使用 JdbcSimpleCall 从存储过程中获取 pl/sql 结构数组

Question

我正在尝试使用 SimpleJDBCCall 执行 oracle 存储过程，所有表和存储过程都在餐厅模式中，表如下所示：

CREATE TABLE STAFF
(
    STAFF_ID             NUMBER(5),
    STAFF_FIRST_NAME     VARCHAR2(10 BYTE)        NOT NULL,
    STAFF_LAST_NAME      VARCHAR2(20 BYTE)        NOT NULL,
    STAFF_ROLE           VARCHAR2(20 BYTE)        NOT NULL,
    STAFF_OTHER_DETAILS  VARCHAR2(50 BYTE)
);

我的类型包：

CREATE OR REPLACE PACKAGE Staff_Types 
AS
    TYPE Staff_Collection IS TABLE OF Staff%ROWTYPE;  
END Staff_Types;

我的访问包：

CREATE OR REPLACE PACKAGE Staff_TAPI 
AS
    FUNCTION getAllStaff RETURN Staff_Types.Staff_Collection;
END Staff_TAPI;

CREATE OR REPLACE PACKAGE BODY Staff_Tapi
AS
    FUNCTION getAllStaff 
        RETURN Staff_Types.Staff_Collection
    IS
        all_staff Staff_Types.Staff_Collection;
    BEGIN
        SELECT * 
        BULK COLLECT INTO all_staff
        FROM Staff;

        RETURN all_staff;
    END;

END Staff_Tapi;

Java 访问：

@Component
@Qualifier("staffJdbcDAO")
public class StaffJDBCDAO implements StaffDAO {
    JdbcTemplate jdbcTemplate;
    SimpleJdbcCall getAllMembersSP;

    @Autowired
    @Qualifier("dataSource")
    DataSource dataSource;

    @Autowired
    @Qualifier("jdbcTemplate")
    public void setJdbcTemplate(JdbcTemplate jdbcTemplate) {
        this.jdbcTemplate = jdbcTemplate;
        initializeStoredProceduresCalls();
    }

    private void initializeStoredProceduresCalls() {
        getAllMembersSP = new SimpleJdbcCall(jdbcTemplate);
        getAllMembersSP.withCatalogName("Staff_Tapi");
        getAllMembersSP.withFunctionName("getAllStaff");
        getAllMembersSP.declareParameters(
            new SqlOutParameter("return",
                    Types.OTHER,
                    "Staff_Types.Staff_Collection",
                    new SqlReturnStructArray<>( new StaffMapper() )
            )
        );
        getAllMembersSP.compile();

    }

    @Override
    public List<Staff> getAllMembers() {
        Staff[] staff = getAllMembersSP.executeFunction(Staff[].class,new   HashMap<String,Object>() );
        return Arrays.asList(staff);
    }
}

映射类：

public class StaffMapper implements StructMapper<Staff> {

    @Override
    public STRUCT toStruct(Staff staff, Connection connection, String typeName) throws SQLException {
        StructDescriptor descriptor = StructDescriptor.createDescriptor(typeName, connection);

        Object[] attributes = new Object[5];
        attributes[0] = new Integer( staff.getId() );
        attributes[1] = new String("STAFF_FIRST_NAME");
        attributes[2] = new String("STAFF_LAST_NAME");
        attributes[3] = new String("STAFF_ROLE");
        attributes[4] = new String("STAFF_OTHER_DETAILS");
        Struct staffStruct = connection.createStruct(typeName,attributes);
        return new STRUCT(descriptor,connection,attributes);
    }

