From d6e203b3dcef8f291b57021903e629d3e30e1f0b Mon Sep 17 00:00:00 2001 From: Ismael Mejia Date: Fri, 6 Sep 2013 16:26:22 +0200 Subject: [PATCH] [subtitles] fixed multiple subtitles language separated by comma after merge As mentioned in the pull request, I forgot to include this changes. https://github.com/rg3/youtube-dl/commit/aa6a10c44a8e2e86f709c5301f9ea6ac3f01f002 --- test/test_dailymotion_subtitles.py | 13 +++++++++--- test/test_youtube_subtitles.py | 13 +++++++++--- youtube_dl/YoutubeDL.py | 2 +- youtube_dl/__init__.py | 12 +++++++---- youtube_dl/extractor/subtitles.py | 33 +++++++++++++++++------------- youtube_dl/extractor/youtube.py | 2 +- 6 files changed, 49 insertions(+), 26 deletions(-) diff --git a/test/test_dailymotion_subtitles.py b/test/test_dailymotion_subtitles.py index efc4e574f..bcd9f79f6 100644 --- a/test/test_dailymotion_subtitles.py +++ b/test/test_dailymotion_subtitles.py @@ -36,20 +36,20 @@ class TestDailymotionSubtitles(unittest.TestCase): self.assertEqual(md5(subtitles['en']), '976553874490cba125086bbfea3ff76f') def test_subtitles_lang(self): self.DL.params['writesubtitles'] = True - self.DL.params['subtitleslang'] = 'fr' + self.DL.params['subtitleslangs'] = ['fr'] subtitles = self.getSubtitles() self.assertEqual(md5(subtitles['fr']), '594564ec7d588942e384e920e5341792') def test_allsubtitles(self): self.DL.params['allsubtitles'] = True subtitles = self.getSubtitles() self.assertEqual(len(subtitles.keys()), 5) - def test_list_subtitles(self): #ojo + def test_list_subtitles(self): self.DL.params['listsubtitles'] = True info_dict = self.getInfoDict() self.assertEqual(info_dict, None) def test_automatic_captions(self): self.DL.params['writeautomaticsub'] = True - self.DL.params['subtitleslang'] = 'en' + self.DL.params['subtitleslang'] = ['en'] subtitles = self.getSubtitles() self.assertTrue(len(subtitles.keys()) == 0) def test_nosubtitles(self): @@ -57,6 +57,13 @@ class TestDailymotionSubtitles(unittest.TestCase): 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, u'Subtitles for \'%s\' not extracted' % lang) if __name__ == '__main__': unittest.main() diff --git a/test/test_youtube_subtitles.py b/test/test_youtube_subtitles.py index e40243077..5632871ac 100644 --- a/test/test_youtube_subtitles.py +++ b/test/test_youtube_subtitles.py @@ -37,7 +37,7 @@ class TestYoutubeSubtitles(unittest.TestCase): self.assertEqual(md5(subtitles['en']), '4cd9278a35ba2305f47354ee13472260') def test_youtube_subtitles_lang(self): self.DL.params['writesubtitles'] = True - self.DL.params['subtitleslang'] = 'it' + self.DL.params['subtitleslangs'] = ['it'] subtitles = self.getSubtitles() self.assertEqual(md5(subtitles['it']), '164a51f16f260476a05b50fe4c2f161d') def test_youtube_allsubtitles(self): @@ -61,7 +61,7 @@ class TestYoutubeSubtitles(unittest.TestCase): def test_youtube_automatic_captions(self): self.url = '8YoUxe5ncPo' self.DL.params['writeautomaticsub'] = True - self.DL.params['subtitleslang'] = 'it' + self.DL.params['subtitleslangs'] = ['it'] subtitles = self.getSubtitles() self.assertTrue(subtitles['it'] is not None) def test_youtube_nosubtitles(self): @@ -69,7 +69,14 @@ class TestYoutubeSubtitles(unittest.TestCase): 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, u'Subtitles for \'%s\' not extracted' % lang) if __name__ == '__main__': unittest.main() diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py index 1fd610a6e..e9f29e680 100644 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -76,7 +76,7 @@ class YoutubeDL(object): allsubtitles: Downloads all the subtitles of the video listsubtitles: Lists all available subtitles for the video subtitlesformat: Subtitle format [srt/sbv/vtt] (default=srt) - subtitleslang: Language of the subtitles to download + subtitleslangs: Language of the subtitles to download keepvideo: Keep the video file after post-processing daterange: A DateRange object, download only if the upload_date is in the range. skip_download: Skip the actual download of the video file diff --git a/youtube_dl/__init__.py b/youtube_dl/__init__.py index 5d686a928..2c2fd441c 100644 --- a/youtube_dl/__init__.py +++ b/youtube_dl/__init__.py @@ -83,6 +83,9 @@ def