java - Spring/Hibernate - 在 Java 中不使用 @Entity 从存储过程中获取结果

Question

我们在 MS SQL 中有一个存储过程，它返回一些与任何数据库表都不匹配的结果集。我们希望在不使用 @Entity 注释的情况下将结果映射到普通的 java POJO。但是如果我们打电话

entityManager.createStoredProcedureQuery("schema.myProcedure",
                        PlainPojo.class);

它将失败 org.hibernate.MappingException: Unknown entity异常(exception)

我知道有 @SqlResultSetMapping 注释，但它需要放在 Entity 类上？这不起作用

@Repository
@SqlResultSetMapping(name = "Mapping")
public class MyRepositoryImpl implements MyRepository

只有这个

@Enity
@SqlResultSetMapping(name = "Mapping")
public class MyEntity

但我们不想要实体类...所以把它放在另一个实体上？？ (不完全是)

可以把它放在其他地方吗？

或者有可能在代码中做到吗？

另一种解决方案？

是否可以制作类似 transient 实体的东西？ -> 使用@Entity 注释而不创建数据库表等

Answer 1

我没有在 stackoverflow 上找到复杂的解决方案并重新发明轮子，也许这对需要在项目中多次调用存储过程的人有所帮助。

主要思想是编写类似于@StoredProcedure的自定义注解，但独立于@Entity

@Target(TYPE)
@Retention(RUNTIME)
public @interface CustomNamedStoredProcedureQueries {
     /**
     * Array of CustomNamedStoredProcedureQuery annotations.
     */
     CustomNamedStoredProcedureQuery[] value();
}
@Target(TYPE)
@Retention(RUNTIME)
public @interface CustomNamedStoredProcedureQuery {
    /*
    * The name of the stored procedure for call.
    */
    String name();
    /*
    * The name of the stored procedure in the database.
    */
    String procedureName();
    /*
    * The scheme name of the database.
    */
    String schemeName();
    /*
    * The name of the package that contains stored procedure in the database.
    */
    String packageName();
    /*
    * Information about all parameters of the stored procedure.
    */
    CustomStoredProcedureParameter[] parameters() default {};
    /*
    * The names of one or more result set mappings, as defined in metadata, when you need custom mapping.
    */
    ProcedureRowMapper[] resultSetRowMappers() default {};
    /*
    * The names of one or more result set mappings, as defined in metadata, when you need 1 to 1 mapping.
    */
    ProcedureResultClass[] resultSetMappers() default {};
}
@Target({})
@Retention(RUNTIME)
public @interface CustomStoredProcedureParameter {
    /** Name of strored procedure parameter.*/
    String name() default "";
    /** JDBC type of the parameter.*/
    int type();
    /** Parameter mode.*/
    ParameterMode mode() default ParameterMode.IN;
}
@Target({})
@Retention(RUNTIME)
public @interface ProcedureResultClass {
    /* Name of procedure parameter*/
    String name();
    /* Class of object using for mapping */
    Class<?> resultClass();
}
@Target({})
@Retention(RUNTIME)
public @interface ProcedureRowMapper {
    /* Name of cursor field for use mapper*/
    String name();
    /* Class of object using for mapping, must extend BeanPropertyRowMapper */
    Class<? extends BeanPropertyRowMapper> mapper();
}

如前所述，当您需要将 resultSet 1 映射到 1 时，请使用 ProcedureResultClass，字段必须具有相同的名称。

当你需要一些自定义映射时使用 ProcedureRowMapper，你可以@Override mapRow() 方法，并手动映射它，或者您可以@Override initBeanWrapper() 如果您只需要以特定方式转换某些字段。

所以，接下来我们需要一个让它工作的方法。把它放在你的应用程序的 init 方法中，比如 main 方法的 @PostConstruct 。

ProjectUtils {
    private JdbcTemplate jdbcTemplate;

    @Autowired
    public ProjectUtils(DataSource dataSource) {
        jdbcTemplate = new JdbcTemplate(DataSource);
    }

    private static Map<String, SimpleJdbcCall> storedProcedureContainer = new HashMap<>();


    public void initStoredProcedure() {
        Reflections reflections = new Reflections("my.package");
        Set<Class<?>> classList = reflections.getTypesAnnotatedWith(CustomNamedStoredProcedureQueries.class);

        for (Class<?> clazz : classList) {
            CustomNamedStoredProcedureQueries queriesAnnotation = clazz.getAnnotation(CustomNamedStoredProcedureQueries.class);

            for (CustomNamedStoredProcedureQuery queryAnnotation : queriesAnnotation.value()) {
                SimpleJdbcCall jdbcCall = new SimpleJdbcCall(jdbcTemplate)
                        .withSchemaName(queryAnnotation.schemeName())
                        .withCatalogName(queryAnnotation.packageName())
                        .withProcedureName(queryAnnotation.procedureName());
                Set<SqlParameter> parameters = this.getDeclaredParameters(queryAnnotation.parameters());
                jdbcCall.declareParameters(parameters.toArray(new SqlParameter[0]));

