Spring-Data-Jpa使用总结
- 2019 年 10 月 17 日
- 筆記
参考资源列表
- 官方文档：https://docs.spring.io/spring-data/jpa/docs/2.1.5.RELEASE/reference/html/
- 《Spring Data JPA入门到精通》
前言
JPA是Java Persistence API的简称，是Spring在Hibernate的基础上进行的二次封装框架，为了更好更方便的融入Spring大家庭，同时也提供了一些Hibernate没有的特性，与其他ORM框架一起构成SpringData，统一封装了ORM层，使开发人员使用起来更加方便快捷。
备注：本文所有代码都是基于SpringBoot 2.1.5版本
JPA的使用
基本单表操作的使用
对于操作单表来说，jpa提供了非常方便使用的封装，我们只需要按规范编写Repository接口同时继承JpaRepository就可以享用jpa基本功能了。代码如下：
User实体:
package com.yizhu.entity; import lombok.Builder; import lombok.Data; import lombok.ToString; import javax.persistence.*; import java.io.Serializable; import java.util.List; import java.util.Set; @Entity @Table(name = "t_user") @Data @Builder public class User implements Serializable { private static final long serialVersionUID = 1L; @Id @GeneratedValue(strategy = GenerationType.IDENTITY) private Long id; private String name; private String password; private Integer age; private Integer sex; } repository接口：
package com.yizhu.repository; import com.yizhu.entity.User; public interface UserRepository extends JpaRepository<User, Long>{ }下面看看jpa为我们提供了哪些默认操作单表数据的方法
package org.springframework.data.jpa.repository; @NoRepositoryBean public interface JpaRepository<T, ID> extends PagingAndSortingRepository<T, ID>, QueryByExampleExecutor<T> { List<T> findAll(); // 查询全表数据 List<T> findAll(Sort var1); // 查询全表数据，支持排序 List<T> findAllById(Iterable<ID> var1); // 根据id字段查询所有匹配数据 <S extends T> List<S> saveAll(Iterable<S> var1); // 批量保存或更新数据 void flush(); // 刷新本地缓存到数据库 <S extends T> S saveAndFlush(S var1); // 保存或更新单挑数据及刷新本地缓存到数据库 void deleteInBatch(Iterable<T> var1); // 批量删除数据 void deleteAllInBatch(); // 批量删除全表数据 T getOne(ID var1); // 根据id查询一条匹配数据 <S extends T> List<S> findAll(Example<S> ar1); // Example方式的查询指定实体 <S extends T> List<S> findAll(Example<S> var1, Sort var2); // Example方式的查询指定实体并排序 }package org.springframework.data.repository; @NoRepositoryBean public interface PagingAndSortingRepository<T, ID> extends CrudRepository<T, ID> { Iterable<T> findAll(Sort var1); // 根据排序查询全表，返回类型是任意集合 Page<T> findAll(Pageable var1); // 根据分页参数分页查询 }package org.springframework.data.repository; @NoRepositoryBean public interface CrudRepository<T, ID> extends Repository<T, ID> { <S extends T> S save(S var1); // 保存或更新单条数据 <S extends T> Iterable<S> saveAll(Iterable<S> var1); // 批量保存或更新 Optional<T> findById(ID var1); // 根据id查询数据，返回类型是Optional boolean existsById(ID var1); // 根据id判断数据是否存在 Iterable<T> findAll(); // 查询全表数据，返回类型为集合 Iterable<T> findAllById(Iterable<ID> var1); // 根据id集合查询数据 long count(); // 统计全表数据量 void deleteById(ID var1); // 根据id删除数据 void delete(T var1); // 删除单条数据 void deleteAll(Iterable<? extends T> var1); // 删除指定集合数据 void deleteAll(); // 删除全表数据 }package org.springframework.data.repository.query; public interface QueryByExampleExecutor<T> { <S extends T> Optional<S> findOne(Example<S> var1); // 根据Example查询一条 <S extends T> Iterable<S> findAll(Example<S> var1); // 根据Example查询所有数据 <S extends T> Iterable<S> findAll(Example<S> var1, Sort var2); // 根据Example查询所有数据，并排序 <S extends T> Page<S> findAll(Example<S> var1, Pageable var2); // 根据Example分页查询 <S extends T> long count(Example<S> var1); // 根据Example统计 <S extends T> boolean exists(Example<S> var1); // 根据Example判断数据是否存在 }除此之外，jpa提供了一套新的生成sql的机制，非常方便好用，jpa根据Repository接口的方法中的关键字、实体字段及出入参，自动生成sql，这种方式启动容器的时候就可以检查语法是否正确，简单使用例子如下：
