perf: 优化文档列表与分块索引查询

- 列表页仅查询展示字段并复用文档分块统计

- 移除文档分页对分块表的关联聚合

- 增加文档分块 document_id 索引和 SQL 回归测试
This commit is contained in:
2026-08-10 23:56:21 +08:00
parent 402c0f16b8
commit 742a4b1647
3 changed files with 129 additions and 17 deletions

View File

@@ -5,6 +5,7 @@ import com.easyagents.core.store.StoreResult;
import com.mybatisflex.core.query.QueryWrapper;
import org.junit.Assert;
import org.junit.Test;
import org.mockito.ArgumentCaptor;
import org.mockito.Mockito;
import tech.easyflow.ai.config.SearcherFactory;
import tech.easyflow.ai.entity.Document;
@@ -21,6 +22,8 @@ import tech.easyflow.common.web.exceptions.BusinessException;
import java.lang.reflect.Field;
import java.math.BigInteger;
import java.util.List;
import java.util.Locale;
import java.util.regex.Pattern;
/**
* {@link DocumentServiceImpl} 文档维护回归测试。
@@ -30,6 +33,70 @@ import java.util.List;
*/
public class DocumentServiceImplTest {
/**
* 验证管理端列表只查询展示字段,并直接使用文档表中的分块统计。
*
* @throws Exception 反射注入异常
*/
@Test
public void getDocumentListShouldAvoidChunkJoinAndLargeFields() throws Exception {
DocumentMapper documentMapper = Mockito.mock(DocumentMapper.class);
DocumentServiceImpl service = new DocumentServiceImpl();
setField(service, "documentMapper", documentMapper);
service.getDocumentList("1001", 10, 1, null);
ArgumentCaptor<QueryWrapper> queryCaptor =
ArgumentCaptor.forClass(QueryWrapper.class);
Mockito.verify(documentMapper).paginateAs(
Mockito.eq(1),
Mockito.eq(10),
queryCaptor.capture(),
Mockito.eq(Document.class)
);
String sql = normalizeSql(queryCaptor.getValue().toSQL());
Assert.assertFalse(sql.contains("tb_document_chunk"));
Assert.assertFalse(containsSqlIdentifier(sql, "content"));
Assert.assertFalse(containsSqlIdentifier(sql, "options"));
Assert.assertFalse(sql.contains("group by"));
Assert.assertTrue(containsSqlIdentifier(sql, "total_chunks"));
Assert.assertTrue(containsSqlIdentifier(sql, "chunk_count"));
}
/**
* 验证公开文档分页继续返回历史详情字段,同时不再关联分块表聚合。
*
* @throws Exception 反射注入异常
*/
@Test
public void getDocumentListByIdShouldKeepDetailFieldsWithoutChunkJoin()
throws Exception {
DocumentMapper documentMapper = Mockito.mock(DocumentMapper.class);
DocumentServiceImpl service = new DocumentServiceImpl();
setField(service, "documentMapper", documentMapper);
service.getDocumentListById(
"1001",
10,
1,
BigInteger.valueOf(2002)
);
ArgumentCaptor<QueryWrapper> queryCaptor =
ArgumentCaptor.forClass(QueryWrapper.class);
Mockito.verify(documentMapper).paginateAs(
Mockito.eq(1),
Mockito.eq(10),
queryCaptor.capture(),
Mockito.eq(Document.class)
);
String sql = normalizeSql(queryCaptor.getValue().toSQL());
Assert.assertFalse(sql.contains("tb_document_chunk"));
Assert.assertFalse(sql.contains("group by"));
Assert.assertTrue(sql.contains("*"));
Assert.assertTrue(containsSqlIdentifier(sql, "chunk_count"));
}
/**
* 验证删除链路在外部索引和分块清理后删除文档主记录。
*
@@ -252,4 +319,27 @@ public class DocumentServiceImplTest {
field.setAccessible(true);
field.set(target, value);
}
/**
* 统一 SQL 文本格式,便于断言查询结构。
*
* @param sql 原始 SQL
* @return 去除标识符引号并转为小写的 SQL
*/
private static String normalizeSql(String sql) {
return sql.replace("`", "").toLowerCase(Locale.ROOT);
}
/**
* 判断 SQL 是否包含完整列标识符,避免与同前缀列名混淆。
*
* @param sql 已标准化的 SQL
* @param identifier 列标识符
* @return 包含完整标识符时返回 true
*/
private static boolean containsSqlIdentifier(String sql, String identifier) {
return Pattern.compile("\\b" + Pattern.quote(identifier) + "\\b")
.matcher(sql)
.find();
}
}