yt_dlp/extractor/neteasemusic.py

@@ -7,6 +7,8 @@ from random import randint
 
 from .common import InfoExtractor
 from ..aes import aes_ecb_encrypt, pkcs7_padding
+from ..compat import compat_urllib_parse_urlencode
+from ..networking import Request
 from ..utils import (
     ExtractorError,
     clean_html,
@@ -22,11 +24,12 @@ from ..utils import (
 
 
 class NetEaseMusicBaseIE(InfoExtractor):
-    _FORMATS = ['bMusic', 'mMusic', 'hMusic']
+    _FORMATS = ['bMusic', 'mMusic', 'hMusic', 'sqMusic', 'hrMusic']
+    _NETEASE_SALT = '3go8&$8*3*3h0k(2)2'
     _API_BASE = 'http://music.163.com/api/'
 
-    def _create_eapi_cipher(self, api_path, query_body, cookies):
-        request_text = json.dumps({**query_body, 'header': cookies}, separators=(',', ':'))
+    def _create_eapi_cipher(self, api_path, query, cookies):
+        request_text = json.dumps({**query, 'header': cookies}, separators=(',', ':'))
 
         message = f'nobody{api_path}use{request_text}md5forencrypt'.encode('latin1')
         msg_digest = md5(message).hexdigest()
@@ -36,7 +39,7 @@ class NetEaseMusicBaseIE(InfoExtractor):
         encrypted = bytes(aes_ecb_encrypt(data, list(b'e82ckenh8dichen8')))
         return f'params={encrypted.hex().upper()}'.encode()
 
-    def _download_eapi_json(self, path, video_id, query_body, headers={}, **kwargs):
+    def _download_eapi_json(self, path, song_id, query, headers={}, **kwargs):
         cookies = {
             'osver': 'undefined',
             'deviceId': 'undefined',
@@ -57,12 +60,13 @@ class NetEaseMusicBaseIE(InfoExtractor):
             **headers,
         }
         url = urljoin('https://interface3.music.163.com/', f'/eapi{path}')
-        data = self._create_eapi_cipher(f'/api{path}', query_body, cookies)
-        return self._download_json(url, video_id, data=data, headers=headers, **kwargs)
+        data = self._create_eapi_cipher(f'/api{path}', query, cookies)
+        return self._download_json(url, song_id, data=data, headers=headers, **kwargs)
 
     def _call_player_api(self, song_id, bitrate):
         return self._download_eapi_json(
-            '/song/enhance/player/url', song_id, {'ids': f'[{song_id}]', 'br': bitrate},
+            '/song/enhance/player/url', song_id,
+            {'ids': f'[{song_id}]', 'br': bitrate},
             note=f'Downloading song URL info: bitrate {bitrate}')
 
     def extract_formats(self, info):
@@ -73,8 +77,10 @@ class NetEaseMusicBaseIE(InfoExtractor):
             details = info.get(song_format)
             if not details:
                 continue
+
             bitrate = int_or_none(details.get('bitrate')) or 999000
-            for song in traverse_obj(self._call_player_api(song_id, bitrate), ('data', ...)):
+            data = self._call_player_api(song_id, bitrate)
+            for song in traverse_obj(data, ('data', ...)):
                 song_url = traverse_obj(song, ('url', {url_or_none}))
                 if not song_url:
                     continue
@@ -82,26 +88,34 @@ class NetEaseMusicBaseIE(InfoExtractor):
                     formats.append({
                         'url': song_url,
                         'format_id': song_format,
-                        'asr': traverse_obj(details, ('sr', {int_or_none})),
                         **traverse_obj(song, {
                             'ext': ('type', {str}),
                             'abr': ('br', {lambda i: int_or_none(i, scale=1000)}),
                             'filesize': ('size', {int_or_none}),
                         }),
+                        **traverse_obj(details, {
+                            'asr': ('sr', {int_or_none}),
+                        }),
                     })
                 elif err == 0:
                     err = try_get(song, lambda x: x['code'], int)
 
         if not formats:
+            msg = 'No media links found'
             if err != 0 and (err < 200 or err >= 400):
-                raise ExtractorError(f'No media links found (site code {err})', expected=True)
+                raise ExtractorError(
+                    '%s (site code %d)' % (msg, err, ), expected=True)
             else:
-                self.raise_geo_restricted('No media links found: probably due to geo restriction.')
+                self.raise_geo_restricted(
+                    msg + ': probably this video is not available from your location due to geo restriction.',
+                    countries=['CN'])
+
         return formats
 
