From 3cd3bea1cb42d17be7ca1e6651a145c5ef315805 Mon Sep 17 00:00:00 2001
From: hailin <hailin@gdzx.xyz>
Date: Sat, 10 May 2025 03:17:00 +0800
Subject: [PATCH] .

---
 app/api/search.py | 22 +++++++++++++---------
 1 file changed, 13 insertions(+), 9 deletions(-)

diff --git a/app/api/search.py b/app/api/search.py
index d695b3f..0d59bea 100644
--- a/app/api/search.py
+++ b/app/api/search.py
@@ -23,8 +23,12 @@ def read_file_with_detected_encoding(file_path: str):
         raw_data = f.read()
         result = chardet.detect(raw_data)
         encoding = result['encoding']  # 获取检测到的编码
-    with open(file_path, "r", encoding=encoding) as f:
-        return f.read()
+    try:
+        with open(file_path, "r", encoding=encoding, errors='ignore') as f:
+            return f.read()
+    except UnicodeDecodeError:
+        logger.error(f"UnicodeDecodeError: Unable to decode the file using {encoding}.")
+        raise HTTPException(status_code=500, detail="Error reading the file due to encoding issues.")
 
 @router.post("/search")
 def search_docs(request: QueryRequest, user_id: str = Query(..., description="用户ID")):
@@ -44,14 +48,14 @@ def search_docs(request: QueryRequest, user_id: str = Query(..., description="
         logger.info(f"Loading Faiss vector store from path: {index_path}")
         storage_context = StorageContext.from_defaults(persist_dir=index_path)
 
-        # 自动读取文件内容并解码
-        vector_store_file = os.path.join(index_path, "vector_store.json")
-        if os.path.exists(vector_store_file):
-            file_content = read_file_with_detected_encoding(vector_store_file)
-            logger.info(f"Successfully read vector_store.json with detected encoding.")
+        # 自动读取 docstore.json 文件内容并解码
+        docstore_file = os.path.join(index_path, "docstore.json")
+        if os.path.exists(docstore_file):
+            file_content = read_file_with_detected_encoding(docstore_file)
+            logger.info(f"Successfully read docstore.json with detected encoding.")
         else:
-            logger.error(f"vector_store.json not found at {vector_store_file}")
-            raise HTTPException(status_code=404, detail="vector_store.json not found")
+            logger.error(f"docstore.json not found at {docstore_file}")
+            raise HTTPException(status_code=404, detail="docstore.json not found")
 
         # 加载 Faiss 向量存储
         faiss_store = FaissVectorStore.from_persist_path(storage_context)