다대다 데이터 조회 질문입니다.

강사님 강의를 보고 최근에 프로젝트를 하다 궁금한게 생겨 질문 남기게 되었습니다.

Product <-> Category Entity 다대다 매핑을 위해

product(OneToMany) <-> product_category(ManyToOne) <-> category(OneToMany)

위와 같은 테이블로 매핑한 상태이고 join 상속 전략으로 district_category 테이블도 생성 하였습니다.

district_category를 쿼리 파라미터로 받고 해당 district_category를 갖고 있는 상품들을 검색하고 페이징을 하는 로직을 짠다고 했을때 두가지 방법이 생각났는데 둘 중 어느게 더 적합한지 모르겠습니다

1. 검색어를 바탕으로 product_category에서 Product를 EntityGraph로 같이 찾은 후에 찾은 ProductCategory.getproduct() 와 같은 방식
   
   SearchService

Category category = districtCategoryRepository.findByDistrictEnum(districtEnum)
        .orElseThrow(() -> new ProductException(ProductExceptionType.CATEGORY_NOT_FOUND));

return new ProductCategoryToProductPage(productCategoryRepository.findAllByCategory(pageable, category));

ProductCategoryRepository

@EntityGraph(attributePaths = "product")
Page<ProductCategory> findAllByCategory(Pageable pageable, Category category);

ProductCategoryToProduct

    public ProductCategoryToProductPage(Page<ProductCategory> page) {
        this.content.addAll(page.getContent().stream()
                .map(ProductCategory::getProduct)
                .map(ProductListGetResponseDTO::new)
                .collect(toList()));
        this.totalPages = page.getTotalPages();
        this.totalElements = page.getTotalElements();
        this.pageNumber = page.getNumber() + 1;
        this.size = page.getSize();
    }

발생 쿼리문
select

districtca0_.category_id as category2_1_,

districtca0_1_.parent_id as parent_i3_1_,

districtca0_.district_enum as district1_2_

from

district_category districtca0_

inner join

category districtca0_1_

on districtca0_.category_id=districtca0_1_.category_id

where

districtca0_.district_enum=?



select

productcat0_.product_category_id as product_1_8_0_,

product1_.product_id as product_1_7_1_,

productcat0_.category_id as category2_8_0_,

productcat0_.product_id as product_3_8_0_,

product1_.created_date as created_2_7_1_,

product1_.content_detail as content_3_7_1_,

product1_.product_content as product_4_7_1_,

product1_.product_name as product_5_7_1_,

product1_.product_price as product_6_7_1_,

product1_.product_status as product_7_7_1_,

product1_.product_thumbnail as product_8_7_1_

from

product_category productcat0_

left outer join

product product1_

on productcat0_.product_id=product1_.product_id

where

productcat0_.category_id=? limit ?



select

count(productcat0_.product_category_id) as col_0_0_

from

product_category productcat0_

where

productcat0_.category_id=?

2. 검색어를 바탕으로 product에서 직접 찾기 (데이터 뻥튀기의 문제는 쿼리dsl 이용 productId로 groupBy로 해결) 글 쓰고 생각해보니 A카테고리는 B라는 상품 안에서는 하나밖에 있을 수가 없으니 굳이 groupBy를 안써도 될거 같네요
   
   SearchService

        Category category = districtCategoryRepository.findByDistrictEnum(districtEnum)
                .orElseThrow(() -> new ProductException(ProductExceptionType.CATEGORY_NOT_FOUND));

        return productRepository.findAllByCategory(pageable, category);

ProductRepository

    @Override
    public Page<Product> findAllByCategory(Pageable pageable, Category category) {
        List<Product> content = queryFactory.selectFrom(product)
                .join(product.productCategories, productCategory)
                .where(productCategory.category.categoryId.eq(category.getCategoryId()))
                .groupBy(product.productId)
                .offset(pageable.getOffset())
                .limit(pageable.getPageSize())
                .fetch();

        Long total = queryFactory
                .select(Wildcard.count)
                .from(product)
                .join(product.productCategories, productCategory)
                .where(productCategory.category.categoryId.eq(category.getCategoryId()))
                .fetchOne();

        return new PageImpl<>(content, pageable, total);
    }

발생 쿼리문
select

districtca0_.category_id as category2_1_,

districtca0_1_.parent_id as parent_i3_1_,

districtca0_.district_enum as district1_2_

from

district_category districtca0_

inner join

category districtca0_1_

on districtca0_.category_id=districtca0_1_.category_id

where

districtca0_.district_enum=?



select

product0_.product_id as product_1_7_,

product0_.created_date as created_2_7_,

product0_.content_detail as content_3_7_,

product0_.product_content as product_4_7_,

product0_.product_name as product_5_7_,

product0_.product_price as product_6_7_,

product0_.product_status as product_7_7_,

product0_.product_thumbnail as product_8_7_

from

product product0_

inner join

product_category productcat1_

on product0_.product_id=productcat1_.product_id

where

productcat1_.category_id=?

group by

product0_.product_id limit ?



select

count(*) as col_0_0_

from

product product0_

inner join

product_category productcat1_

on product0_.product_id=productcat1_.product_id

where

productcat1_.category_id=?