     def query_api(self, endpoint, video_id, note):
-        result = self._download_json(
-            f'{self._API_BASE}{endpoint}', video_id, note, headers={'Referer': self._API_BASE})
+        req = Request('%s%s' % (self._API_BASE, endpoint))
+        req.headers['Referer'] = self._API_BASE
+        result = self._download_json(req, video_id, note)
         if result['code'] == -462:
             self.raise_login_required(f'Login required to download: {result["message"]}')
         elif result['code'] != 200:
@@ -123,7 +137,6 @@ class NetEaseMusicIE(NetEaseMusicBaseIE):
             'timestamp': 1522944000,
             'upload_date': '20180405',
             'description': 'md5:3650af9ee22c87e8637cb2dde22a765c',
-            'subtitles': {'lyric': [{'ext': 'lrc'}]},
             "duration": 256,
             'thumbnail': r're:^http.*\.jpg',
         },
@@ -137,6 +150,7 @@ class NetEaseMusicIE(NetEaseMusicBaseIE):
             'creator': 'Dustin O\'Halloran',
             'upload_date': '20080211',
             'timestamp': 1202745600,
+            'description': 'md5:f12945b0f6e0365e3b73c5032e1b0ff4',
             'duration': 263,
             'thumbnail': r're:^http.*\.jpg',
         },
@@ -151,7 +165,6 @@ class NetEaseMusicIE(NetEaseMusicBaseIE):
             'upload_date': '19911130',
             'timestamp': 691516800,
             'description': 'md5:1ba2f911a2b0aa398479f595224f2141',
-            'subtitles': {'lyric': [{'ext': 'lrc'}]},
             'duration': 268,
             'alt_title': '伴唱:现代人乐队 合唱:总政歌舞团',
             'thumbnail': r're:^http.*\.jpg',
@@ -166,8 +179,7 @@ class NetEaseMusicIE(NetEaseMusicBaseIE):
             'creator': 'Taylor Swift / Kendrick Lamar',
             'upload_date': '20150516',
             'timestamp': 1431792000,
-            'description': 'md5:21535156efb73d6d1c355f95616e285a',
-            'subtitles': {'lyric': [{'ext': 'lrc'}]},
+            'description': 'md5:25fc5f27e47aad975aa6d36382c7833c',
             'duration': 199,
             'thumbnail': r're:^http.*\.jpg',
         },
@@ -183,7 +195,6 @@ class NetEaseMusicIE(NetEaseMusicBaseIE):
             'upload_date': '20100127',
             'timestamp': 1264608000,
             'description': 'md5:79d99cc560e4ca97e0c4d86800ee4184',
-            'subtitles': {'lyric': [{'ext': 'lrc'}]},
             'duration': 229,
             'alt_title': '说出愿望吧(Genie)',
             'thumbnail': r're:^http.*\.jpg',
@@ -192,54 +203,47 @@ class NetEaseMusicIE(NetEaseMusicBaseIE):
     }]
 
     def _process_lyrics(self, lyrics_info):
-        original = traverse_obj(lyrics_info, ('lrc', 'lyric', {str}))
-        translated = traverse_obj(lyrics_info, ('tlyric', 'lyric', {str}))
+        original = lyrics_info.get('lrc', {}).get('lyric')
+        translated = lyrics_info.get('tlyric', {}).get('lyric')
 
-        if original.strip() == '[99:00.00]纯音乐，请欣赏':
-            return None
-
-        if not translated or not original:
+        if not translated:
             return original
 
         lyrics_expr = r'(\[[0-9]{2}:[0-9]{2}\.[0-9]{2,}\])([^\n]+)'
         original_ts_texts = re.findall(lyrics_expr, original)
-        translation_ts_dict = {
-            timestamp: text for timestamp, text in re.findall(lyrics_expr, translated)
-        }
-
-        for i in range(len(original_ts_texts)):
-            timestamp, text = original_ts_texts[i]
-            if translation_ts_dict.get(timestamp):
-                original_ts_texts[i] = timestamp, f'{text} / {translation_ts_dict[timestamp]}'
-        lyrics = '\n'.join([''.join(i) for i in original_ts_texts])
+        translation_ts_dict = dict(
+            (time_stamp, text) for time_stamp, text in re.findall(lyrics_expr, translated)
+        )
+        lyrics = '\n'.join([
+            '%s%s / %s' % (time_stamp, text, translation_ts_dict.get(time_stamp, ''))
+            for time_stamp, text in original_ts_texts
+        ])
         return lyrics
 
     def _real_extract(self, url):
         song_id = self._match_id(url)
 
+        params = {
+            'id': song_id,
+            'ids': '[%s]' % song_id
+        }
         info = self.query_api(
-            f'song/detail?id={song_id}&ids=%5B{song_id}%5D', song_id, 'Downloading song info')['songs'][0]
+            'song/detail?' + compat_urllib_parse_urlencode(params),
+            song_id, 'Downloading song info')['songs'][0]
 
         formats = self.extract_formats(info)
 