package com.yizhu.repository; import com.yizhu .entity.User; import org.springframework.data.jpa.repository.EntityGraph; import org.springframework.data.jpa.repository.JpaRepository; import org.springframework.data.jpa.repository.JpaSpecificationExecutor; import org.springframework.data.jpa.repository.Query; import org.springframework.data.repository.query.Param; import java.util.List; public interface UserRepository extends JpaRepository<User, Long>{ /** * 根据年龄查询用户信息 * @param age * @return */ List<User> findAllByAge(Integer age); /** * 根据用户性别和所属组织名称查询用户信息 * @param userSex * @param orgName * @return */ List<User> findBySexAndOrg(@Param("sex") Integer sex, @Param("name") String name); /** * 根据用户名模糊查询 * @return */ List<User> findAllByNameLike(@Param("name") String name); }除了find 、By、And之外，还有一些关键词，全部定义在PartTree、Part类，组装起来可以生成各种各样的sql，下面截取部分代码，感兴趣的同学可以打开源码去阅读
package org.springframework.data.repository.query.parser; public class PartTree implements Streamable<PartTree.OrPart> { private static final String KEYWORD_TEMPLATE = "(%s)(?=(\p{Lu}|\P{InBASIC_LATIN}))"; private static final String QUERY_PATTERN = "find|read|get|query|stream"; private static final String COUNT_PATTERN = "count"; private static final String EXISTS_PATTERN = "exists"; private static final String DELETE_PATTERN = "delete|remove"; private static final Pattern PREFIX_TEMPLATE = Pattern.compile("^(find|read|get|query|stream|count|exists|delete|remove)((\p{Lu}.*?))??By"); private final PartTree.Subject subject; private final PartTree.Predicate predicate; ... private static String[] split(String text, String keyword) { Pattern pattern = Pattern.compile(String.format("(%s)(?=(\p{Lu}|\P{InBASIC_LATIN}))", keyword)); return pattern.split(text); } private static class Predicate implements Streamable<PartTree.OrPart> { private static final Pattern ALL_IGNORE_CASE = Pattern.compile("AllIgnor(ing|e)Case"); private static final String ORDER_BY = "OrderBy"; private final List<PartTree.OrPart> nodes; private final OrderBySource orderBySource; private boolean alwaysIgnoreCase; public Predicate(String predicate, Class<?> domainClass) { String[] parts = PartTree.split(this.detectAndSetAllIgnoreCase(predicate), "OrderBy"); if (parts.length > 2) { throw new IllegalArgumentException("OrderBy must not be used more than once in a method name!"); } else { this.nodes = (List)Arrays.stream(PartTree.split(parts[0], "Or")).filter(StringUtils::hasText).map((part) -> { return new PartTree.OrPart(part, domainClass, this.alwaysIgnoreCase); }).collect(Collectors.toList()); this.orderBySource = parts.length == 2 ? new OrderBySource(parts[1], Optional.of(domainClass)) : OrderBySource.EMPTY; } } ... } private static class Subject { private static final String DISTINCT = "Distinct"; private static final Pattern COUNT_BY_TEMPLATE = Pattern.compile("^count(\p{Lu}.*?)??By"); private static final Pattern EXISTS_BY_TEMPLATE = Pattern.compile("^(exists)(\p{Lu}.*?)