服务器之家:专注于服务器技术及软件下载分享
分类导航

PHP教程|ASP.NET教程|JAVA教程|ASP教程|编程技术|正则表达式|C/C++|IOS|C#|Swift|Android|JavaScript|易语言|

服务器之家 - 编程语言 - JAVA教程 - 通过Spring Boot + Mybatis + Redis快速搭建现代化Web项目

通过Spring Boot + Mybatis + Redis快速搭建现代化Web项目

2021-02-27 14:26roncoo JAVA教程

本篇文章介绍了如何通过Spring Boot、Mybatis以及Redis快速搭建一个现代化的Web项目,并且同时介绍了如何在Spring Boot下优雅地书写单元测试来保证我们的代码质量。具体内容详情大家通过本文学习下吧

背景

springboot因其提供了各种开箱即用的插件,使得它成为了当今最为主流的java web开发框架之一。mybatis是一个十分轻量好用的orm框架。redis是当今十分主流的分布式key-value型数据库,在web开发中,我们常用它来缓存数据库的查询结果。

本篇博客将介绍如何使用springboot快速搭建一个web应用,并且采用mybatis作为我们的orm框架。为了提升性能,我们将redis作为mybatis的二级缓存。为了测试我们的代码,我们编写了单元测试,并且用h2内存数据库来生成我们的测试数据。通过该项目,我们希望读者可以快速掌握现代化java web开发的技巧以及最佳实践。

本文的示例代码可在github中下载:https://github.com/lovelcp/spring-boot-mybatis-with-redis/tree/master

环境

开发环境:mac 10.11
ide:intellij 2017.1
jdk:1.8
spring-boot:1.5.3.release
redis:3.2.9
mysql:5.7

spring-boot

新建项目

首先,我们需要初始化我们的spring-boot工程。通过intellij的spring initializer,新建一个spring-boot工程变得十分简单。首先我们在intellij中选择new一个project:

通过Spring Boot + Mybatis + Redis快速搭建现代化Web项目

然后在选择依赖的界面,勾选web、mybatis、redis、mysql、h2:

通过Spring Boot + Mybatis + Redis快速搭建现代化Web项目

新建工程成功之后,我们可以看到项目的初始结构如下图所示:

通过Spring Boot + Mybatis + Redis快速搭建现代化Web项目

spring initializer已经帮我们自动生成了一个启动类——springbootmybatiswithredisapplication。该类的代码十分简单:

?
1
2
3
4
5
6
@springbootapplication
public class springbootmybatiswithredisapplication {
 public static void main(string[] args) {
 springapplication.run(springbootmybatiswithredisapplication.class, args);
 }
}

@springbootapplication注解表示启用spring boot的自动配置特性。好了,至此我们的项目骨架已经搭建成功,感兴趣的读者可以通过intellij启动看看效果。

新建api接口

接下来,我们要编写web api。假设我们的web工程负责处理商家的产品(product)。我们需要提供根据product id返回product信息的get接口和更新product信息的put接口。首先我们定义product类,该类包括产品id,产品名称name以及价格price:

?
1
2
3
4
5
6
7
public class product implements serializable {
 private static final long serialversionuid = 1435515995276255188l;
 private long id;
 private string name;
 private long price;
 // getters setters
}

然后我们需要定义controller类。由于spring boot内部使用spring mvc作为它的web组件,所以我们可以通过注解的方式快速开发我们的接口类:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
@restcontroller
@requestmapping("/product")
public class productcontroller {
 @getmapping("/{id}")
 public product getproductinfo(
   @pathvariable("id")
     long productid) {
  // todo
  return null;
 }
 @putmapping("/{id}")
 public product updateproductinfo(
   @pathvariable("id")
     long productid,
   @requestbody
     product newproduct) {
  // todo
  return null;
 }
}

我们简单介绍一下上述代码中所用到的注解的作用:

@restcontroller:表示该类为controller,并且提供rest接口,即所有接口的值以json格式返回。该注解其实是@controller和@responsebody的组合注解,便于我们开发rest api。

@requestmapping、@getmapping、@putmapping:表示接口的url地址。标注在类上的@requestmapping注解表示该类下的所有接口的url都以/product开头。@getmapping表示这是一个get http接口,@putmapping表示这是一个put http接口。

@pathvariable、@requestbody:表示参数的映射关系。假设有个get请求访问的是/product/123,那么该请求会由getproductinfo方法处理,其中url里的123会被映射到productid中。同理,如果是put请求的话,请求的body会被映射到newproduct对象中。

