fix apple music

boofilsic/settings.py

@@ -449,6 +449,7 @@ LANGUAGE_CODE, PREFERRED_LANGUAGES = _init_language_settings(
 
 if TESTING:  # force en if testing
     LANGUAGE_CODE = "en"
+    PREFERRED_LANGUAGES = ["en"]
 
 LOCALE_PATHS = [os.path.join(BASE_DIR, "locale")]
 
@@ -580,7 +581,7 @@ SEARCH_INDEX_NEW_ONLY = False
 
 INDEX_ALIASES = env("INDEX_ALIASES")
 
-DOWNLOADER_SAVEDIR = env("NEODB_DOWNLOADER_SAVE_DIR", default="/tmp")  # type: ignore
+DOWNLOADER_SAVEDIR = env("NEODB_DOWNLOADER_SAVE_DIR", default="")  # type: ignore
 
 DISABLE_MODEL_SIGNAL = False  # disable index and social feeds during importing/etc
 

catalog/music/tests.py

@@ -6,6 +6,8 @@ from catalog.music.utils import *
 
 
 class BasicMusicTest(TestCase):
+    databases = "__all__"
+
     def test_gtin(self):
         self.assertIsNone(upc_to_gtin_13("018771208112X"))
         self.assertIsNone(upc_to_gtin_13("999018771208112"))
@@ -15,6 +17,8 @@ class BasicMusicTest(TestCase):
 
 
 class SpotifyTestCase(TestCase):
+    databases = "__all__"
+
     def test_parse(self):
         t_id_type = IdType.Spotify_Album
         t_id_value = "65KwtzkJXw7oT819NFWmEP"
@@ -48,6 +52,8 @@ class SpotifyTestCase(TestCase):
 
 
 class DoubanMusicTestCase(TestCase):
+    databases = "__all__"
+
     def test_parse(self):
         t_id_type = IdType.DoubanMusic
         t_id_value = "33551231"
@@ -74,6 +80,8 @@ class DoubanMusicTestCase(TestCase):
 
 
 class MultiMusicSitesTestCase(TestCase):
+    databases = "__all__"
+
     @use_local_response
     def test_albums(self):
         url1 = "https://music.douban.com/subject/33551231/"
@@ -92,6 +100,8 @@ class MultiMusicSitesTestCase(TestCase):
 
 
 class BandcampTestCase(TestCase):
+    databases = "__all__"
+
     def test_parse(self):
         t_id_type = IdType.Bandcamp
         t_id_value = "intlanthem.bandcamp.com/album/in-these-times"
@@ -119,6 +129,8 @@ class BandcampTestCase(TestCase):
 
 
 class DiscogsReleaseTestCase(TestCase):
+    databases = "__all__"
+
     def test_parse(self):
         t_id_type = IdType.Discogs_Release
         t_id_value = "25829341"
@@ -155,6 +167,8 @@ class DiscogsReleaseTestCase(TestCase):
 
 
 class DiscogsMasterTestCase(TestCase):
+    databases = "__all__"
+
     def test_parse(self):
         t_id_type = IdType.Discogs_Master
         t_id_value = "469004"
@@ -182,6 +196,8 @@ class DiscogsMasterTestCase(TestCase):
 
 
 class AppleMusicTestCase(TestCase):
+    databases = "__all__"
+
     def test_parse(self):
         t_id_type = IdType.AppleMusic
         t_id_value = "1284391545"
@@ -201,8 +217,10 @@ class AppleMusicTestCase(TestCase):
         self.assertEqual(site.ready, False)
         site.get_resource_ready()
         self.assertEqual(site.ready, True)
-        self.assertEqual(site.resource.metadata["title"], "Kids Only")
+        self.assertEqual(
+            site.resource.metadata["localized_title"][0]["text"], "Kids Only"
+        )
         self.assertEqual(site.resource.metadata["artist"], ["Leah Dou"])
         self.assertIsInstance(site.resource.item, Album)
-        self.assertEqual(site.resource.item.genre, ["Pop"])
+        self.assertEqual(site.resource.item.genre, ["Pop", "Music"])
-        self.assertEqual(site.resource.item.duration, 2371628)
+        self.assertEqual(site.resource.item.duration, 2368000)

catalog/sites/apple_music.py

@@ -9,8 +9,9 @@ Scraping the website directly.
 """
 
 import json
+from datetime import timedelta
 
-import dateparser
+from django.utils.dateparse import parse_duration
 from loguru import logger
 
 from catalog.common import *
@@ -18,7 +19,6 @@ from catalog.models import *
 from common.models.lang import (
     SITE_DEFAULT_LANGUAGE,
     SITE_PREFERRED_LANGUAGES,
-    detect_language,
 )
 from common.models.misc import uniq
 
