mirror of
https://gitee.com/mybatis-flex/mybatis-flex.git
synced 2025-12-07 00:58:24 +08:00
874 lines
20 KiB
Markdown
874 lines
20 KiB
Markdown
# 灵活的 QueryWrapper
|
||
在 [增删改](./add-delete-update) 和 [查询和分页](./query) 章节中，我们随时能看到 QueryWrapper 的身影，QueryWrapper 是用于构造 Sql 的
|
||
强有力工具，也是 MyBatis-Flex 的亮点和特色。
|
||
|
||
::: tip 提示
|
||
QueryWrapper 可以被序列化通过 RPC 进行传输，因此，在微服务项目中，我们可以在客户端（网关、Controller 层等）构造出 QueryWrapper，传给
|
||
Provider 层进行查询返回数据。
|
||
:::
|
||
|
||
## QueryWrapper 的使用
|
||
|
||
以下代码是一个完整 Spring Controller 的示例：
|
||
|
||
```java
|
||
@RestController
|
||
public class AccountController {
|
||
|
||
@Autowired
|
||
AccountMapper accountMapper;
|
||
|
||
@GetMapping("/accounts")
|
||
List<Account> selectList() {
|
||
|
||
//构造 QueryWrapper，也支持使用 QueryWrapper.create() 构造，效果相同
|
||
QueryWrapper query = new QueryWrapper();
|
||
query.where(ACCOUNT.ID.ge(100));
|
||
|
||
//通过 query 查询数据列表返回
|
||
return accountMapper.selectListByQuery(query);
|
||
}
|
||
}
|
||
```
|
||
在以上的示例中，其核心代码如下所示：
|
||
|
||
```java
|
||
//构造 QueryWrapper
|
||
QueryWrapper query = new QueryWrapper();
|
||
query.where(ACCOUNT.ID.ge(100));
|
||
|
||
//通过 query 查询数据列表
|
||
accountMapper.selectListByQuery(query);
|
||
```
|
||
以上代码执行的 Sql 如下：
|
||
|
||
```sql
|
||
select * from tb_account
|
||
where id >= 100
|
||
```
|
||
|
||
::: tip 问题：以上示例中，`ACCOUNT.ID.ge(100)` 中的 `ACCOUNT` 是怎么来的？
|
||
MyBatis-Flex 使用了 APT 技术，这个 `ACCOUNT` 是自动生成的。
|
||
参考：《[MyBatis-Flex APT 配置](../others/apt.md)》章节。
|
||
:::
|
||
|
||
## select *
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper();
|
||
query.select(ACCOUNT.ID, ACCOUNT.USER_NAME)
|
||
.from(ACCOUNT)
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT id, user_name FROM tb_account
|
||
```
|
||
|
||
## select ... as
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.select(
|
||
ACCOUNT.ID.as("accountId")
|
||
, ACCOUNT.USER_NAME)
|
||
.from(ACCOUNT.as("a"));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT a.id as accountId, a.user_name
|
||
FROM tb_account AS a
|
||
```
|
||
|
||
## select 多张表
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.select(
|
||
ACCOUNT.ID
|
||
, ACCOUNT.USER_NAME
|
||
, ARTICLE.ID.as("articleId")
|
||
, ARTICLE.TITLE)
|
||
.from(ACCOUNT.as("a"), ARTICLE.as("b"))
|
||
.where(ACCOUNT.ID.eq(ARTICLE.ACCOUNT_ID));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT a.id, a.user_name, b.id AS articleId, b.title
|
||
FROM tb_account AS a, tb_article AS b
|
||
WHERE a.id = b.account_id
|
||
```
|
||
|
||
## select function（SQL 函数）
|
||
|
||
示例 ：
|
||
|
||
```java
|
||
QueryWrapper query=new QueryWrapper()
|
||
.select(
|
||
ACCOUNT.ID,
|
||
ACCOUNT.USER_NAME,
|
||
max(ACCOUNT.BIRTHDAY),
|
||
avg(ACCOUNT.SEX).as("sex_avg")
|
||
).from(ACCOUNT);
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT id, user_name, MAX(birthday), AVG(sex) AS sex_avg
|
||
FROM tb_account
|
||
```
|
||
|
||
|
||
|
||
目前，MyBatis-Flex 内置的函数支持如下：
|
||
|
||
- count
|
||
- max
|
||
- min
|
||
- avg
|
||
- sum
|
||
- year
|
||
- month
|
||
- day
|
||
- convert
|
||
- concat
|
||
|
||
更多的函数，用户可以参考 [QueryMethods](https://gitee.com/mybatis-flex/mybatis-flex/blob/main/mybatis-flex-core/src/main/java/com/mybatisflex/core/query/QueryMethods.java)
|
||
，然后再自己的项目里进行自定义扩展。
|
||
|
||
|
||
## select 列计算
|
||
|
||
|
||
