Question

我正在尝试为数据库表正确实现搜索功能。我尝试过这种方法：

控制器：

    @GetMapping
    public Page<TransactionDTO> find(TransactionFilterDTO filter, Pageable page) {
        return searchRepository
                .findTransactionsByFilter(mapper.toFilter(filter), page)
                .map(mapper::toDTO);
    }

Filer DTO：

public class TransactionFilterDTO {

    private String name;

    private Integer id;

    @DateTimeFormat(iso = DateTimeFormat.ISO.DATE_TIME)
    private LocalDateTime from;

    @DateTimeFormat(iso = DateTimeFormat.ISO.DATE_TIME)
    private LocalDateTime to;
    ... // getters and setter
}

搜索实现：

@Repository
public class TransactionSearchRepositoryImpl implements TransactionSearchRepository {

    @Autowired
    private TransactionRepository transactionRepository;

    @Autowired
    private TransactionSpecification specification;

    @Override
    public Page<Transaction> findTransactionsByFilter(TransactionFilter filter, @Nullable Pageable page) {


        List<Transaction> transactions = transactionRepository
                .findAll(specification.getFilter(filter));

        int totalCount = transactions.size();

        if(page != null) {
           transactions = transactions
                   .stream()
                   .skip(page.getOffset())
                   .limit(page.getPageSize())
                   .collect(Collectors.toList());
        }

        return new PageImpl<>(transactions, page, totalCount);
    }
}

存储库：

public interface TransactionSearchRepository {

    Page<Transaction> findTransactionsByFilter(TransactionFilter filter, Pageable page);
}

是否有更好的方法来实现搜索功能？我认为这种解决方案非常难看。

Answer 1

我可以推荐几种解决此任务的方法：

1）简单但不灵活：只需根据您的过滤器属性（name，from，to）在控制器中使用请求参数，并在存储库中准备相应的查询，例如：

控制器：

@GetMapping("/q")
public List<ResponseDto> getAllByQuery(
    @RequestParam(value = "name", required = false) String name,
    @RequestParam(value = "from", required = false) @DateTimeFormat(iso = ISO.DATE) LocalDate from,
    @RequestParam(value = "to", required = false) @DateTimeFormat(iso = ISO.DATE) LocalDate to,
    Pageable pageable
) {
    return service.getByQuery(name, from, to, pageable);
}

服务：

public Page<ResponseDto> getByQuery(String name, LocalDate from, LocalDate to, Pageable pageable) {
    return repo.getByQuery(name, from, to, pageable).map(mapper::toResponseDto);
}

存储库：

@Query("select m from MyEntity m where " +
       "(?1 is null or upper(m.name) like concat('%', upper(?1), '%')) " +
       "and (?2 is null or m.createdAt >= ?2) " +
       "and (?3 is null or m.createdAt <= ?3)")
Page<MyEntity> getByQuery(String name, final LocalDate from, final LocalDate to, final Pageable pageable);

然后执行请求：

GET http://localhost:8080/q?name=john&from=2019-04-19&to=2019-04-19

2）使用 QueryDsl 。您应该将其添加到项目中（可以找到详细信息here），从QuerydslPredicateExecutor和QuerydslBinderCustomizer扩展存储库，并向其中添加一些“调整”：

public interface MyEntityRepo extends JpaRepository<MyEntity, Integer>, QuerydslPredicateExecutor<MyEntity>, QuerydslBinderCustomizer<QMyEntity> {
    @Override
    default void customize(@NonNull QuerydslBindings bindings, @NonNull QMyEntity entity) {

        // Make case-insensitive 'like' filter for all string properties
        bindings.bind(String.class).first((SingleValueBinding<StringPath, String>) StringExpression::containsIgnoreCase);

        // Add 'between' and 'greater or equal' filter date property
        bindings.bind(entity.createdAt).all((path, value) -> {
            Iterator<? extends LocalDate> it = value.iterator();
            LocalDate from = it.next();
            if (value.size() >= 2) {
                LocalDate to = it.next();
                return Optional.of(path.between(from, to)); // between
            } else {
                return Optional.of(path.goe(from)); // greater than or equal
            }
        });
    }

添加服务方法：

public Page<ResponseDto> getAllByQueryDsl(Predicate predicate, Pageable pageable) {
    return repo.findAll(predicate, pageable).map(mapper::toResponseDto);
}

