]> jfr.im git - yt-dlp.git/blobdiff - yt_dlp/extractor/hidive.py
[ie/crunchyroll] Fix stream extraction (#10005)
[yt-dlp.git] / yt_dlp / extractor / hidive.py
index f26f80265678576217994746d9286482c5184a3d..df6868df65d294a73c981d35aebab4194353bb80 100644 (file)
@@ -1,20 +1,15 @@
-# coding: utf-8
-from __future__ import unicode_literals
-
-import re
-
 from .common import InfoExtractor
-from ..compat import compat_str
 from ..utils import (
     ExtractorError,
     int_or_none,
+    try_get,
     url_or_none,
     urlencode_postdata,
 )
 
 
 class HiDiveIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?hidive\.com/stream/(?P<title>[^/]+)/(?P<key>[^/?#&]+)'
+    _VALID_URL = r'https?://(?:www\.)?hidive\.com/stream/(?P<id>(?P<title>[^/]+)/(?P<key>[^/?#&]+))'
     # Using X-Forwarded-For results in 403 HTTP error for HLS fragments,
     # so disabling geo bypass completely
     _GEO_BYPASS = False
@@ -37,75 +32,78 @@ class HiDiveIE(InfoExtractor):
         'skip': 'Requires Authentication',
     }]
 
-    def _real_initialize(self):
-        email, password = self._get_login_info()
-        if email is None:
-            return
-
+    def _perform_login(self, username, password):
         webpage = self._download_webpage(self._LOGIN_URL, None)
         form = self._search_regex(
             r'(?s)<form[^>]+action="/account/login"[^>]*>(.+?)</form>',
-            webpage, 'login form')
+            webpage, 'login form', default=None)
+        if not form:
+            return
         data = self._hidden_inputs(form)
         data.update({
-            'Email': email,
+            'Email': username,
             'Password': password,
         })
-        self._download_webpage(
+        login_webpage = self._download_webpage(
             self._LOGIN_URL, None, 'Logging in', data=urlencode_postdata(data))
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        title, key = mobj.group('title', 'key')
-        video_id = '%s/%s' % (title, key)
-
-        settings = self._download_json(
-            'https://www.hidive.com/play/settings', video_id,
+        # If the user has multiple profiles on their account, select one. For now pick the first profile.
+        profile_id = self._search_regex(
+            r'<button [^>]+?data-profile-id="(\w+)"', login_webpage, 'profile id', default=None)
+        if profile_id is None:
+            return  # If only one profile, Hidive auto-selects it
+        self._request_webpage(
+            'https://www.hidive.com/ajax/chooseprofile', None,
             data=urlencode_postdata({
-                'Title': title,
-                'Key': key,
-                'PlayerId': 'f4f895ce1ca713ba263b91caeb1daa2d08904783',
+                'profileId': profile_id,
+                'hash': self._search_regex(
+                    r'\<button [^>]+?data-hash="(\w+)"', login_webpage, 'profile id hash'),
+                'returnUrl': '/dashboard'
             }))
 
+    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 _real_extract(self, url):
+        video_id, title, key = self._match_valid_url(url).group('id', 'title', 'key')
+        settings = self._call_api(video_id, title, key)
+
         restriction = settings.get('restrictionReason')
         if restriction == 'RegionRestricted':
             self.raise_geo_restricted()
-
         if restriction and restriction != 'None':
             raise ExtractorError(
                 '%s said: %s' % (self.IE_NAME, restriction), expected=True)
 
-        formats = []
-        subtitles = {}
+        formats, parsed_urls = [], {None}
         for rendition_id, rendition in settings['renditions'].items():
-            bitrates = rendition.get('bitrates')
-            if not isinstance(bitrates, dict):
-                continue
-            m3u8_url = url_or_none(bitrates.get('hls'))
-            if not m3u8_url:
-                continue
-            formats.extend(self._extract_m3u8_formats(
-                m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native',
-                m3u8_id='%s-hls' % rendition_id, fatal=False))
-            cc_files = rendition.get('ccFiles')
-            if not isinstance(cc_files, list):
-                continue
-            for cc_file in cc_files:
-                if not isinstance(cc_file, list) or len(cc_file) < 3:
-                    continue
-                cc_lang = cc_file[0]
-                cc_url = url_or_none(cc_file[2])
-                if not isinstance(cc_lang, compat_str) or not cc_url:
-                    continue
-                subtitles.setdefault(cc_lang, []).append({
-                    'url': cc_url,
-                })
-        self._sort_formats(formats)
+            audio, version, extra = rendition_id.split('_')
+            m3u8_url = url_or_none(try_get(rendition, lambda x: x['bitrates']['hls']))
+            if m3u8_url not in parsed_urls:
+                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:
+                    f['language'] = audio
+                    f['format_note'] = f'{version}, {extra}'
+                formats.extend(frmt)
 
-        season_number = int_or_none(self._search_regex(
-            r's(\d+)', key, 'season number', default=None))
-        episode_number = int_or_none(self._search_regex(
-            r'e(\d+)', key, 'episode number', default=None))
+        subtitles = {}
+        for rendition_id, rendition in settings['renditions'].items():
+            audio, version, extra = rendition_id.split('_')
+            for cc_file in rendition.get('ccFiles') or []:
+                cc_url = url_or_none(try_get(cc_file, lambda x: x[2]))
+                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})
 
         return {
             'id': video_id,
@@ -113,6 +111,9 @@ def _real_extract(self, url):
             'subtitles': subtitles,
             'formats': formats,
             'series': title,
-            'season_number': season_number,
-            'episode_number': episode_number,
+            'season_number': int_or_none(
+                self._search_regex(r's(\d+)', key, 'season number', default=None)),
+            'episode_number': int_or_none(
+                self._search_regex(r'e(\d+)', key, 'episode number', default=None)),
+            'http_headers': {'Referer': url}
         }