#### 示例 1：
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.select(ACCOUNT.ID.add(100).as("x100"))
|
||
.from(ACCOUNT);
|
||
|
||
String sql = query.toSQL();
|
||
```
|
||
> 列计算的 **加减乘除** 对应的方法分别为：add / subtract / multiply / divide
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT (`id` + 100) AS `x100` FROM `tb_account`
|
||
```
|
||
|
||
|
||
#### 示例 2：
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.select(sum(ACCOUNT.ID.multiply(ACCOUNT.AGE)).as("total_x"))
|
||
.from(ACCOUNT);
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT SUM(`id` * `age`) AS `total_x` FROM `tb_account`
|
||
```
|
||
|
||
|
||
#### 示例 3：
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.select(ACCOUNT.ID.add(ACCOUNT.AGE.add(100)).as("x100"))
|
||
.from(ACCOUNT);
|
||
|
||
String sql = query.toSQL();
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT (`id` + (`age` + 100)) AS `x100` FROM `tb_account`
|
||
```
|
||
|
||
|
||
## select case...when
|
||
|
||
**示例 1：**
|
||
|
||
```java
|
||
QueryWrapper wrapper = QueryWrapper.create()
|
||
.select(
|
||
ACCOUNT.ID
|
||
,case_()
|
||
.when(ACCOUNT.ID.ge(2)).then("x2")
|
||
.when(ACCOUNT.ID.ge(1)).then("x1")
|
||
.else_("x100")
|
||
.end().as("xName")
|
||
)
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT `id`,
|
||
(CASE WHEN `id` >= 2 THEN 'x2'
|
||
WHEN `id` >= 1 THEN 'x1'
|
||
ELSE 'x100'
|
||
END) AS `xName`
|
||
FROM `tb_account`
|
||
```
|
||
|
||
SQL 执行的结果如下：
|
||
|
||
```
|
||
|id |xName |
|
||
|1 |x1 |
|
||
|2 |x2 |
|
||
```
|
||
|
||
|
||
**示例 2：**
|
||
|
||
```java
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select(
|
||
ACCOUNT.ALL_COLUMNS,
|
||
case_(ACCOUNT.ID)
|
||
.when(100).then(100)
|
||
.when(200).then(200)
|
||
.else_(300).end().as("result")
|
||
)
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.USER_NAME.like("michael"));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT *,
|
||
(CASE `id`
|
||
WHEN 100 THEN 100
|
||
WHEN 200 THEN 200
|
||
ELSE 300 END) AS `result`
|
||
FROM `tb_account`
|
||
WHERE `user_name` LIKE ?
|
||
```
|
||
|
||
::: tip 提示
|
||
在以上示例中，由于 `case` 和 `else` 属于 Java 关键字，无法使用其进行方法命名，因此会添加一个下划线小尾巴 `"_"` 变成 `case_` 和 `else_`，这是无奈之举。
|
||
在以后的 QueryWrapper 构建中，遇到 java 关键字也会采用类似的解决方法。
|
||
:::
|
||
|
||
|
||
## with...select
|
||
|
||
示例 1：
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.with("CTE").asSelect(
|
||
select().from(ARTICLE).where(ARTICLE.ID.ge(100))
|
||
)
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.SEX.eq(1));
|
||
|
||
System.out.println(query.toSQL());
|
||
```
|
||
|
||
生成的 SQL 如下：
|
||
|
||
```sql
|
||
WITH CTE AS (SELECT * FROM `tb_article` WHERE `id` >= 100)
|
||
SELECT * FROM `tb_account` WHERE `sex` = 1
|
||
```
|
||
|
||
|
||
示例 2：
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.with("xxx", "id", "name").asValues(
|
||
Arrays.asList("a", "b"),
|
||
union(
|
||
select().from(ARTICLE).where(ARTICLE.ID.ge(200))
|
||
)
|
||
)
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.SEX.eq(1));
|
||
|
||
System.out.println(query.toSQL());
|
||
```
|
||
|
||
生成的 SQL 如下：
|
||
|
||
```sql
|
||
WITH xxx(id, name)
|
||
AS (VALUES (a, b) UNION (SELECT * FROM `tb_article` WHERE `id` >= 200))
|
||
SELECT * FROM `tb_account` WHERE `sex` = 1
|
||
```
|
||
|
||
|
||
|
||
|
||
## with recursive...select
|
||
|
||
示例 1：
