fix: list_datasets 改为本地表分页，修复 page_size 超限 500

去掉 service 层硬编码 page_size=1000 拉取 RAGFlow 全部数据集的写法
（撞 SDK 上限 100 导致 500），改为直接从 user_datasets 表分页返回。

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

services/knowledge_base_service.py

@@ -108,7 +108,7 @@ class KnowledgeBaseService:
         search: str = None
     ) -> Dict[str, Any]:
         """
-        获取用户的数据集列表（从本地数据库过滤）
+        获取用户的数据集列表（本地数据库分页）
 
         Args:
             user_id: 用户 ID
@@ -123,10 +123,8 @@ class KnowledgeBaseService:
 
         pool = get_db_pool_manager().pool
 
-        # 从本地数据库获取用户的数据集 ID 列表
         async with pool.connection() as conn:
             async with conn.cursor() as cursor:
-                # 构建查询条件
                 where_conditions = ["user_id = %s"]
                 params = [user_id]
 
@@ -136,49 +134,33 @@ class KnowledgeBaseService:
 
                 where_clause = " AND ".join(where_conditions)
 
-                # 获取总数
                 await cursor.execute(f"""
                     SELECT COUNT(*) FROM user_datasets
                     WHERE {where_clause}
                 """, params)
                 total = (await cursor.fetchone())[0]
 
-                # 获取分页数据
                 offset = (page - 1) * page_size
                 await cursor.execute(f"""
-                    SELECT dataset_id, dataset_name, created_at
+                    SELECT dataset_id, dataset_name, owner, created_at
                     FROM user_datasets
                     WHERE {where_clause}
                     ORDER BY created_at DESC
                     LIMIT %s OFFSET %s
                 """, params + [page_size, offset])
 
-                user_datasets = await cursor.fetchall()
+                rows = await cursor.fetchall()
 
-        if not user_datasets:
+        items = [
-            return {
+            {
-                "items": [],
+                "dataset_id": row[0],
-                "total": 0,
+                "name": row[1],
-                "page": page,
+                "description": None,
-                "page_size": page_size
+                "owner": row[2],
+                "created_at": row[3],
             }
-
+            for row in rows
-        # 获取数据集 ID 列表，从 RAGFlow 获取详情
+        ]
-        dataset_ids = [row[0] for row in user_datasets]
-        dataset_names = {row[0]: row[1] for row in user_datasets}
-
-        # 从 RAGFlow 获取完整的数据集信息
-        ragflow_result = await self.repository.list_datasets(
-            page=1,
-            page_size=1000  # 获取所有数据集，然后在本地过滤
-        )
-
-        # 过滤出属于该用户的数据集
-        user_dataset_ids_set = set(dataset_ids)
-        items = []
-        for item in ragflow_result["items"]:
-            if item.get("dataset_id") in user_dataset_ids_set:
-                items.append(item)
 
         return {
             "items": items,