtitles['en']), '2154f31ff9b9f89a0aa671537559c21d')
+        self.assertFalse(subtitles)
+

 
 

-    def test_subtitles_lang(self):
+class TestLyndaSubtitles(BaseTestSubtitles):
+    url = 'http://www.lynda.com/Bootstrap-tutorials/Using-exercise-files/110885/114408-4.html'
+    IE = LyndaIE
+
+    def test_allsubtitles(self):

         self.DL.params['writesubtitles'] = True
         self.DL.params['writesubtitles'] = True

-        self.DL.params['subtitleslangs'] = ['fr']
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertEqual(md5(subtitles['fr']), '7616cbc6df20ec2c1204083c83871cf6')
+        self.assertEqual(set(subtitles.keys()), set(['en']))
+        self.assertEqual(md5(subtitles['en']), '09bbe67222259bed60deaa26997d73a7')
+
+
+class TestNPOSubtitles(BaseTestSubtitles):
+    url = 'http://www.npo.nl/nos-journaal/28-08-2014/POW_00722860'
+    IE = NPOIE

 
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
 
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()

-        self.assertEqual(len(subtitles.keys()), 28)
+        self.assertEqual(set(subtitles.keys()), set(['nl']))
+        self.assertEqual(md5(subtitles['nl']), 'fc6435027572b63fb4ab143abd5ad3f4')

 
 

-    def test_list_subtitles(self):
-        self.DL.expect_warning(u'Automatic Captions not supported by this server')
-        self.DL.params['listsubtitles'] = True
-        info_dict = self.getInfoDict()
-        self.assertEqual(info_dict, None)

 
 

-    def test_automatic_captions(self):
-        self.DL.expect_warning(u'Automatic Captions not supported by this server')
-        self.DL.params['writeautomaticsub'] = True
-        self.DL.params['subtitleslang'] = ['en']
+class TestMTVSubtitles(BaseTestSubtitles):
+    url = 'http://www.cc.com/video-clips/kllhuv/stand-up-greg-fitzsimmons--uncensored---too-good-of-a-mother'
+    IE = ComedyCentralIE
+
+    def getInfoDict(self):
+        return super(TestMTVSubtitles, self).getInfoDict()['entries'][0]
+
+    def test_allsubtitles(self):
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        self.assertEqual(set(subtitles.keys()), set(['en']))
+        self.assertEqual(md5(subtitles['en']), 'b9f6ca22a6acf597ec76f61749765e65')
+
+
+class TestNRKSubtitles(BaseTestSubtitles):
+    url = 'http://tv.nrk.no/serie/ikke-gjoer-dette-hjemme/DMPV73000411/sesong-2/episode-1'
+    IE = NRKTVIE
+
+    def test_allsubtitles(self):
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        self.assertTrue(len(subtitles.keys()) == 0)
+        self.assertEqual(set(subtitles.keys()), set(['no']))
+        self.assertEqual(md5(subtitles['no']), '1d221e6458c95c5494dcd38e6a1f129a')
+

 
 

-    def test_multiple_langs(self):
+class TestRaiSubtitles(BaseTestSubtitles):
+    url = 'http://www.rai.tv/dl/RaiTV/programmi/media/ContentItem-cb27157f-9dd0-4aee-b788-b1f67643a391.html'
+    IE = RaiIE
+
+    def test_allsubtitles(self):

         self.DL.params['writesubtitles'] = True
         self.DL.params['writesubtitles'] = True

-        langs = ['es', 'fr', 'de']
-        self.DL.params['subtitleslangs'] = langs
+        self.DL.params['allsubtitles'] = True

         subtitles = self.getSubtitles()
         subtitles = self.getSubtitles()

-        for lang in langs:
-            self.assertTrue(subtitles.get(lang) is not None, u'Subtitles for \'%s\' not extracted' % lang)
+        self.assertEqual(set(subtitles.keys()), set(['it']))
+        self.assertEqual(md5(subtitles['it']), 'b1d90a98755126b61e667567a1f6680a')
+
+
+class TestVikiSubtitles(BaseTestSubtitles):
+    url = 'http://www.viki.com/videos/1060846v-punch-episode-18'
+    IE = VikiIE
+
+    def test_allsubtitles(self):
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        self.assertEqual(set(subtitles.keys()), set(['en']))
+        self.assertEqual(md5(subtitles['en']), '53cb083a5914b2d84ef1ab67b880d18a')
+
+
+class TestThePlatformSubtitles(BaseTestSubtitles):
+    # from http://www.3playmedia.com/services-features/tools/integrations/theplatform/
+    # (see http://theplatform.com/about/partners/type/subtitles-closed-captioning/)
+    url = 'theplatform:JFUjUE1_ehvq'
+    IE = ThePlatformIE
+
+    def test_allsubtitles(self):
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        self.assertEqual(set(subtitles.keys()), set(['en']))
+        self.assertEqual(md5(subtitles['en']), '97e7670cbae3c4d26ae8bcc7fdd78d4b')
+
+
+class TestRtveSubtitles(BaseTestSubtitles):
+    url = 'http://www.rtve.es/alacarta/videos/los-misterios-de-laura/misterios-laura-capitulo-32-misterio-del-numero-17-2-parte/2428621/'
+    IE = RTVEALaCartaIE
+
+    def test_allsubtitles(self):
+        print('Skipping, only available from Spain')
+        return
+        self.DL.params['writesubtitles'] = True
+        self.DL.params['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        self.assertEqual(set(subtitles.keys()), set(['es']))
+        self.assertEqual(md5(subtitles['es']), '69e70cae2d40574fb7316f31d6eb7fca')
+

 
 if __name__ == '__main__':
     unittest.main()
 
 if __name__ == '__main__':
     unittest.main()