-        lyrics = self._process_lyrics(self.query_api(
-            f'song/lyric?id={song_id}&lv=-1&tv=-1', song_id, 'Downloading lyrics data'))
-        lyric_data = {
-            'description': lyrics,
-            'subtitles': {
-                'lyric': [{
-                    'data': lyrics,
-                    'ext': 'lrc',
-                }]
-            }
-        } if lyrics else {}
+        lyrics_info = self.query_api(
+            'song/lyric?id=%s&lv=-1&tv=-1' % song_id,
+            song_id, 'Downloading lyrics data')
+        lyrics = self._process_lyrics(lyrics_info)
 
         return {
             'id': song_id,
+            'description': lyrics,
             'formats': formats,
             'alt_title': '/'.join(traverse_obj(info, (('transNames', 'alias'), ...))) or None,
             'creator': ' / '.join(traverse_obj(info, ('artists', ..., 'name'))),
-            **lyric_data,
             **traverse_obj(info, {
                 'title': ('name', {str}),
                 'timestamp': ('album', 'publishTime', {lambda i: int_or_none(i, scale=1000)}),
@@ -318,28 +322,21 @@ class NetEaseMusicSingerIE(NetEaseMusicBaseIE):
             'title': '李昇基 - 이승기',
         },
         'playlist_count': 50,
-    }, {
-        'note': 'Singer with both translated and alias',
-        'url': 'https://music.163.com/#/artist?id=159692',
-        'info_dict': {
-            'id': '159692',
-            'title': '初音ミク - 初音未来;Hatsune Miku',
-        },
-        'playlist_count': 50,
     }]
 
     def _real_extract(self, url):
         singer_id = self._match_id(url)
 
         info = self.query_api(
-            f'artist/{singer_id}?id={singer_id}', singer_id, note='Downloading singer data')
+            'artist/%s?id=%s' % (singer_id, singer_id),
+            singer_id, 'Downloading singer data')
 
-        name_and_aliases = traverse_obj(info, (
-            'artist', ('name', 'trans', ('alias', ...)), {str}, {lambda i: i or None}))
-        if len(name_and_aliases) > 1:
-            name = f'{name_and_aliases[0]} - {";".join(name_and_aliases[1:])}'
-        else:
-            name = traverse_obj(name_and_aliases, 0)
+        artist_info = info.get('artist', {})
+        name = artist_info.get('name', '')
+        if artist_info.get('trans'):
+            name = '%s - %s' % (name, info['artist']['trans'])
+        if artist_info.get('alias'):
+            name = '%s - %s' % (name, ';'.join(map(str, info['artist']['alias'])))
 
         entries = [
             self.url_result('http://music.163.com/#/song?id=%s' % song['id'],
@@ -457,10 +454,11 @@ class NetEaseMusicMvIE(NetEaseMusicBaseIE):
         mv_id = self._match_id(url)
 
         info = self.query_api(
-            f'mv/detail?id={mv_id}&type=mp4', mv_id, 'Downloading mv info')['data']
+            'mv/detail?id=%s&type=mp4' % mv_id,
+            mv_id, 'Downloading mv info')['data']
 
         formats = [
-            {'url': mv_url, 'ext': 'mp4', 'format_id': f'{brs}p', 'height': int_or_none(brs)}
+            {'url': mv_url, 'ext': 'mp4', 'format_id': '%sp' % brs, 'height': int(brs)}
             for brs, mv_url in info['brs'].items()
         ]
 
@@ -534,7 +532,8 @@ class NetEaseMusicProgramIE(NetEaseMusicBaseIE):
         program_id = self._match_id(url)
 
         info = self.query_api(
-            f'dj/program/detail?id={program_id}', program_id, note='Downloading program info')['program']
+            'dj/program/detail?id=%s' % program_id,
+            program_id, 'Downloading program info')['program']
 
         metainfo = traverse_obj(info, {
             'title': ('name', {str}),
@@ -585,8 +584,9 @@ class NetEaseMusicDjRadioIE(NetEaseMusicBaseIE):
         entries = []
         for offset in itertools.count(start=0, step=self._PAGE_SIZE):
             info = self.query_api(
-                f'dj/program/byradio?asc=false&limit={self._PAGE_SIZE}&radioId={dj_id}&offset={offset}',
-                dj_id, note=f'Downloading dj programs - {offset}')
+                'dj/program/byradio?asc=false&limit=%d&radioId=%s&offset=%d'
+                % (self._PAGE_SIZE, dj_id, offset),
+                dj_id, 'Downloading dj programs - %d' % offset)
 
             entries.extend([
                 self.url_result(