测试环境搭建

主表 Dimension

@Getter
@Setter
@NoArgsConstructor
@AllArgsConstructor
@SuperBuilder
@TableName(value = "sys_dimension", subTables = {"sys_dimension_unit"})
public class Dimension extends BaseComponentEntity {

    @NotBlank(message = "维度不能为空")
    @Length(max = 32, message = "维度不能超过32个字符")
    private String code;

    @NotBlank(message = "名称不能为空")
    @Length(max = 32, message = "名称不能超过32个字符")
    private String name;

    @NotNull(message = "分类不能为空")
    private DimensionCategoryEnum category;

}

@TableName 添加属性 subTables = {"sys_dimension_unit"} 表示子表是 sys_dimension_unit

那么，当删除的时候就会级联删除子表的数据。子表的外键是约定大于配置，采用“表名小写_id”。本例中的外键是 dimension_id
子表 Unit

@Getter
@Setter
@NoArgsConstructor
@AllArgsConstructor
@SuperBuilder
@TableName("sys_dimension_unit")
public class Unit extends BaseComponentEntity {

    @NotBlank(message = "单位code不能为空")
    @Length(max = 32, message = "维度不能超过32个字符")
    private String code;

    @NotBlank(message = "名称不能为空")
    @Length(max = 32, message = "名称不能超过32个字符")
    private String name;

    @NotNull(message = "分子不能为空")
    private Integer n;

    @NotNull(message = "分母不能为空")
    private Integer d;

    private Integer e;

    private BigDecimal k;

    @NotNull(message = "维度id不能为空")
    private Long dimensionId;

}

dimensionId 表示外键，对应的表字段是 dimension_id。

编程式级联查询

• selectAsSubTable

public void selectAsSubTable(List<Map<String, Object>> masterData, String property, String refColumnName) {
    Map<Long, List<T>> refColumnNameMap = this.groupByColumnName(refColumnName, (Collection)masterData.stream().map((rowx) -> {
        return rowx.get("id");
    }).collect(Collectors.toSet()));
    Iterator var5 = masterData.iterator();

    while(var5.hasNext()) {
        Map<String, Object> row = (Map)var5.next();
        List<T> subTableList = (List)refColumnNameMap.get(row.get("id"));
        row.put(property, CollectionUtils.isEmpty(subTableList) ? Collections.emptyList() : subTableList);
    }

}

子表操作，通过从masterData中获取Id值后，获取对应的子表数据。property作为可以放到Map中。
这个一般配合主表的查询。

public Grid<Map<String, Object>> list(String code, String name) {
    Grid<Map<String, Object>> grid = GridUtils.list(warehouseDAO.getSelectSQL() + " WHERE name like :name AND code = :code",
            Params.builder(2).pv("code", code).pv("name", name).pv(EntityConstants.LOGIC_DELETE_COLUMN_NAME, false).build());

    warehouseSpaceDAO.selectAsSubTable(grid.getRows(), "spaces", "warehouse_id");
    return grid;
}

• groupByColumnName

public Map<Long, List<T>> groupByColumnName(String refColumnName, Collection<?> refValues) {
    Map<Long, List<T>> refColumnNameMap = (Map)this.selectByParams(Params.builder(1).pv("refColumnName", refValues).build(), refColumnName + " IN (:refColumnName)").stream().collect(Collectors.groupingBy((t) -> {
        return (Long)this.getPropertyValue(t, (String)this.columnNameToPropertyNameMap.get(refColumnName));
    }));
    return refColumnNameMap;
}

根据字段进行分组，并通过 Map 进行收集。