parseOpts(overrideArguments=None): return "".join(opts) + def _comma_separated_values_options_callback(option, opt_str, value, parser): + setattr(parser.values, option.dest, value.split(',')) + def _find_term_columns(): columns = os.environ.get('COLUMNS', None) if columns: @@ -203,9 +206,10 @@ def parseOpts(overrideArguments=None): subtitles.add_option('--sub-format', action='store', dest='subtitlesformat', metavar='FORMAT', help='subtitle format (default=srt) ([sbv/vtt] youtube only)', default='srt') - subtitles.add_option('--sub-lang', '--srt-lang', - action='store', dest='subtitleslang', metavar='LANG', - help='language of the subtitles to download (optional) use IETF language tags like \'en\'') + subtitles.add_option('--sub-lang', '--sub-langs', '--srt-lang', + action='callback', dest='subtitleslangs', metavar='LANGS', type='str', + default=[], callback=_comma_separated_values_options_callback, + help='languages of the subtitles to download (optional) separated by commas, use IETF language tags like \'en,pt\'') downloader.add_option('-r', '--rate-limit', dest='ratelimit', metavar='LIMIT', help='maximum download rate (e.g. 50k or 44.6m)') @@ -570,7 +574,7 @@ def _real_main(argv=None): 'allsubtitles': opts.allsubtitles, 'listsubtitles': opts.listsubtitles, 'subtitlesformat': opts.subtitlesformat, - 'subtitleslang': opts.subtitleslang, + 'subtitleslangs': opts.subtitleslangs, 'matchtitle': decodeOption(opts.matchtitle), 'rejecttitle': decodeOption(opts.rejecttitle), 'max_downloads': opts.max_downloads, diff --git a/youtube_dl/extractor/subtitles.py b/youtube_dl/extractor/subtitles.py index caacea5fe..c10cdf266 100644 --- a/youtube_dl/extractor/subtitles.py +++ b/youtube_dl/extractor/subtitles.py @@ -21,24 +21,29 @@ class SubtitlesIE(InfoExtractor): def _extract_subtitles(self, video_id): """ returns {sub_lang: sub} or {} if subtitles not found """ - sub_lang_list = self._get_available_subtitles(video_id) - if not sub_lang_list: # error, it didn't get the available subtitles + available_subs_list = self._get_available_subtitles(video_id) + if not available_subs_list: # error, it didn't get the available subtitles return {} + if self._downloader.params.get('allsubtitles', False): + sub_lang_list = available_subs_list + else: + if self._downloader.params.get('writesubtitles', False): + if self._downloader.params.get('subtitleslangs', False): + requested_langs = self._downloader.params.get('subtitleslangs') + elif 'en' in available_subs_list: + requested_langs = ['en'] + else: + requested_langs = [list(available_subs_list.keys())[0]] - if self._downloader.params.get('writesubtitles', False): - if self._downloader.params.get('subtitleslang', False): - sub_lang = self._downloader.params.get('subtitleslang') - elif 'en' in sub_lang_list: - sub_lang = 'en' - else: - sub_lang = list(sub_lang_list.keys())[0] - if not sub_lang in sub_lang_list: - self._downloader.report_warning(u'no closed captions found in the specified language "%s"' % sub_lang) - return {} - sub_lang_list = {sub_lang: sub_lang_list[sub_lang]} + sub_lang_list = {} + for sub_lang in requested_langs: + if not sub_lang in available_subs_list: + self._downloader.report_warning(u'no closed captions found in the specified language "%s"' % sub_lang) + continue + sub_lang_list[sub_lang] = available_subs_list[sub_lang] subtitles = {} - for sub_lang, url in sub_lang_list.iteritems(): + for sub_lang, url in sub_lang_list.items(): subtitle = self._request_subtitle_url(sub_lang, url) if subtitle: subtitles[sub_lang] = subtitle diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 370cc64cc..b3400df0a 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -160,7 +160,7 @@ class YoutubeSubtitlesIE(SubtitlesIE): def _request_automatic_caption(self, video_id, webpage): """We need the webpage for getting the captions url, pass it as an argument to speed up the process.""" - sub_lang = self._downloader.params.get('subtitleslang') or 'en' + sub_lang = (self._downloader.params.get('subtitleslangs') or ['en'])[0] sub_format = self._downloader.params.get('subtitlesformat') self.to_screen(u'%s: Looking for automatic captions' % video_id) mobj = re.search(r';ytplayer.config = ({.*?});', webpage)