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.withRecursive("CTE").asSelect(
|
||
select().from(ARTICLE).where(ARTICLE.ID.ge(100))
|
||
)
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.SEX.eq(1));
|
||
|
||
System.out.println(query.toSQL());
|
||
```
|
||
|
||
生成的 SQL 如下：
|
||
|
||
```sql
|
||
WITH RECURSIVE CTE AS (SELECT * FROM `tb_article` WHERE `id` >= 100)
|
||
SELECT * FROM `tb_account` WHERE `sex` = 1
|
||
```
|
||
|
||
|
||
示例 2：
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.withRecursive("CTE").asSelect(
|
||
select().from(ARTICLE).where(ARTICLE.ID.ge(100))
|
||
)
|
||
.with("xxx", "id", "name").asValues(
|
||
Arrays.asList("a", "b"),
|
||
union(
|
||
select().from(ARTICLE).where(ARTICLE.ID.ge(200))
|
||
)
|
||
)
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.SEX.eq(1));
|
||
|
||
System.out.println(query.toSQL());
|
||
```
|
||
|
||
生成的 SQL 如下：
|
||
|
||
```sql
|
||
WITH RECURSIVE
|
||
CTE AS (SELECT * FROM `tb_article` WHERE `id` >= 100),
|
||
xxx(id, name) AS (VALUES (a, b) UNION (SELECT * FROM `tb_article` WHERE `id` >= 200))
|
||
SELECT * FROM `tb_account` WHERE `sex` = 1
|
||
```
|
||
|
||
|
||
|
||
|
||
## where
|
||
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.ID.ge(100))
|
||
.and(ACCOUNT.USER_NAME.like("michael"));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE id >= ?
|
||
AND user_name LIKE ?
|
||
```
|
||
|
||
## where 动态条件 1
|
||
|
||
```java 1,4
|
||
boolean flag = false;
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select().from(ACCOUNT)
|
||
.where(flag ? ACCOUNT.ID.ge(100) : noCondition())
|
||
.and(ACCOUNT.USER_NAME.like("michael"));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE user_name LIKE ?
|
||
```
|
||
|
||
## where 动态条件 2
|
||
|
||
```java 1,4
|
||
boolean flag = false;
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select().from(ACCOUNT)
|
||
.where(ACCOUNT.ID.ge(100).when(flag)) // when....
|
||
.and(ACCOUNT.USER_NAME.like("michael"));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE user_name LIKE ?
|
||
```
|
||
|
||
## where 动态条件 3
|
||
|
||
```java 1,5
|
||
String name = null;
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select().from(ACCOUNT)
|
||
.where(ACCOUNT.ID.ge(100)) // when....
|
||
.and(ACCOUNT.USER_NAME.like(name).when(StringUtil::isNotBlank));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE id >= ?
|
||
```
|
||
|
||
## where 动态条件 4
|
||
|
||
```java 1,5
|
||
String name = null;
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select().from(ACCOUNT)
|
||
.where(ACCOUNT.ID.ge(100))
|
||
.and(ACCOUNT.USER_NAME.like(name, If::hasText));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE id >= ?
|
||
```
|
||
|
||
## where select
|
||
```java
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.ID.ge(
|
||
select(ARTICLE.ACCOUNT_ID).from(ARTICLE).where(ARTICLE.ID.ge(100))
|
||
));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE id >=
|
||
(SELECT account_id FROM tb_article WHERE id >= ? )
|
||
```
|
||
|
||
## where exists, not exists
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.ID.ge(100))
|
||
.and(
|
||
exists( // or notExists(...)
|
||
selectOne().from(ARTICLE).where(ARTICLE.ID.ge(100))
|
||
)
|
||
);
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE id >= ?
|
||
AND EXISTS (
|
||
SELECT 1 FROM tb_article WHERE id >= ?
