Merge branch 'subtitles-rework'

(Closes PR #4964)

3 files changed, 185 insertions, 166 deletions

diff --git a/test/parameters.json b/test/parameters.json
index af77b89b4..cbff9bd16 100644
--- a/test/parameters.json
+++ b/test/parameters.json
@@ -28,7 +28,7 @@
     "retries": 10, 
     "simulate": false, 
     "subtitleslang": null, 
-    "subtitlesformat": "srt",
+    "subtitlesformat": "best",
     "test": true, 
     "updatetime": true, 
     "usenetrc": false, 
diff --git a/test/test_YoutubeDL.py b/test/test_YoutubeDL.py
index b1cd6a69f..055e42555 100644
--- a/test/test_YoutubeDL.py
+++ b/test/test_YoutubeDL.py
@@ -337,6 +337,65 @@ class TestFormatSelection(unittest.TestCase):
         downloaded = ydl.downloaded_info_dicts[0]
         self.assertEqual(downloaded['format_id'], 'G')
 
+    def test_subtitles(self):
+        def s_formats(lang, autocaption=False):
+            return [{
+                'ext': ext,
+                'url': 'http://localhost/video.%s.%s' % (lang, ext),
+                '_auto': autocaption,
+            } for ext in ['vtt', 'srt', 'ass']]
+        subtitles = dict((l, s_formats(l)) for l in ['en', 'fr', 'es'])
+        auto_captions = dict((l, s_formats(l, True)) for l in ['it', 'pt', 'es'])
+        info_dict = {
+            'id': 'test',
+            'title': 'Test',
+            'url': 'http://localhost/video.mp4',
+            'subtitles': subtitles,
+            'automatic_captions': auto_captions,
+            'extractor': 'TEST',
+        }
+
+        def get_info(params={}):
+            params.setdefault('simulate', True)
+            ydl = YDL(params)
+            ydl.report_warning = lambda *args, **kargs: None
+            return ydl.process_video_result(info_dict, download=False)
+
+        result = get_info()
+        self.assertFalse(result.get('requested_subtitles'))
+        self.assertEqual(result['subtitles'], subtitles)
+        self.assertEqual(result['automatic_captions'], auto_captions)
+
+        result = get_info({'writesubtitles': True})
+        subs = result['requested_subtitles']
+        self.assertTrue(subs)
+        self.assertEqual(set(subs.keys()), set(['en']))
+        self.assertTrue(subs['en'].get('data') is None)
+        self.assertEqual(subs['en']['ext'], 'ass')
+
+        result = get_info({'writesubtitles': True, 'subtitlesformat': 'foo/srt'})
+        subs = result['requested_subtitles']
+        self.assertEqual(subs['en']['ext'], 'srt')
+
+        result = get_info({'writesubtitles': True, 'subtitleslangs': ['es', 'fr', 'it']})
+        subs = result['requested_subtitles']
+        self.assertTrue(subs)
+        self.assertEqual(set(subs.keys()), set(['es', 'fr']))
+
+        result = get_info({'writesubtitles': True, 'writeautomaticsub': True, 'subtitleslangs': ['es', 'pt']})
+        subs = result['requested_subtitles']
+        self.assertTrue(subs)
+        self.assertEqual(set(subs.keys()), set(['es', 'pt']))
+        self.assertFalse(subs['es']['_auto'])
+        self.assertTrue(subs['pt']['_auto'])
+
+        result = get_info({'writeautomaticsub': True, 'subtitleslangs': ['es', 'pt']})
+        subs = result['requested_subtitles']
+        self.assertTrue(subs)
+        self.assertEqual(set(subs.keys()), set(['es', 'pt']))
+        self.assertTrue(subs['es']['_auto'])
+        self.assertTrue(subs['pt']['_auto'])
+
     def test_add_extra_info(self):
         test_dict = {
             'extractor': 'Foo',
diff --git a/test/test_subtitles.py b/test/test_subtitles.py
index bcc69a778..7f93f0a75 100644
--- a/test/test_subtitles.py
+++ b/test/test_subtitles.py
@@ -18,6 +18,13 @@ from youtube_dl.extractor import (
     VimeoIE,
     WallaIE,
     CeskaTelevizeIE,
+    LyndaIE,
+    NPOIE,
+    ComedyCentralIE,
+    NRKTVIE,
+    RaiIE,
+    VikiIE,
+    ThePlatformIE,
 )
 
 
@@ -27,42 +34,38 @@ class BaseTestSubtitles(unittest.TestCase):
 
     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):
-        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['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
 