@@ -39,7 +39,6 @@ class AppleMusic(AbstractSite):
     headers = {
         "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:107.0) Gecko/20100101 Firefox/107.0",
         "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
-        "Accept-Language": BasicDownloader.get_accept_language(),
         "Accept-Encoding": "gzip, deflate",
         "Connection": "keep-alive",
         "DNT": "1",
@@ -70,80 +69,63 @@ class AppleMusic(AbstractSite):
         return locales
 
     def scrape(self):
-        matched_content = None
+        matched_schema_data = None
         localized_title = []
         localized_desc = []
         for lang, locales in self.get_locales().items():
             for loc in locales:  # waterfall thru all locales
                 url = f"https://music.apple.com/{loc}/album/{self.id_value}"
                 try:
+                    tl = f"{lang}-{loc}" if lang == "zh" else lang
+                    headers = {
+                        "Accept-Language": tl,
+                    }
+                    headers.update(self.headers)
                     content = (
                         BasicDownloader(url, headers=self.headers).download().html()
                     )
-                    logger.info(f"got localized content from {url}")
+                    logger.debug(f"got localized content from {url}")
-                    elem = content.xpath(
+                    txt: str = content.xpath(
-                        "//script[@id='serialized-server-data']/text()"
+                        "//script[@id='schema:music-album']/text()"
-                    )
+                    )[0]  # type:ignore
-                    txt: str = elem[0]  # type:ignore
+                    schema_data = json.loads(txt)
-                    page_data = json.loads(txt)[0]
+                    title = schema_data["name"]
-                    album_data = page_data["data"]["sections"][0]["items"][0]
+                    if title:
-                    title = album_data["title"]
-                    brief = album_data.get("modalPresentationDescriptor", {}).get(
-                        "paragraphText", ""
-                    )
-                    tl = detect_language(title + " " + brief)
                         localized_title.append({"lang": tl, "text": title})
+                    try:
+                        txt: str = content.xpath(
+                            "//script[@id='serialized-server-data']/text()"
+                        )[0]  # type:ignore
+                        server_data = json.loads(txt)
+                        brief = server_data[0]["data"]["sections"][0]["items"][0][
+                            "modalPresentationDescriptor"
+                        ]["paragraphText"]
                         if brief:
                             localized_desc.append({"lang": tl, "text": brief})
-                    if lang == SITE_DEFAULT_LANGUAGE or not matched_content:
+                    except Exception:
-                        matched_content = content
+                        server_data = brief = None
+                    if lang == SITE_DEFAULT_LANGUAGE or not matched_schema_data:
+                        matched_schema_data = schema_data
                     break
                 except Exception:
                     pass
-        if matched_content is None:
+        if matched_schema_data is None:  # no schema data found
             raise ParseError(self, f"localized content for {self.url}")
-        elem = matched_content.xpath("//script[@id='serialized-server-data']/text()")
+        artist = [a["name"] for a in matched_schema_data.get("byArtist", [])]
-        txt: str = elem[0]  # type:ignore
+        release_date = matched_schema_data.get("datePublished", None)
-        page_data = json.loads(txt)[0]
+        genre = matched_schema_data.get("genre", [])
-        album_data = page_data["data"]["sections"][0]["items"][0]
+        image_url = matched_schema_data.get("image", None)
-        title = album_data["title"]
+        track_list = [t["name"] for t in matched_schema_data.get("tracks", [])]
-        brief = album_data.get("modalPresentationDescriptor")
+        duration = round(
-        brief = brief.get("paragraphText") if brief else None
+            sum(
-        artist_list = album_data["subtitleLinks"]
+                (parse_duration(t["duration"]) or timedelta()).total_seconds() * 1000
-        artist = [item["title"] for item in artist_list]
+                for t in matched_schema_data.get("tracks", [])
-
+            )
-        track_data = page_data["data"]["seoData"]
-        date_elem = track_data.get("musicReleaseDate")
-        release_datetime = dateparser.parse(date_elem.strip()) if date_elem else None
-        release_date = (
-            release_datetime.strftime("%Y-%m-%d") if release_datetime else None
         )
-
-        track_list = [
-            f"{i}. {track['attributes']['name']}"
-            for i, track in enumerate(track_data["ogSongs"], 1)
-        ]
-        duration_list = [
-            track["attributes"].get("durationInMillis", 0)
-            for track in track_data["ogSongs"]
-        ]
-        duration = int(sum(duration_list))
-        genre = track_data["schemaContent"].get("genre")
-        if genre:
-            genre = [
-                genre[0]
-            ]  # apple treat "Music" as a genre. Thus, only the first genre is obtained.
-
-        images = matched_content.xpath("//source[@type='image/jpeg']/@srcset")
-        image_elem: str = images[0] if images else ""  # type:ignore
-        image_url = image_elem.split(" ")[0] if image_elem else None
-
         pd = ResourceContent(
             metadata={
                 "localized_title": uniq(localized_title),
                 "localized_description": uniq(localized_desc),
-                "title": title,
-                "brief": brief,
                 "artist": artist,
                 "genre": genre,
                 "release_date": release_date,