这里我们只定义了接口,实际的处理逻辑还未完成,因为product的信息都存在数据库中。接下来我们将在项目中集成mybatis,并且与数据库做交互。

集成mybatis

配置数据源

首先我们需要在配置文件中配置我们的数据源。我们采用mysql作为我们的数据库。这里我们采用yaml作为我们配置文件的格式。我们在resources目录下新建application.yml文件:

?
1
2
3
4
5
6
7
spring:
# 数据库配置
 datasource:
 url: jdbc:mysql://{your_host}/{your_db}
 username: {your_username}
 password: {your_password}
 driver-class-name: org.gjt.mm.mysql.driver

由于spring boot拥有自动配置的特性,我们不用新建一个datasource的配置类,sping boot会自动加载配置文件并且根据配置文件的信息建立数据库的连接池,十分便捷。

笔者推荐大家采用yaml作为配置文件的格式。xml显得冗长,properties没有层级结构,yaml刚好弥补了这两者的缺点。这也是spring boot默认就支持yaml格式的原因。

配置mybatis

我们已经通过spring initializer在pom.xml中引入了mybatis-spring-boot-starte库,该库会自动帮我们初始化mybatis。首先我们在application.yml中填写mybatis的相关配置:

?
1
2
3
4
5
6
7
# mybatis配置
mybatis:
 # 配置映射类所在包名
 type-aliases-package: com.wooyoo.learning.dao.domain
 # 配置mapper xml文件所在路径,这里是一个数组
 mapper-locations:
 - mappers/productmapper.xml

然后,再在代码中定义productmapper类:

?
1
2
3
4
5
6
7
@mapper
public interface productmapper {
 product select(
   @param("id")
     long id);
 void update(product product);
}

这里,只要我们加上了@mapper注解,spring boot在初始化mybatis时会自动加载该mapper类。

spring boot之所以这么流行,最大的原因是它自动配置的特性。开发者只需要关注组件的配置(比如数据库的连接信息),而无需关心如何初始化各个组件,这使得我们可以集中精力专注于业务的实现,简化开发流程。

访问数据库

完成了mybatis的配置之后,我们就可以在我们的接口中访问数据库了。我们在productcontroller下通过@autowired引入mapper类,并且调用对应的方法实现对product的查询和更新操作,这里我们以查询接口为例:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
@restcontroller
@requestmapping("/product")
public class productcontroller {
 @autowired
 private productmapper productmapper;
 @getmapping("/{id}")
 public product getproductinfo(
   @pathvariable("id")
     long productid) {
  return productmapper.select(productid);
 }
 // 避免篇幅过长,省略updateproductinfo的代码
}

然后在你的mysql中插入几条product的信息,就可以运行该项目看看是否能够查询成功了。

至此,我们已经成功地在项目中集成了mybatis,增添了与数据库交互的能力。但是这还不够,一个现代化的web项目,肯定会上缓存加速我们的数据库查询。接下来,将介绍如何科学地将redis集成到mybatis的二级缓存中,实现数据库查询的自动缓存。

集成redis

配置redis

同访问数据库一样,我们需要配置redis的连接信息。在application.yml文件中增加如下配置:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
spring:
 redis:
 # redis数据库索引(默认为0),我们使用索引为3的数据库,避免和其他数据库冲突
 database: 3
 # redis服务器地址(默认为localhost)
 host: localhost
 # redis端口(默认为6379
 port: 6379
 # redis访问密码(默认为空)
 password:
 # redis连接超时时间(单位为毫秒)
 timeout: 0
 # redis连接池配置
 pool:
  # 最大可用连接数(默认为8,负数表示无限)
  max-active: 8
  # 最大空闲连接数(默认为8,负数表示无限)
  max-idle: 8
  # 最小空闲连接数(默认为0,该值只有为正数才有作用)
  min-idle: 0
  # 从连接池中获取连接最大等待时间(默认为-1,单位为毫秒,负数表示无限)
  max-wait: -1

上述列出的都为常用配置,读者可以通过注释信息了解每个配置项的具体作用。由于我们在pom.xml中已经引入了spring-boot-starter-data-redis库,所以spring boot会帮我们自动加载redis的连接,具体的配置类
org.springframework.boot.autoconfigure.data.redis.redisautoconfiguration。通过该配置类,我们可以发现底层默认使用jedis库,并且提供了开箱即用的redistemplate和stringtemplate。