添加控制器方法：

@GetMapping("/query-dsl")
public Page<ResponseDto> getAllByQueryDsl(
        @QuerydslPredicate(root = MyEntity.class, bindings = MyEntityRepo.class) Predicate predicate,
        Pageable pageable
) {
    return service.getAllByQueryDsl(predicate, pageable);
}

并在实体的“日期”属性中添加@DateTimeFormat注释：

@Entity
public class MyEntity {
    // ...
    @DateTimeFormat(iso = ISO.DATE) private LocalDate createdAt;
}

然后您可以执行这样的请求：

GET http://localhost:8080/query-dsl?name=john&createdAt=2019-04-15&createdAt=2019-04-19

第一个日期是“ from”参数，第二个日期是“ to”参数。如果您仅使用一个日期-它将是'from'参数（大于或等于）。

3）使用 specification-arg-resolver 库。将其添加到您的项目中（请参阅说明：1和2），然后从JpaSpecificationExecutor扩展存储库：

public interface MyEntityRepo extends JpaRepository<MyEntity, Integer>, JpaSpecificationExecutor<MyEntity> {}

将这样的方法添加到您的控制器中：

@GetMapping("/specification")
public Page<ResponseDto> getAllBySpecification(
        @And({
                @Spec(path = "name", spec = LikeIgnoreCase.class),
                @Spec(path = "createdAt", params = "from", spec = GreaterThanOrEqual.class),
                @Spec(path = "createdAt", params = "to", spec = LessThanOrEqual.class)
        }) Specification<MyEntity> specification,
        Pageable pageable
) {
    return service.getAllBySpecification(specification, pageable);
}

更新您的服务：

public Page<ResponseDto> getAllBySpecification(final Specification<MyEntity> specification, final Pageable pageable) {
    return repo.findAll(specification, pageable).map(mapper::toResponseDto);
}

然后请求您的数据：

GET http://localhost:8080/specification?name=john&from=2019-04-10&to=2019-04-19

4）手动建立规范：

创建过滤器类：

@Data
public class MyFilter implements Specification<MyEntity> {
    private String name;
    @DateTimeFormat(iso = ISO.DATE) private LocalDate from;
    @DateTimeFormat(iso = ISO.DATE) private LocalDate to;

    @Override
    public Predicate toPredicate(Root<MyEntity> root, CriteriaQuery<?> query, CriteriaBuilder builder) {
        List<Predicate> predicates = new ArrayList<>();
        if (name != null) predicates.add(builder.like(builder.upper(root.get("name")), "%" + name.toUpperCase() + "%"));
        if (from != null) predicates.add(builder.greaterThanOrEqualTo(root.get("createdAt"), from));
        if (to != null) predicates.add(builder.lessThanOrEqualTo(root.get("createdAt"), to));
        return builder.and(predicates.toArray(new Predicate[0]));
    }
}

创建控制器方法：

@GetMapping("/filter")
public Page<ResponseDto> getAllByMyFilter(MyFilter filter, Pageable pageable) {
    return service.getAllBySpecification(filter, pageable);
}

然后运行请求：

GET http://localhost:8080/filter?name=john&from=2019-04-10&to=2019-04-19

Answer 2

看看query by example。您只需使用

就可以摆脱一些代码

transactionRepository.findAll(Example.of(transaction));

它还支持分页和更高级的API检查ExampleMatcher类。

Answer 3

有几种方法可以解决此问题。您也可以实现以下解决方案：例如在您的存储库中：

@Query(value="FROM Students AS student "
            + " LEFT JOIN FETCH student.school AS school"
            .......................
            ...................
            + " WHERE  LOWER(student.name) LIKE LOWER(CONCAT('%',:criteria, '%')) OR"
            + " LOWER(student.code) LIKE LOWER(CONCAT('%',:criteria, '%')) OR"
            .........................
            .........................
            .........................
            + " LOWER(school.name) LIKE LOWER(CONCAT('%',:criteria, '%')) "         
            + " ORDER BY student.name ASC",
            countQuery=" SELECT COUNT(student) from Students AS student ")
    public Page<Students> getAllStudentsBySearchCriteria(@Param("criteria")String search, Pageable pageable);

为您服务：

public Page<Students> getAll(Page  pagable) {
    return studentRepository.getAllStudentsBySearchCriteria(searchString, pageable);

}

为JPA和Spring Boot实施搜索功能

3 个答案: