feat(dataset): 实现数据集文件可见性过滤功能

- 添加派生文件识别逻辑，通过元数据中的derived_from_file_id字段判断
- 实现applyVisibleFileCounts方法为数据集批量设置可见文件数量
- 修改数据集统计接口使用过滤后的可见文件进行统计计算
- 添加normalizeFilePath工具方法统一路径格式处理
- 更新文件查询逻辑支持派生文件过滤功能
- 新增DatasetFileCount DTO用于文件计数统计返回

backend/services/data-management-service/src/main/resources/mappers/DatasetFileMapper.xml

@@ -42,6 +42,13 @@
         SELECT COUNT(*) FROM t_dm_dataset_files WHERE dataset_id = #{datasetId}
     </select>
 
+    <select id="countNonDerivedByDatasetId" parameterType="string" resultType="long">
+        SELECT COUNT(*)
+        FROM t_dm_dataset_files
+        WHERE dataset_id = #{datasetId}
+          AND (metadata IS NULL OR JSON_EXTRACT(metadata, '$.derived_from_file_id') IS NULL)
+    </select>
+
     <select id="countCompletedByDatasetId" parameterType="string" resultType="long">
         SELECT COUNT(*) FROM t_dm_dataset_files WHERE dataset_id = #{datasetId} AND status = 'COMPLETED'
     </select>
@@ -110,4 +117,16 @@
           AND metadata IS NOT NULL
           AND JSON_EXTRACT(metadata, '$.derived_from_file_id') IS NOT NULL
     </select>
+
+    <select id="countNonDerivedByDatasetIds" resultType="com.datamate.datamanagement.infrastructure.persistence.repository.dto.DatasetFileCount">
+        SELECT dataset_id AS datasetId,
+               COUNT(*) AS fileCount
+        FROM t_dm_dataset_files
+        WHERE dataset_id IN
+        <foreach collection="datasetIds" item="datasetId" open="(" separator="," close=")">
+            #{datasetId}
+        </foreach>
+          AND (metadata IS NULL OR JSON_EXTRACT(metadata, '$.derived_from_file_id') IS NULL)
+        GROUP BY dataset_id
+    </select>
 </mapper>