如何配置@ResultsetMapping，使@ManyToOne关联无数据时属性为null

我有一个Location类，包含如下@ManyToOne关联属性：

public class Location {
......
@ManyToOne(cascade = {CascadeType.PERSIST, CascadeType.MERGE, CascadeType.REFRESH}, fetch = FetchType.LAZY)
@JoinColumn(name = "ledgerId")
@JsonIgnore
protected LedgerValue ledgerValue;

由于查询逻辑复杂，我使用原生查询配合@ResultSetMapping来映射LedgerValue实体，配置如下：

@ConstructorResult(
        targetClass = LedgerValue.class,
        columns = {
            @ColumnResult(name = "lvid"),
            @ColumnResult(name = "mainc", type = Long.class),
            @ColumnResult(name = "emailonly", type = Boolean.class),
            @ColumnResult(name = "copy", type = Boolean.class),
        }
    )

目前功能正常，Location类可被填充LedgerValue实体，但当Location无关联的LedgerValue时，仍会得到一个所有属性为null的LedgerValue对象。我希望此时ledgerValue属性本身为null，除了自行编写完整SQL或使用ResultSetExtractor外，是否有简便方法实现该需求？

方法1：改用@EntityResult替代@ConstructorResult #

如果LedgerValue是标准JPA实体，优先使用@EntityResult映射结果。JPA在处理左关联无匹配数据时，会自动将关联属性设为null，而非实例化全空对象。配置示例：

@SqlResultSetMapping(
    name = "LocationWithLedgerMapping",
    entities = {
        @EntityResult(entityClass = Location.class),
        @EntityResult(
            entityClass = LedgerValue.class,
            fields = {
                @FieldResult(name = "id", column = "lvid"),
                @FieldResult(name = "mainc", column = "mainc"),
                @FieldResult(name = "emailonly", column = "emailonly"),
                @FieldResult(name = "copy", column = "copy")
            }
        )
    }
)

需确保SQL使用左外连接关联Location和LedgerValue，且无匹配时lvid等字段为null，JPA会自动识别并将ledgerValue设为null。

方法2：使用@FactoryResult通过静态工厂控制返回值 #

通过@FactoryResult替代@ConstructorResult，借助静态工厂方法判断字段是否为空，返回null而非全空对象。示例如下：

1. 定义静态工厂类 #

public class LedgerValueFactory {
    public static LedgerValue createLedgerValue(Long lvid, Long mainc, Boolean emailonly, Boolean copy) {
        // 以主键lvid为空作为无关联的判断依据
        if (lvid == null) {
            return null;
        }
        LedgerValue lv = new LedgerValue();
        lv.setId(lvid);
        lv.setMainc(mainc);
        lv.setEmailonly(emailonly);
        lv.setCopy(copy);
        return lv;
    }
}

2. 修改@SqlResultSetMapping配置 #

@SqlResultSetMapping(
    name = "LocationLedgerMapping",
    entities = @EntityResult(entityClass = Location.class),
    factoryResults = @FactoryResult(
        targetClass = LedgerValueFactory.class,
        methodName = "createLedgerValue",
        columns = {
            @ColumnResult(name = "lvid"),
            @ColumnResult(name = "mainc", type = Long.class),
            @ColumnResult(name = "emailonly", type = Boolean.class),
            @ColumnResult(name = "copy", type = Boolean.class),
        }
    )
)

当lvid为null时，工厂方法返回null，JPA会自动将Location的ledgerValue属性设为null。

方法3：SQL层面配合空值判断 #

在原生SQL中通过CASE语句，确保只有当关联存在时才返回LedgerValue的字段值，配合上述工厂方法进一步处理：

SELECT 
    loc.*,
    CASE WHEN lv.lvid IS NOT NULL THEN lv.lvid END AS lvid,
    CASE WHEN lv.lvid IS NOT NULL THEN lv.mainc END AS mainc,
    CASE WHEN lv.lvid IS NOT NULL THEN lv.emailonly END AS emailonly,
    CASE WHEN lv.lvid IS NOT NULL THEN lv.copy END AS copy
FROM Location loc
LEFT JOIN LedgerValue lv ON loc.ledgerId = lv.lvid

这种方式能强化空值判断逻辑，避免因SQL返回默认空值导致的全空对象实例化。

内容的提问来源于stack exchange，提问作者Mathias