将redis作为二级缓存

mybatis的二级缓存原理本文不再赘述,读者只要知道,mybatis的二级缓存可以自动地对数据库的查询做缓存,并且可以在更新数据时同时自动地更新缓存。

实现mybatis的二级缓存很简单,只需要新建一个类实现org.apache.ibatis.cache.cache接口即可。

该接口共有以下五个方法:

string getid():mybatis缓存操作对象的标识符。一个mapper对应一个mybatis的缓存操作对象。

void putobject(object key, object value):将查询结果塞入缓存。

object getobject(object key):从缓存中获取被缓存的查询结果。

object removeobject(object key):从缓存中删除对应的key、value。只有在回滚时触发。一般我们也可以不用实现,具体使用方式请参考:org.apache.ibatis.cache.decorators.transactionalcache。

void clear():发生更新时,清除缓存。

int getsize():可选实现。返回缓存的数量。

readwritelock getreadwritelock():可选实现。用于实现原子性的缓存操作。

接下来,我们新建rediscache类,实现cache接口:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
public class rediscache implements cache {
 private static final logger logger = loggerfactory.getlogger(rediscache.class);
 private final readwritelock readwritelock = new reentrantreadwritelock();
 private final string id; // cache instance id
 private redistemplate redistemplate;
 private static final long expire_time_in_minutes = 30; // redis过期时间
 public rediscache(string id) {
  if (id == null) {
   throw new illegalargumentexception("cache instances require an id");
  }
  this.id = id;
 }
 @override
 public string getid() {
  return id;
 }
 /**
  * put query result to redis
  *
  * @param key
  * @param value
  */
 @override
 @suppresswarnings("unchecked")
 public void putobject(object key, object value) {
  redistemplate redistemplate = getredistemplate();
  valueoperations opsforvalue = redistemplate.opsforvalue();
  opsforvalue.set(key, value, expire_time_in_minutes, timeunit.minutes);
  logger.debug("put query result to redis");
 }
 /**
  * get cached query result from redis
  *
  * @param key
  * @return
  */
 @override
 public object getobject(object key) {
  redistemplate redistemplate = getredistemplate();
  valueoperations opsforvalue = redistemplate.opsforvalue();
  logger.debug("get cached query result from redis");
  return opsforvalue.get(key);
 }
 /**
  * remove cached query result from redis
  *
  * @param key
  * @return
  */
 @override
 @suppresswarnings("unchecked")
 public object removeobject(object key) {
  redistemplate redistemplate = getredistemplate();
  redistemplate.delete(key);
  logger.debug("remove cached query result from redis");
  return null;
 }
 /**
  * clears this cache instance
  */
 @override
 public void clear() {
  redistemplate redistemplate = getredistemplate();
  redistemplate.execute((rediscallback) connection -> {
   connection.flushdb();
   return null;
  });
  logger.debug("clear all the cached query result from redis");
 }
 @override
 public int getsize() {
  return 0;
 }
 @override
 public readwritelock getreadwritelock() {
  return readwritelock;
 }
 private redistemplate getredistemplate() {
  if (redistemplate == null) {
   redistemplate = applicationcontextholder.getbean("redistemplate");
  }
  return redistemplate;
 }
}

讲解一下上述代码中一些关键点:

自己实现的二级缓存,必须要有一个带id的构造函数,否则会报错。

我们使用spring封装的redistemplate来操作redis。网上所有介绍redis做二级缓存的文章都是直接用jedis库,但是笔者认为这样不够spring style,而且,redistemplate封装了底层的实现,未来如果我们不用jedis了,我们可以直接更换底层的库,而不用修改上层的代码。更方便的是,使用redistemplate,我们不用关心redis连接的释放问题,否则新手很容易忘记释放连接而导致应用卡死。

需要注意的是,这里不能通过autowire的方式引用redistemplate,因为rediscache并不是spring容器里的bean。所以我们需要手动地去调用容器的getbean方法来拿到这个bean,具体的实现方式请参考github中的代码。

我们采用的redis序列化方式是默认的jdk序列化。所以数据库的查询对象(比如product类)需要实现serializable接口。
这样,我们就实现了一个优雅的、科学的并且具有spring style的redis缓存类。

开启二级缓存

接下来,我们需要在productmapper.xml中开启二级缓存:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
<?xml version="1.0" encoding="utf-8" ?>
<!doctype mapper
  public "-//mybatis.org//dtd mapper 3.0//en"
  "http://mybatis.org/dtd/mybatis-3-mapper.dtd">
<mapper namespace="com.wooyoo.learning.dao.mapper.productmapper">
 <!-- 开启基于redis的二级缓存 -->
 <cache type="com.wooyoo.learning.util.rediscache"/>
 <select id="select" resulttype="product">
  select * from products where id = #{id} limit 1
 </select>
 <update id="update" parametertype="product" flushcache="true">
  update products set name = #{name}, price = #{price} where id = #{id} limit 1
 </update>
</mapper>