??By"); private static final Pattern DELETE_BY_TEMPLATE = Pattern.compile("^(delete|remove)(\p{Lu}.*?)??By"); private static final String LIMITING_QUERY_PATTERN = "(First|Top)(\d*)?"; private static final Pattern LIMITED_QUERY_TEMPLATE = Pattern.compile("^(find|read|get|query|stream)(Distinct)?(First|Top)(\d*)?(\p{Lu}.*?)??By"); private final boolean distinct; private final boolean count; private final boolean exists; private final boolean delete; private final Optional<Integer> maxResults; public Subject(Optional<String> subject) { this.distinct = (Boolean)subject.map((it) -> { return it.contains("Distinct"); }).orElse(false); this.count = this.matches(subject, COUNT_BY_TEMPLATE); this.exists = this.matches(subject, EXISTS_BY_TEMPLATE); this.delete = this.matches(subject, DELETE_BY_TEMPLATE); this.maxResults = this.returnMaxResultsIfFirstKSubjectOrNull(subject); } private Optional<Integer> returnMaxResultsIfFirstKSubjectOrNull(Optional<String> subject) { return subject.map((it) -> { Matcher grp = LIMITED_QUERY_TEMPLATE.matcher(it); return !grp.find() ? null : StringUtils.hasText(grp.group(4)) ? Integer.valueOf(grp.group(4)) : 1; }); } ... private boolean matches(Optional<String> subject, Pattern pattern) { return (Boolean)subject.map((it) -> { return pattern.matcher(it).find(); }).orElse(false); } } }package org.springframework.data.repository.query.parser; public class Part { private static final Pattern IGNORE_CASE = Pattern.compile("Ignor(ing|e)Case"); private final PropertyPath propertyPath; private final Part.Type type; private Part.IgnoreCaseType ignoreCase; ... public static enum Type { BETWEEN(2, new String[]{"IsBetween", "Between"}), IS_NOT_NULL(0, new String[]{"IsNotNull", "NotNull"}), IS_NULL(0, new String[]{"IsNull", "Null"}), LESS_THAN(new String[]{"IsLessThan", "LessThan"}), LESS_THAN_EQUAL(new String[]{"IsLessThanEqual", "LessThanEqual"}), GREATER_THAN(new String[]{"IsGreaterThan", "GreaterThan"}), GREATER_THAN_EQUAL(new String[]{"IsGreaterThanEqual", "GreaterThanEqual"}), BEFORE(new String[]{"IsBefore", "Before"}), AFTER(new String[]{"IsAfter", "After"}), NOT_LIKE(new String[]{"IsNotLike", "NotLike"}), LIKE(new String[]{"IsLike", "Like"}), STARTING_WITH(new String[]{"IsStartingWith", "StartingWith", "StartsWith"}), ENDING_WITH(new String[]{"IsEndingWith", "EndingWith", "EndsWith"}), IS_NOT_EMPTY(0, new String[]{"IsNotEmpty", "NotEmpty"}), IS_EMPTY(0, new String[]{"IsEmpty", "Empty"}), NOT_CONTAINING(new String[]{"IsNotContaining", "NotContaining", "NotContains"}), CONTAINING(new String[]{"IsContaining", "Containing", "Contains"}), NOT_IN(new String[]{"IsNotIn", "NotIn"}), IN(new String[]{"IsIn", "In"}), NEAR(new String[]{"IsNear", "Near"}), WITHIN(new String[]{"IsWithin", "Within"}), REGEX(new String[]{"MatchesRegex", "Matches", "Regex"}), EXISTS(0, new String[]{"Exists"}), TRUE(0, new String[]{"IsTrue", "True"}), FALSE(0, new String[]{"IsFalse", "False"}), NEGATING_SIMPLE_PROPERTY(new String[]{"IsNot", "Not"}), SIMPLE_PROPERTY(new String[]{"Is", "Equals"}); private static final List<Part.Type> ALL = Arrays.asList(IS_NOT_NULL, IS_NULL, BETWEEN, LESS_THAN, LESS_THAN_EQUAL, GREATER_THAN, GREATER_THAN_EQUAL, BEFORE, AFTER, NOT_LIKE, LIKE, STARTING_WITH, ENDING_WITH, IS_NOT_EMPTY, IS_EMPTY, NOT_CONTAINING, CONTAINING, NOT_IN, IN, NEAR, WITHIN, REGEX, EXISTS, TRUE, FALSE, NEGATING_SIMPLE_PROPERTY, SIMPLE_PROPERTY); public static final Collection<String> ALL_KEYWORDS; private final List<String> keywords; private final int numberOfArguments; ... static { List<String> allKeywords = new ArrayList(); Iterator var1 = ALL.iterator(); while(var1.hasNext()) { Part.Type type = (Part.Type)var1.next(); allKeywords.addAll(type.keywords); } ALL_KEYWORDS = Collections.unmodifiableList(allKeywords); } } }可以看到单表操作的大部分需求jpa都为我们提供了现成的实现，但也支持我们使用@Query注解自定义查询sql，方便有sql基础的同学使用，sql可控性强。
package com.yizhu.repository; import com.yizhu .entity.User; import org.springframework.data.jpa.repository.EntityGraph; import org.springframework.data.jpa.repository.JpaRepository; import org.springframework.data.jpa.repository.JpaSpecificationExecutor; import org.springframework.data.jpa.repository.Query; import org.springframework.data.repository.query.Param; import java.util.List; public interface UserRepository extends JpaRepository<User, Long>{ /** * 查询所有用户信息 * @return */ @Query(value = "from User u") List<User> findAll(); /** * 根据年龄查询用户信息 * @param age * @return */ @Query(value = "select * from t_user u where u.user_age = ?1", nativeQuery = true) List<User> findAllByAge(Integer age); /** * 根据用户性别和所属组织名称查询用户信息 * @param userSex * @param orgName * @return */ @Query(value = "select u from User u left join u.org o where u.userSex = :userSex and o.orgName = :orgName") List<User> findUsersBySexAndOrg(@Param("userSex") Integer userSex, @Param("orgName") String orgName); }多表关联
@OneToOne、@OneToMany、@ManyToOne、@ManyToMany
@Entity @Table(name = "t_user") @NamedEntityGraph(name = "User.findUsers", attributeNodes = {@NamedAttributeNode("jobs"), @NamedAttributeNode("roles")}) @Data @Builder @NoArgsConstructor @AllArgsConstructor public class User implements Serializable { private static final long serialVersionUID = 1L; @ApiModelProperty(hidden = true) @Id @GeneratedValue(strategy = GenerationType.IDENTITY) private Long id; @ApiModelProperty(value = "用户名") @Column(name = "user_name") private String name; @ApiModelProperty(value = "用户密码") @Column(name = "user_password") private String password; @ApiModelProperty(value = "用户年龄") @Column(name = "user_age") private Integer age; @ApiModelProperty(value = "用户性别") @Column(name = "user_sex") private Integer sex; @ApiModelProperty(value = "所属组织id") @Column(name = "org_id") private Long orgId; @ApiModelProperty(value = "用户信息") @OneToOne @JoinColumn(name = "id", updatable = false, insertable = false) private UserInfo userInfo; @ApiModelProperty(value = "用户所属组织") @ManyToOne @JoinColumn(name = "org_id", updatable = false, insertable = false) private Organization org; @ApiModelProperty(value = "用户角色") @OneToMany @JoinColumn(name = "user_id", referencedColumnName = "id", insertable = false, updatable = false) @NotFound(action = NotFoundAction.IGNORE) private Set<Role> roles; @ApiModelProperty(value = "用户工作") @ManyToMany @JoinTable( name = "t_user_job", joinColumns = @JoinColumn(name = "user_id", referencedColumnName = "id"), inverseJoinColumns = @JoinColumn(name = "job_id", referencedColumnName = "id") ) @NotFound(action = NotFoundAction.IGNORE) private Set<Job> jobs;
private Set<Role> roles;和private Set<Job> jobs;不能用同时使用List集合代替，会报错org.hibernate.loader.MultipleBagFetchException: cannot simultaneously fetch multiple bags: [com.yizhu.entity.User.jobs, com.yizhu.entity.User.roles]
动态查询
package com.yizhu.repository; import com.yizhu.dto.UserQueryDto; import com.yizhu.entity.Organization; import com.yizhu.entity.User; import org.springframework.data.jpa.domain.Specification; import javax.persistence.criteria.Join; import javax.persistence.criteria.JoinType; import javax.persistence.criteria.Predicate; import java.util.ArrayList; import java.util.List; import java.util.Optional; public class UserSpecs { public static Specification<User> listQuerySpec(UserQueryDto userQueryDto){ return (root, query, builder) -> { List<Predicate> predicates = new ArrayList<>(); Optional.ofNullable(userQueryDto.getId()).ifPresent(i -> predicates.add(builder.equal(root.get("id"), i))); Optional.ofNullable(userQueryDto.getName()).ifPresent(n -> predicates.add(builder.equal(root.get("name"), n))); Optional.ofNullable(userQueryDto.getAge()).ifPresent(a -> predicates.add(builder.equal(root.get("age"), a))); Optional.ofNullable(userQueryDto.getOrgId()).ifPresent(oi -> predicates.add(builder.equal(root.get("orgId"), oi))); Optional.ofNullable(userQueryDto.getOrgName()).ifPresent(on -> { Join<User, Organization> userJoin = root.join(root.getModel().getSingularAttribute("org", Organization.class), JoinType.LEFT); predicates.add(builder.equal(userJoin.get("orgName"), on)); }); return builder.and(predicates.toArray(new Predicate[predicates.size()])); }; } }package com.yizhu.service; import com.yizhu.dto.UserQueryDto; import com.yizhu.entity.User; import com.yizhu.repository.UserRepository; import com.yizhu.repository.UserSpecs; import org.springframework.beans.factory.annotation.Autowired; import org.springframework.stereotype.Service; import java.util.List; @Service public class UserService { @Autowired private UserRepository userRepository; public List<User> findUsersDynamic(UserQueryDto userQueryDto){ return userRepository.findAll(UserSpecs.listQuerySpec(userQueryDto)); } }审计功能使用
在启动类添加@EnableJpaAuditing注解表示开启jpa审计功能
package com.yizhu; import org.springframework.boot.SpringApplication; import org.springframework.boot.autoconfigure.SpringBootApplication; import org.springframework.data.jpa.repository.config.EnableJpaAuditing; @EnableJpaAuditing @SpringBootApplication public class DemoApplication { public static void main(String[] args) { SpringApplication.run(DemoApplication.class, args); } }在需要使用审计功能的实体类添加@EntityListeners(AuditingEntityListener.class)注解
package com.yizhu.entity; import lombok.Builder; import lombok.Data; import org.springframework.data.jpa.domain.support.AuditingEntityListener; import javax.persistence.*; import java.io.Serializable; import java.util.Date; @Entity @Table(name = "t_role") @Data @Builder @EntityListeners(AuditingEntityListener.class) public class Role implements Serializable { private static final long serialVersionUID=1L; @ApiModelProperty(hidden = true) @Id @GeneratedValue(strategy = GenerationType.AUTO) private Long id; private String roleName; @CreatedDate private Date createTime; @CreatedBy private Long createId; @LastModifiedDate private Date updateTime; @LastModifiedBy private Long updateId; }实现AuditorAware接口，告诉容器当前登录人id
package com.yizhu.configuration; import org.springframework.data.domain.AuditorAware; import org.springframework.web.context.request.RequestContextHolder; import org.springframework.web.context.request.ServletRequestAttributes; import java.util.Optional; public class UserAuditorAwareImpl implements AuditorAware<Long> { @Override public Optional<Long> getCurrentAuditor() { // 从session中获取登录人id ServletRequestAttributes servletRequestAttributes = (ServletRequestAttributes) RequestContextHolder.getRequestAttributes(); Long userId = (Long)servletRequestAttributes.getRequest().getSession().getAttribute("userId"); return Optional.of(userId); } }ok，然后jpa就会根据ID和Version判断当前操作是更新还是新增数据，新增时会注入当前登录人id到标有@CreateBy注解的字段上，当前时间注入到标有@CreateTime注解字段上；更新时则注入到@LastModifiedBy和@LastModifiedDate对应的字段上。想详细了解的可查看org.springframework.data.jpa.domain.support.AuditingEntityListener源码。
常见的坑
- N+1问题，当使用
@ManyToMany、@ManyToOne、@OneToMany、@OneToOne关联
关系的时候，FetchType怎么配置LAZY或者EAGER。SQL真正执行的时
候是由一条主表查询和N条子表查询组成的。这种查询效率一般比较
低下，比如子对象有N个就会执行N+1条SQL。使用JPA 2.1推出来的@EntityGraph、@NamedEntityGraph可以解决该问题。如下。
@ApiModel @Entity @Table(name = "t_user") @NamedEntityGraph(name = "User.findUsers", attributeNodes = {@NamedAttributeNode("jobs"), @NamedAttributeNode("roles")}) @Data @Builder @NoArgsConstructor @AllArgsConstructor public class User implements Serializable { private static final long serialVersionUID = 1L; @ApiModelProperty(hidden = true) @Id @GeneratedValue(strategy = GenerationType.IDENTITY) private Long id; // 省略其他属性 }package com.yizhu.repository; import com.yizhu.entity.User; import org.springframework.data.jpa.repository.EntityGraph; import org.springframework.data.jpa.repository.JpaRepository; import org.springframework.data.jpa.repository.JpaSpecificationExecutor; import org.springframework.data.jpa.repository.Query; import org.springframework.data.repository.query.Param; import java.util.List; public interface UserRepository extends JpaRepository<User, Long>, JpaSpecificationExecutor<User> { /** * 根据id查询用户信息 * @param id * @return */ @EntityGraph(value = "User.findUsers", type = EntityGraph.EntityGraphType.FETCH) User findAllById(Long id); /** * 根据name查询用户信息 * @param name * @return */ @EntityGraph(value = "User.findUsers", type = EntityGraph.EntityGraphType.FETCH) @Query(value = "select * from t_user where user_name = :name", nativeQuery = true) List<User> findAllByUserName(@Param("name") String name); }- 所有的注解要么全配置在字段上，要么全配置在get方法上，不能混用，混用就会启动不起来，但是语法配置没有问题。
- 所有的关联都是支持单向关联和双向关联的，视具体业务场景而定。JSON序列化的时候使用双向注解会产生死循环，需要人为手动转化一次，或者使用@JsonIgnore。
- 在所有的关联查询中，表一般是不需要建立外键索引的。@mappedBy的使用需要注意。
- 级联删除比较危险，建议考虑清楚，或者完全掌握。
- 不同的关联关系的配置，@JoinClumn里面的name、referencedColumnName代表的意思是不一样的，很容易弄混，可以根据打印出来的SQL做调整。
- 当配置这些关联关系的时候建议大家直接在表上面，把外键建好，然后通过后面我们介绍的开发工具直接生成，这样可以减少自己调试的时间。
JPA常用注解
摘自《Spring Data JPA从入门到精通》