-    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)
+        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
@@ -76,12 +79,6 @@ class TestYoutubeSubtitles(BaseTestSubtitles):
         subtitles = self.getSubtitles()
         self.assertEqual(md5(subtitles['en']), '3cb210999d3e021bd6c7f0ea751eab06')
 
-    def test_youtube_list_subtitles(self):
-        self.DL.expect_warning('Video doesn\'t have automatic captions')
-        self.DL.params['listsubtitles'] = True
-        info_dict = self.getInfoDict()
-        self.assertEqual(info_dict, None)
-
     def test_youtube_automatic_captions(self):
         self.url = '8YoUxe5ncPo'
         self.DL.params['writeautomaticsub'] = True
@@ -103,55 +100,22 @@ class TestYoutubeSubtitles(BaseTestSubtitles):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
-        self.assertEqual(len(subtitles), 0)
-
-    def test_youtube_multiple_langs(self):
-        self.url = 'QRS8MkLhQmM'
-        self.DL.params['writesubtitles'] = True
-        langs = ['it', 'fr', 'de']
-        self.DL.params['subtitleslangs'] = langs
-        subtitles = self.getSubtitles()
-        for lang in langs:
-            self.assertTrue(subtitles.get(lang) is not None, 'Subtitles for \'%s\' not extracted' % lang)
+        self.assertFalse(subtitles)
 
 
 class TestDailymotionSubtitles(BaseTestSubtitles):
     url = 'http://www.dailymotion.com/video/xczg00'
     IE = DailymotionIE
 
-    def test_no_writesubtitles(self):
-        subtitles = self.getSubtitles()
-        self.assertEqual(subtitles, None)
-
-    def test_subtitles(self):
-        self.DL.params['writesubtitles'] = True
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['en']), '976553874490cba125086bbfea3ff76f')
-
-    def test_subtitles_lang(self):
-        self.DL.params['writesubtitles'] = True
-        self.DL.params['subtitleslangs'] = ['fr']
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['fr']), '594564ec7d588942e384e920e5341792')
-
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
         self.assertTrue(len(subtitles.keys()) >= 6)
-
-    def test_list_subtitles(self):
-        self.DL.expect_warning('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('Automatic Captions not supported by this server')
-        self.DL.params['writeautomaticsub'] = True
-        self.DL.params['subtitleslang'] = ['en']
-        subtitles = self.getSubtitles()
-        self.assertTrue(len(subtitles.keys()) == 0)
+        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_nosubtitles(self):
         self.DL.expect_warning('video doesn\'t have subtitles')
@@ -159,61 +123,21 @@ class TestDailymotionSubtitles(BaseTestSubtitles):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
-        self.assertEqual(len(subtitles), 0)
-
-    def test_multiple_langs(self):
-        self.DL.params['writesubtitles'] = True
-        langs = ['es', 'fr', 'de']
-        self.DL.params['subtitleslangs'] = langs
-        subtitles = self.getSubtitles()
-        for lang in langs:
-            self.assertTrue(subtitles.get(lang) is not None, 'Subtitles for \'%s\' not extracted' % lang)
+        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):
-        subtitles = self.getSubtitles()
-        self.assertEqual(subtitles, None)
-
-    def test_subtitles(self):
-        self.DL.params['writesubtitles'] = True
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['en']), '4262c1665ff928a2dada178f62cb8d14')
-
-    def test_subtitles_lang(self):
-        self.DL.params['writesubtitles'] = True
-        self.DL.params['subtitleslangs'] = ['fr']
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['fr']), '66a63f7f42c97a50f8c0e90bc7797bb5')
-
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
         self.assertTrue(len(subtitles.keys()) >= 28)
-
-    def test_list_subtitles(self):
-        self.DL.expect_warning('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('Automatic Captions not supported by this server')
-        self.DL.params['writeautomaticsub'] = True
-        self.DL.params['subtitleslang'] = ['en']
-        subtitles = self.getSubtitles()
-        self.assertTrue(len(subtitles.keys()) == 0)
-
-    def test_multiple_langs(self):
-        self.DL.params['writesubtitles'] = True
-        langs = ['es', 'fr', 'de']
-        self.DL.params['subtitleslangs'] = langs
-        subtitles = self.getSubtitles()
-        for lang in langs:
+        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)
 
 
@@ -221,14 +145,7 @@ class TestBlipTVSubtitles(BaseTestSubtitles):
     url = 'http://blip.tv/a/a-6603250'
     IE = BlipTVIE
 
