[Hidive] Fix subtitles broken by 705e7c2005

yt_dlp/extractor/hidive.py

@@ -1,5 +1,5 @@
 # coding: utf-8
-from __future__ import unicode_literals
+import re
 
 from .common import InfoExtractor
 from ..utils import (
@@ -52,15 +52,39 @@ class HiDiveIE(InfoExtractor):
         self._download_webpage(
             self._LOGIN_URL, None, 'Logging in', data=urlencode_postdata(data))
 
+    def _call_api(self, video_id, title, key, data={}, **kwargs):
+        data = {
+            **data,
+            'Title': title,
+            'Key': key,
+            'PlayerId': 'f4f895ce1ca713ba263b91caeb1daa2d08904783',
+        }
+        return self._download_json(
+            'https://www.hidive.com/play/settings', video_id,
+            data=urlencode_postdata(data), **kwargs) or {}
+
+    def _extract_subtitles_from_rendition(self, rendition, subtitles, parsed_urls):
+        for cc_file in rendition.get('ccFiles', []):
+            cc_url = url_or_none(try_get(cc_file, lambda x: x[2]))
+            # name is used since we cant distinguish subs with same language code
+            cc_lang = try_get(cc_file, (lambda x: x[1].replace(' ', '-').lower(), lambda x: x[0]), str)
+            if cc_url not in parsed_urls and cc_lang:
+                parsed_urls.add(cc_url)
+                subtitles.setdefault(cc_lang, []).append({'url': cc_url})
+
+    def _get_subtitles(self, url, video_id, title, key, subtitles, parsed_urls):
+        webpage = self._download_webpage(url, video_id, fatal=False) or ''
+        for caption in set(re.findall(r'data-captions=\"([^\"]+)\"', webpage)):
+            renditions = self._call_api(
+                video_id, title, key, {'Captions': caption}, fatal=False,
+                note=f'Downloading {caption} subtitle information').get('renditions') or {}
+            for rendition_id, rendition in renditions.items():
+                self._extract_subtitles_from_rendition(rendition, subtitles, parsed_urls)
+        return subtitles
+
     def _real_extract(self, url):
         video_id, title, key = self._match_valid_url(url).group('id', 'title', 'key')
-        settings = self._download_json(
+        settings = self._call_api(video_id, title, key)
-            'https://www.hidive.com/play/settings', video_id,
-            data=urlencode_postdata({
-                'Title': title,
-                'Key': key,
-                'PlayerId': 'f4f895ce1ca713ba263b91caeb1daa2d08904783',
-            }))
 
         restriction = settings.get('restrictionReason')
         if restriction == 'RegionRestricted':
@@ -69,12 +93,12 @@ class HiDiveIE(InfoExtractor):
             raise ExtractorError(
                 '%s said: %s' % (self.IE_NAME, restriction), expected=True)
 
-        formats, subtitles, urls = [], {}, {None}
+        formats, subtitles, parsed_urls = [], {}, {None}
         for rendition_id, rendition in settings['renditions'].items():
             audio, version, extra = rendition_id.split('_')
             m3u8_url = url_or_none(try_get(rendition, lambda x: x['bitrates']['hls']))
-            if m3u8_url not in urls:
+            if m3u8_url not in parsed_urls:
-                urls.add(m3u8_url)
+                parsed_urls.add(m3u8_url)
                 frmt = self._extract_m3u8_formats(
                     m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native', m3u8_id=rendition_id, fatal=False)
                 for f in frmt:
@@ -82,19 +106,13 @@ class HiDiveIE(InfoExtractor):
                     f['format_note'] = f'{version}, {extra}'
                 formats.extend(frmt)
 
-            for cc_file in rendition.get('ccFiles', []):
+            self._extract_subtitles_from_rendition(rendition, subtitles, parsed_urls)
-                cc_url = url_or_none(try_get(cc_file, lambda x: x[2]))
-                # name is used since we cant distinguish subs with same language code
-                cc_lang = try_get(cc_file, (lambda x: x[1].replace(' ', '-').lower(), lambda x: x[0]), str)
-                if cc_url not in urls and cc_lang:
-                    urls.add(cc_url)
-                    subtitles.setdefault(cc_lang, []).append({'url': cc_url})
         self._sort_formats(formats)
 
         return {
             'id': video_id,
             'title': video_id,
-            'subtitles': subtitles,
+            'subtitles': self.extract_subtitles(url, video_id, title, key, subtitles, parsed_urls),
             'formats': formats,
             'series': title,
             'season_number': int_or_none(