|
||
)
|
||
```
|
||
|
||
## and (...) or (...)
|
||
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.where(ACCOUNT.ID.ge(100))
|
||
.and(ACCOUNT.SEX.eq(1).or(ACCOUNT.SEX.eq(2)))
|
||
.or(ACCOUNT.AGE.in(18,19,20).and(ACCOUNT.USER_NAME.like("michael")));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
WHERE id >= ?
|
||
AND (sex = ? OR sex = ? )
|
||
OR (age IN (?,?,?) AND user_name LIKE ? )
|
||
```
|
||
|
||
## group by
|
||
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.groupBy(ACCOUNT.USER_NAME);
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
GROUP BY user_name
|
||
```
|
||
|
||
## having
|
||
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.groupBy(ACCOUNT.USER_NAME)
|
||
.having(ACCOUNT.AGE.between(18,25));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
GROUP BY user_name
|
||
HAVING age BETWEEN ? AND ?
|
||
```
|
||
|
||
## orderBy
|
||
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.orderBy(ACCOUNT.AGE.asc(), ACCOUNT.USER_NAME.desc().nullsLast());
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
ORDER BY age ASC, user_name DESC NULLS LAST
|
||
```
|
||
|
||
## hint
|
||
|
||
Hint 是数据库厂商（比如 Oracle、MySQL、达梦等）提供的一种 SQL语法，它允许用户在 SQL 语句中插入相关的语法，从而影响 SQL 的执行方式。
|
||
它是一种【非常规】的直接影响优化器、指定执行计划的 SQL 优化手段。
|
||
|
||
|
||
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select().hint("INDEX_DESC")
|
||
.from(ACCOUNT)
|
||
.orderBy(ACCOUNT.AGE.asc(), ACCOUNT.USER_NAME.desc().nullsLast());
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT /*+ INDEX_DESC */ * FROM tb_account
|
||
ORDER BY age ASC, user_name DESC NULLS LAST
|
||
```
|
||
|
||
|
||
## join（left join，inner join...）
|
||
|
||
```java
|
||
QueryWrapper queryWrapper=QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.leftJoin(ARTICLE).on(ACCOUNT.ID.eq(ARTICLE.ACCOUNT_ID))
|
||
.innerJoin(ARTICLE).on(ACCOUNT.ID.eq(ARTICLE.ACCOUNT_ID))
|
||
.where(ACCOUNT.AGE.ge(10));
|
||
```
|
||
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
LEFT JOIN tb_article ON tb_account.id = tb_article.account_id
|
||
INNER JOIN tb_article ON tb_account.id = tb_article.account_id
|
||
WHERE tb_account.age >= ?
|
||
```
|
||
|
||
## join on 多个条件
|
||
|
||
```java
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.leftJoin(ARTICLE).on(
|
||
ACCOUNT.ID.eq(ARTICLE.ACCOUNT_ID).and(ACCOUNT.AGE.eq(18))
|
||
)
|
||
.where(ACCOUNT.AGE.ge(10));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account LEFT JOIN tb_article
|
||
ON tb_account.id = tb_article.account_id AND tb_account.age = ?
|
||
WHERE tb_account.age >= ?
|
||
```
|
||
|
||
## join select
|
||
|
||
```java
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.leftJoin(
|
||
select().from(ARTICLE).where(ARTICLE.ID.ge(100))
|
||
).as("a").on(
|
||
ACCOUNT.ID.eq(raw("a.id"))
|
||
)
|
||
.where(ACCOUNT.AGE.ge(10));
|
||
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
SELECT * FROM tb_account
|
||
LEFT JOIN (SELECT * FROM tb_article WHERE id >= ? ) AS a
|
||
ON tb_account.id = a.id
|
||
WHERE tb_account.age >= ?
|
||
```
|
||
|
||
## union, union all
|
||
|
||
```java
|
||
QueryWrapper query = new QueryWrapper()
|
||
.select(ACCOUNT.ID)
|
||
.from(ACCOUNT)
|
||
.orderBy(ACCOUNT.ID.desc())
|
||
.union(select(ARTICLE.ID).from(ARTICLE))
|
||
.unionAll(select(ARTICLE.ID).from(ARTICLE));
|
||
```
|
||
|
||
其查询生成的 Sql 如下：
|
||
|
||
```sql
|
||
(SELECT id FROM tb_account ORDER BY id DESC)
|
||
UNION (SELECT id FROM tb_article)
|
||
UNION ALL (SELECT id FROM tb_article)
|
||
```
|
||
|
||
|
||
## limit... offset
|
||
|
||
::: tip 提示
|
||
在 "limit... offset" 的示例中，MyBatis-Flex 能够自动识别当前数据库👍，并根据数据库的类型生成不同的 SQL，用户也可以很轻易的通过 DialectFactory 注册（新增或改写）自己的实现方言。
|
||
:::
|
||
|
||
|
||
```java
|
||
QueryWrapper queryWrapper = QueryWrapper.create()
|
||
.select()
|
||
.from(ACCOUNT)
|
||
.orderBy(ACCOUNT.ID.desc())
|
||
.limit(10)
|
||
.offset(20);
|
||
```
|
||
|
||
MySQL 下执行的代码如下：
|
||
```sql
|
||
SELECT * FROM `tb_account` ORDER BY `id` DESC LIMIT 20, 10
|
||
```
|
||
|
||
PostgreSQL 下执行的代码如下：
|
||
```sql
|
||
SELECT * FROM "tb_account" ORDER BY "id" DESC LIMIT 20 OFFSET 10
|
||
```
|
||
Informix 下执行的代码如下：
|
||
```sql
|
||
SELECT SKIP 20 FIRST 10 * FROM "tb_account" ORDER BY "id" DESC
|
||
```
|
||
|
||
Oracle 下执行的代码如下：
|
||
```sql
|
||
SELECT * FROM (SELECT TEMP_DATAS.*,
|
||
ROWNUM RN FROM (
|
||
SELECT * FROM "tb_account" ORDER BY "id" DESC)
|
||
TEMP_DATAS WHERE ROWNUM <=30)
|
||
WHERE RN >20
|
||
```
|
||
|
||
Db2 下执行的代码如下：
|
||
```sql
|
||
SELECT * FROM "tb_account" ORDER BY "id" DESC
|
||
OFFSET 20 ROWS FETCH NEXT 10 ROWS ONLY
|
||
```
|
||
|
||
Sybase 下执行的代码如下：
|
||
```sql
|
||
SELECT TOP 10 START AT 21 * FROM "tb_account" ORDER BY "id" DESC
|
||
```
|
||
|
||
Firebird 下执行的代码如下：
|
||
```sql
|
||
SELECT * FROM "tb_account" ORDER BY "id" DESC ROWS 20 TO 30
|
||
```
|
||
|
||
## Lambda 扩展
|
||
|
||
虽然 MyBaits-Flex 也支持 lambda 方式，但是并不推荐使用，建议在一些简单的场景下使用，以下是示例：
|
||
|
||
简单示例：
|
||
```java
|
||
QueryWrapper query = QueryWrapper.create();
|
||
query.where(Account::getId).ge(100)
|
||
.and(Account::getUserName).like("michael")
|
||
.or(Account::getUserName).like(" ", If::hasText);
|
||
System.out.println(query.toSQL());
|
||
```
|
||
SQL 输入内容如下：
|
||
|
||
```sql
|
||
SELECT * FROM WHERE `id` >= 100 AND `user_name` LIKE '%michael%'
|
||
```
|
||
|
||
稍微复杂点的示例：
|
||
|
||
```java
|
||
QueryWrapper query = QueryWrapper.create()
|
||
.from(Article.class)
|
||
.leftJoin(Account.class).as("a").on(
|
||
wrapper -> wrapper.where(Account::getId).eq(Article::getAccountId)
|
||
)
|
||
.where(Account::getId).ge(100, If::notEmpty)
|
||
.and(wrapper -> {
|
||
wrapper.where(Account::getId).ge(100)
|
||
.or(Account::getAge).gt(200)
|
||
.and(Article::getAccountId).eq(200)
|
||
.or(wrapper1 -> {
|
||
wrapper1.where(Account::getId).like("a", If::notEmpty);
|
||
})
|
||
;
|
||
});
|
||
System.out.println(query.toSQL());
|
||
```
|
||
SQL 输入内容如下：
|
||
|
||
```sql
|
||
SELECT * FROM `tb_article`
|
||
LEFT JOIN `tb_account` AS `a` ON `a`.`id` = `tb_article`.`account_id`
|
||
WHERE `a`.`id` >= 100 AND
|
||
(`a`.`id` >= 100
|
||
OR `a`.`age` > 200
|
||
AND `tb_article`.`account_id` = 200
|
||
OR (`a`.`id` LIKE '%a%' )
|
||
)
|
||
```
|
||
|
||
|
||
## QueryWrapper 序列化
|
||
|
||
在 `QueryWrapper` 中，由于其定义了 `循环引用` 的一些数据结构，同时，其很多属性都是 `private` 或者 `protected` 修饰且没有 `getter` `setter` 方法，
|
||
这会导致使用一些 json 库在序列化的过程中，出现问题；但这些问题并非 `QueryWrapper` 的问题，而是序列化框架的问题。
|
||
|
||
因此，我们在使用序列化框架时，需要注意其是否支持这些特征，比如在使用 FastJson2 序列化时，需要添加以下配置：
|
||
|
||
序列化：
|
||
|
||
```java
|
||
String json = JSON.toJSONString(queryWrapper
|
||
, JSONWriter.Feature.FieldBased // 基于 field，而非 getter 方法
|
||
, JSONWriter.Feature.ReferenceDetection);
|
||
```
|
||
|
||
反序列化：
|
||
|
||
```java
|
||
QueryWrapper query = JSON.parseObject(json, QueryWrapper.class
|
||
, JSONReader.Feature.FieldBased);
|
||
```
|
||
|
||
`Gson` 、`Jackson` 等其他框架需要自行参考其文档添加相关配置；另外，我们更加建议使用专业的序列化工具去进行序列化，而非使用 json，比如使用 `JDK Serial` 或者 `fst` 等。
|
||
以下是相关的序列化示例代码：
|
||
|
||
`JDK Serial` 序列化：
|
||
|
||
```java
|
||
QueryWrapper queryWrapper = ...
|
||
ByteArrayOutputStream bos = new ByteArrayOutputStream();
|
||
ObjectOutputStream oos = new ObjectOutputStream(bos);
|
||
oos.writeObject(queryWrapper);
|
||
```
|
||
|
||
`JDK Serial` 反序列化：
|
||
|
||
```java
|
||
byte[] bytes = ....
|
||
ByteArrayInputStream bis = new ByteArrayInputStream(bytes);
|
||
ObjectInputStream ois = new ObjectInputStream(bis);
|
||
QueryWrapper queryWrapper = (QueryWrapper) ois.readObject();
|
||
```
|
||
|
||
`fst` 序列化 和 反序列化：
|
||
|
||
```java
|
||
FSTConfiguration fst = FSTConfiguration.createDefaultConfiguration();
|
||
|
||
//序列化得到 bytes 进行存储或者传输
|
||
byte[] bytes = fst.asByteArray(wrapper);
|
||
|
||
//反序列化得到 QueryWrapper
|
||
QueryWrapper newWrapper = (QueryWrapper) fst.asObject(bytes);
|
||
```
|
||
|
||
|
||
|
||
## 特别注意事项!!!
|
||
在 QueryWrapper 的条件构建中，如果传入 null 值，则自动忽略该条件，这有许多的好处，不需要额外的通过 `when()` 方法判断。但是也带来一些额外的知识记忆点，
|
||
因此，正对这一点需要特别注意一下。
|
||
|
||
例如：
|
||
|
||
```java
|
||
String userName = null;
|
||
Integer id = null;
|
||
QueryWrapper query1 = QueryWrapper.create()
|
||
.where(ACCOUNT.AGE.ge(18))
|
||
.and(ACCOUNT.USER_NAME.like(userName))
|
||
.and(ACCOUNT.ID.ge(id));
|
||
|
||
QueryWrapper query2 = QueryWrapper.create()
|
||
.where(ACCOUNT.AGE.ge(18));
|
||
```
|
||
在以上的 `query1` 中，由于 `userName` 和 `id` 都为 null，MyBatis-Flex 会自动忽略 null 值的条件，因此，它们构建出来的 SQL 条件是和 `query2` 完全一致的 。
|
||
|
||
|
||
## 存在疑问？
|
||
|
||
**疑问1：示例代码中的 QueryWrapper 所需要的 "ACCOUNT" 从哪里来的？**
|
||
|
||
答：MyBatis-Flex 使用了 APT（Annotation Processing Tool）在项目编译的时候，
|
||
会自动根据 Entity 类定义的字段生成 "ACCOUNT" 类以及 Entity 对应的 Mapper 类， 通过开发工具构建项目（如下图），
|
||
或者执行 maven 编译命令: `mvn clean package` 都可以自动生成。这个原理和 lombok 一致。
|
||
|
||
|
||
|
||
> 更多关于 APT 的配置，请进入 [APT 配置章节](../others/apt.md) 了解。
|
||
|