    @Override
    public Staff fromStruct(STRUCT struct) throws SQLException {
        StructDescriptor descriptor = struct.getDescriptor();
        ResultSetMetaData metaData = descriptor.getMetaData();
        Object[] attributes = struct.getAttributes();
        Map<String,Object> attributeMap = new HashMap<>();
        int idx = 1;
        for ( Object attribute : attributes )
            attributeMap.put( metaData.getColumnName(idx++),attribute );


        int id = ((Integer)attributeMap.get("STAFF_ID")).intValue();
        String firstName = (String) attributeMap.get("STAFF_FIRST_NAME");
        String lastName = (String) attributeMap.get("STAFF_LAST_NAME");
        String staffRole = (String) attributeMap.get("STAFF_ROLE");
        String otherDetails = (String) attributeMap.get("STAFF_OTHER_DETAILS");
        return new Staff(id,firstName,lastName,staffRole,otherDetails);
    }
}

和工作人员：

public class Staff {
    private int id;
    private String firstName;
    private String lastName;
    private String profession;
    private String otherDetails;


    public Staff(int id, String firstName, String lastName, String profession, String otherDetails) {
        this.id = id;
        this.firstName = firstName;
        this.lastName = lastName;
        this.profession = profession;
        this.otherDetails = otherDetails;
    }

    public int getId() {
        return id;
    }

    public int setId(int id) {
        this.id = id;
    }

    // and others getters and setters
}

当我从 StaffDAO 执行 getAllMembers 时，我不断得到： CallableStatementCallback; SQL 的未分类 SQLException [{? = 调用 STAFF_TAPI.GETALLSTAFF()}]; SQL 状态 [99999]；错误代码 [17004]；无效的列类型：1111；

当我将返回类型参数更改为 Types.Array 时，我得到： CallableStatementCallback; SQL 的未分类 SQLException [{? = 调用 STAFF_TAPI.GETALLSTAFF()}]; SQL 状态 [99999]；错误代码 [17074];
无效的名称模式：restaurant.Staff_Types.Staff_Collection；

我尝试了两种方式，模式“Staff_Types.Staf_collection”得到了相同的结果，我尝试了将近 2 天，但不知道我该怎么做，如果有人有任何建议，我会很感激。

Answer 1

您可能希望在 java 代码中大写您的自定义类型，例如

    getAllMembersSP.declareParameters(
        new SqlOutParameter("return",
                Types.OTHER,
                "STAFF_TYPES.STAFF_COLLECTION",
                new SqlReturnStructArray<>( new StaffMapper() )
        )
    );

Answer 2

您无法通过 JDBC 从存储过程加载 PL/SQL record。事实上，您甚至无法从 Oracle SQL 加载这样的类型。有关详细信息，另请参阅此问题：

• Fetch Oracle table type from stored procedure using JDBC

您只能通过 JDBC 加载 SQL 类型（与 PL/SQL 类型相反）。鉴于您的示例，您需要编写：

-- You cannot really avoid this redundancy
CREATE TYPE STAFF AS OBJECT
(
    STAFF_ID             NUMBER(5),
    STAFF_FIRST_NAME     VARCHAR2(10 BYTE)        NOT NULL,
    STAFF_LAST_NAME      VARCHAR2(20 BYTE)        NOT NULL,
    STAFF_ROLE           VARCHAR2(20 BYTE)        NOT NULL,
    STAFF_OTHER_DETAILS  VARCHAR2(50 BYTE)
);
CREATE TYPE STAFF_TABLE AS TABLE OF STAFF;

然后：

CREATE OR REPLACE PACKAGE Staff_TAPI 
AS
    FUNCTION getAllStaff RETURN STAFF_TABLE;
END Staff_TAPI;

Answer 3

为了轻松集成您的 PL/SQL 调用，并且由于它已经构建为函数：您是否考虑过这样的事情？

select * from TABLE(CAST(Staff_Tapi.getAllStaff() as Staff_Types.Staff_Collection))

这样，您可以轻松地将其作为常规 JDBC 查询执行。完成后，只需处理ResultSet 它返回的fromStruct 方法的一些小变体，以便将List<Staff> 列表返回到您在其之上的任何业务逻辑。希望你觉得这很有用！