                Map<String, RowMapper<?>> rowMappers = getDeclaredRowMappers(queryAnnotation.resultSetRowMappers(), queryAnnotation.resultSetMappers());
                 rowMappers.forEach((parameterName, rowMapper) -> jdbcCall.addDeclaredRowMapper(parameterName, rowMapper));

                jdbcCall.compile();

                storedProcedureContainer.put(queryAnnotation.name(), jdbcCall);
            }
        }
    }

    private Set<SqlParameter> getDeclaredParameters(CustomStoredProcedureParameter[] procedureParameters) {
        Set<SqlParameter> parameters = new HashSet<>();
        for (CustomStoredProcedureParameter procedureParameter : procedureParameters) {
            if (procedureParameter.mode().equals(ParameterMode.IN)) {
                parameters.add(new SqlParameter(procedureParameter.name(), procedureParameter.type()));
            }
            if (procedureParameter.mode().equals(ParameterMode.OUT)) {
                parameters.add(new SqlOutParameter(procedureParameter.name(), procedureParameter.type()));
            }
        }
        return parameters;
    }

    private Map<String, RowMapper<?>> getDeclaredRowMappers(ProcedureRowMapper[] resultSetMappers, ProcedureResultClass[] simpleResultSetMappers) {
        Map<String, RowMapper<?>> mappers = new HashMap<>();
        if (!ArrayUtils.isEmpty(simpleResultSetMappers)) {
            for (ProcedureResultClass procedureResultClass : simpleResultSetMappers) {
                mappers.put(procedureResultClass.name(), BeanPropertyRowMapper.newInstance(procedureResultClass.resultClass()));
            }
            return mappers;
        }
        try {
            for (ProcedureRowMapper resultSetMapper : resultSetMappers) {
                mappers.put(resultSetMapper.name(), resultSetMapper.mapper().newInstance());
            }
        } catch (InstantiationException | IllegalAccessException e) {
            e.printStackTrace();
        }
        return mappers;
    }

    public SimpleJdbcCall getProcedure(String name) {
        return storedProcedureContainer.get(name);
    }
}

假设必须声明@CustomNamedStoredProcedureQueries，即使存储过程只有一个，您也可以根据需要自行更改。

您可以使用ProcedureResultClass 或ProcedureRowMapper，不能同时使用两者，它可以更改为。

示例 (我使用 oracledb，但稍作改动，它可以用于另一个数据库)

@CustomNamedStoredProcedureQueries({
        @CustomNamedStoredProcedureQuery(
                name = "getFoo",
                schemeName = "FooScheme",
                packageName = "FooPackage",
                procedureName = "get_Foo",
                resultSetMappers = @ProcedureResultClass(name = "foo_cursor", resultClass = Foo.class),
                parameters = {
                        @CustomStoredProcedureParameter(name = "foo_id", type = OracleTypes.NUMBER, mode = ParameterMode.IN),
                        @CustomStoredProcedureParameter(name = "error_code", type = OracleTypes.NUMBER, mode = ParameterMode.OUT),
                        @CustomStoredProcedureParameter(name = "foo_cursor", type = OracleTypes.CURSOR, mode = ParameterMode.OUT),
                }
        ),
        @CustomNamedStoredProcedureQuery(
                name = "deleteFoo",
                schemeName = "FooScheme",
                packageName = "FooPackage",
                procedureName = "delete_Foo",
                parameters = {
                        @CustomStoredProcedureParameter(name = "foo_id", type = OracleTypes.NUMBER, mode = ParameterMode.IN),
                        @CustomStoredProcedureParameter(name = "error_code", type = OracleTypes.NUMBER, mode = ParameterMode.OUT),
                }
        )
})
public class Foo {
     long id;
     String name;
     int age;

     public void setId(long id) { 
       this.id= id;
     }
     public long getId() { 
       return this.id;
     }

     public void setName(String name) { 
       this.name = name;
     }
     public String getName() { 
       return this.name;
     }

     public void setAge(int age) { 
       this.age = age;
     }
     public int getAge() { 
       return this.age;
     }
}

叫它:

public Foo getFoo(Long fooId) {
        SqlParameterSource sqlParams = new MapSqlParameterSource()
                .addValue("foo_id", fooId);
        SimpleJdbcCall procedure = projectUtils.getProcedure("getFoo");
        Map<String, Object> result = procedure.execute(sqlParams);
        return ((List<Foo>) result.get("foo_cursor")).get(0);
    }

提示 - 它可以与自定义类型(存储在 bd 中的类型)一起使用，例如 Array，因为您需要创建扩展的 AbstractSqlTypeValue 的自定义类型并覆盖 createTypeValue() ，然后像其他参数一样传递它( 仅在 oracle 上经过仔细测试)。

CustomType 必须在方案级别和更高级别上声明，在包级别它不起作用，方案和类型必须以大写形式调用。

    @Override
    protected Object createTypeValue(Connection connection, int sqlType, String typeName) throws SQLException {
        if (connection.isWrapperFor(OracleConnection.class)) {
            OracleConnection oracleConnection = connection.unwrap(OracleConnection.class);
            return oracleConnection.createARRAY("FOOSCHEME.NUMBERARRAY", this.array); // array pass when customType class init
        }
        return connection.createArrayOf("FOOSCHEME.NUMBERARRAY", this.array);
    }