<cache type="com.wooyoo.learning.util.rediscache"/>表示开启基于redis的二级缓存,并且在update语句中,我们设置flushcache为true,这样在更新product信息时,能够自动失效缓存(本质上调用的是clear方法)。

测试

配置h2内存数据库

至此我们已经完成了所有代码的开发,接下来我们需要书写单元测试代码来测试我们代码的质量。我们刚才开发的过程中采用的是mysql数据库,而一般我们在测试时经常采用的是内存数据库。这里我们使用h2作为我们测试场景中使用的数据库。

要使用h2也很简单,只需要跟使用mysql时配置一下即可。在application.yml文件中:

?
1
2
3
4
5
6
7
8
9
10
11
---
spring:
 profiles: test
 # 数据库配置
 datasource:
 url: jdbc:h2:mem:test
 username: root
 password: 123456
 driver-class-name: org.h2.driver
 schema: classpath:schema.sql
 data: classpath:data.sql

为了避免和默认的配置冲突,我们用---另起一段,并且用profiles: test表明这是test环境下的配置。然后只要在我们的测试类中加上@activeprofiles(profiles = "test")注解来启用test环境下的配置,这样就能一键从mysql数据库切换到h2数据库。

在上述配置中,schema.sql用于存放我们的建表语句,data.sql用于存放insert的数据。这样当我们测试时,h2就会读取这两个文件,初始化我们所需要的表结构以及数据,然后在测试结束时销毁,不会对我们的mysql数据库产生任何影响。这就是内存数据库的好处。另外,别忘了在pom.xml中将h2的依赖的scope设置为test。

使用spring boot就是这么简单,无需修改任何代码,轻松完成数据库在不同环境下的切换。

编写测试代码

因为我们是通过spring initializer初始化的项目,所以已经有了一个测试类——springbootmybatiswithredisapplicationtests。

spring boot提供了一些方便我们进行web接口测试的工具类,比如testresttemplate。然后在配置文件中我们将log等级调成debug,方便观察调试日志。具体的测试代码如下:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
@runwith(springrunner.class)
@springboottest(webenvironment = springboottest.webenvironment.random_port)
@activeprofiles(profiles = "test")
public class springbootmybatiswithredisapplicationtests {
 @localserverport
 private int port;
 @autowired
 private testresttemplate resttemplate;
 @test
 public void test() {
  long productid = 1;
  product product = resttemplate.getforobject("http://localhost:" + port + "/product/" + productid, product.class);
  assertthat(product.getprice()).isequalto(200);
  product newproduct = new product();
  long newprice = new random().nextlong();
  newproduct.setname("new name");
  newproduct.setprice(newprice);
  resttemplate.put("http://localhost:" + port + "/product/" + productid, newproduct);
  product testproduct = resttemplate.getforobject("http://localhost:" + port + "/product/" + productid, product.class);
  assertthat(testproduct.getprice()).isequalto(newprice);
 }
}

在上述测试代码中:

我们首先调用get接口,通过assert语句判断是否得到了预期的对象。此时该product对象会存入redis中。

然后我们调用put接口更新该product对象,此时redis缓存会失效。

最后我们再次调用get接口,判断是否获取到了新的product对象。如果获取到老的对象,说明缓存失效的代码执行失败,代码存在错误,反之则说明我们代码是ok的。

书写单元测试是一个良好的编程习惯。虽然会占用你一定的时间,但是当你日后需要做一些重构工作时,你就会感激过去写过单元测试的自己。

查看测试结果

我们在intellij中点击执行测试用例,测试结果如下:

通过Spring Boot + Mybatis + Redis快速搭建现代化Web项目

显示的是绿色,说明测试用例执行成功了。

总结

本篇文章介绍了如何通过spring boot、mybatis以及redis快速搭建一个现代化的web项目,并且同时介绍了如何在spring boot下优雅地书写单元测试来保证我们的代码质量。当然这个项目还存在一个问题,那就是mybatis的二级缓存只能通过flush整个db来实现缓存失效,这个时候可能会把一些不需要失效的缓存也给失效了,所以具有一定的局限性。

原文链接:http://www.roncoo.com/article/detail/131302

延伸 · 阅读

精彩推荐