-    def test_list_subtitles(self):
-        self.DL.expect_warning('Automatic Captions not supported by this server')
-        self.DL.params['listsubtitles'] = True
-        info_dict = self.getInfoDict()
-        self.assertEqual(info_dict, None)
-
     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()
@@ -240,39 +157,13 @@ class TestVimeoSubtitles(BaseTestSubtitles):
     url = 'http://vimeo.com/76979871'
     IE = VimeoIE
 
-    def test_no_writesubtitles(self):
-        subtitles = self.getSubtitles()
-        self.assertEqual(subtitles, None)
-
-    def test_subtitles(self):
-        self.DL.params['writesubtitles'] = True
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['en']), '8062383cf4dec168fc40a088aa6d5888')
-
-    def test_subtitles_lang(self):
-        self.DL.params['writesubtitles'] = True
-        self.DL.params['subtitleslangs'] = ['fr']
-        subtitles = self.getSubtitles()
-        self.assertEqual(md5(subtitles['fr']), 'b6191146a6c5d3a452244d853fde6dc8')
-
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
         self.assertEqual(set(subtitles.keys()), set(['de', 'en', 'es', 'fr']))
-
-    def test_list_subtitles(self):
-        self.DL.expect_warning('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('Automatic Captions not supported by this server')
-        self.DL.params['writeautomaticsub'] = True
-        self.DL.params['subtitleslang'] = ['en']
-        subtitles = self.getSubtitles()
-        self.assertTrue(len(subtitles.keys()) == 0)
+        self.assertEqual(md5(subtitles['en']), '8062383cf4dec168fc40a088aa6d5888')
+        self.assertEqual(md5(subtitles['fr']), 'b6191146a6c5d3a452244d853fde6dc8')
 
     def test_nosubtitles(self):
         self.DL.expect_warning('video doesn\'t have subtitles')
@@ -280,27 +171,13 @@ class TestVimeoSubtitles(BaseTestSubtitles):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
-        self.assertEqual(len(subtitles), 0)
-
-    def test_multiple_langs(self):
-        self.DL.params['writesubtitles'] = True
-        langs = ['es', 'fr', 'de']
-        self.DL.params['subtitleslangs'] = langs
-        subtitles = self.getSubtitles()
-        for lang in langs:
-            self.assertTrue(subtitles.get(lang) is not None, 'Subtitles for \'%s\' not extracted' % lang)
+        self.assertFalse(subtitles)
 
 
 class TestWallaSubtitles(BaseTestSubtitles):
     url = 'http://vod.walla.co.il/movie/2705958/the-yes-men'
     IE = WallaIE
 
-    def test_list_subtitles(self):
-        self.DL.expect_warning('Automatic Captions not supported by this server')
-        self.DL.params['listsubtitles'] = True
-        info_dict = self.getInfoDict()
-        self.assertEqual(info_dict, None)
-
     def test_allsubtitles(self):
         self.DL.expect_warning('Automatic Captions not supported by this server')
         self.DL.params['writesubtitles'] = True
@@ -315,19 +192,13 @@ class TestWallaSubtitles(BaseTestSubtitles):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
-        self.assertEqual(len(subtitles), 0)
+        self.assertFalse(subtitles)
 
 
 class TestCeskaTelevizeSubtitles(BaseTestSubtitles):
     url = 'http://www.ceskatelevize.cz/ivysilani/10600540290-u6-uzasny-svet-techniky'
     IE = CeskaTelevizeIE
 
-    def test_list_subtitles(self):
-        self.DL.expect_warning('Automatic Captions not supported by this server')
-        self.DL.params['listsubtitles'] = True
-        info_dict = self.getInfoDict()
-        self.assertEqual(info_dict, None)
-
     def test_allsubtitles(self):
         self.DL.expect_warning('Automatic Captions not supported by this server')
         self.DL.params['writesubtitles'] = True
@@ -342,7 +213,96 @@ class TestCeskaTelevizeSubtitles(BaseTestSubtitles):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
         subtitles = self.getSubtitles()
-        self.assertEqual(len(subtitles), 0)
+        self.assertFalse(subtitles)
+
+
+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['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        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()
+        self.assertEqual(set(subtitles.keys()), set(['nl']))
+        self.assertEqual(md5(subtitles['nl']), 'fc6435027572b63fb4ab143abd5ad3f4')
+
+
+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()
+        self.assertEqual(set(subtitles.keys()), set(['no']))
+        self.assertEqual(md5(subtitles['no']), '1d221e6458c95c5494dcd38e6a1f129a')
+
+
+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['allsubtitles'] = True
+        subtitles = self.getSubtitles()
+        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')
 
 
 if __name__ == '__main__':