fix

api_routes.py

@@ -1,6 +1,7 @@
 import asyncio
 import base64
 import functools
+import glob
 import hashlib
 import inspect
 import io
@@ -3614,6 +3615,28 @@ def _flatten_chinese_celeb_dataset_dir(target_dir: str) -> bool:
     return True
 
 
+def _cleanup_chinese_celeb_hidden_files(target_dir: str) -> int:
+    """
+    删除解压后遗留的 macOS 资源分叉文件（._*），避免污染后续处理。
+    """
+    pattern = os.path.join(target_dir, "._*")
+    removed = 0
+    for hidden_path in glob.glob(pattern):
+        try:
+            if os.path.isdir(hidden_path):
+                shutil.rmtree(hidden_path, ignore_errors=True)
+            else:
+                os.remove(hidden_path)
+            removed += 1
+        except FileNotFoundError:
+            continue
+        except OSError as exc:
+            logger.warning("清理隐藏文件失败: %s (%s)", hidden_path, exc)
+    if removed:
+        logger.info("已清理 chinese_celeb_dataset 隐藏文件 %d 个 (pattern=%s)", removed, pattern)
+    return removed
+
+
 def extract_chinese_celeb_dataset_sync() -> Dict[str, Any]:
     """
     同步执行 chinese_celeb_dataset 解压操作，供启动流程或其他同步场景复用。
@@ -3634,6 +3657,7 @@ def extract_chinese_celeb_dataset_sync() -> Dict[str, Any]:
 
     extract_result = _extract_tar_archive(archive_path, target_dir)
     flattened = _flatten_chinese_celeb_dataset_dir(target_dir)
+    hidden_removed = _cleanup_chinese_celeb_hidden_files(target_dir)
 
     return {
         "success": True,
@@ -3644,6 +3668,7 @@ def extract_chinese_celeb_dataset_sync() -> Dict[str, Any]:
         "stdout": extract_result.get("stdout"),
         "stderr": extract_result.get("stderr"),
         "normalized": flattened,
+        "hidden_removed": hidden_removed,
     }