Go 工程化(十一) 如何优雅的写出 repo 层代码

注:本文已发布超过一年,请注意您所使用工具的相关版本是否适用

注:本文所有示例代码都可以在 blog-code 仓库中找到

本系列为 Go 进阶训练营 笔记,访问 博客: Go进阶训练营, 即可查看当前更新进度,部分文章篇幅较长,使用 PC 大屏浏览体验更佳。

还是以获取一篇文章为例,我们在获取文章的时候大部分时候可能都是通过 id 获取的,但是我们也可能通过标题等其它信息获取文章的数据,这时候我们的 repo 层代码怎么写呢?

最简单的方式,就是我们直接在 repo 这里写两个方法

1
2
3
4
5
// IArticleRepo IArticleRepo
type IArticleRepo interface {
GetArticleByTitle(ctx context.Context, title string) (*Article, error)
GetArticleByID(ctx context.Context, id int) (*Article, error)
}

这样最简单也最直观,但是问题是我们的实际的业务需求往往比我们的例子复杂,如果我们需要通过 id 或者标题获取呢?再添加一个 GetArticleByIDOrTitle ?

这么做的话也不是不行,但是这么做的话就会让我们的 repo 的代码随着时间的增长越来越多不说,命名也是问题,因为组合的方式可能是多种多样的

接下来给大家提供一种我们正在使用的一种思路,利用 Function Options 这种 Go 常见的编程范式,使我们的 repo 更优雅,也可扩展

DBOption

注意: 笔者这里使用的是 GORM,但是这种方式不仅仅适用于 orm 的情况,只是相对方便一点而已

1
2
3
4
5
6
7
8
type DBOption func(*gorm.DB) *gorm.DB

// IArticleRepo IArticleRepo
type IArticleRepo interface {
WithByID(id uint) DBOption
WithByTitle(title string) DBOption
GetArticle(ctx context.Context, opts ...DBOption) (*Article, error)
}

我们定义一个的 DBOption 这个 Option 方法会作为我们 repo 层方法中的最后一个参数,这样我们在定义方法的时候就可以简洁一些,就不必定义很多 GetArticleByXXX 方法了,而是通过定义很多 WithByXXXOption 方法来解决。

这样在 usecase 层,我们只需要这么调用即可

1
2
3
4
func (u *article) GetArticle(ctx context.Context, id int) (*domain.Article, error) {
// 这里可能有其他业务逻辑...
return u.repo.GetArticle(ctx, u.repo.WithByID(uint(id)))
}

优点

复用: 虽然看上去我们只是把 GetArticleByXXX 换成了 WithByXXX 该有的方法并没有变少,但是我们拆分之后会发现很多可以复用的方法,例如 WithByID 这种几乎是每个实体都会有的方法,我们就不用重复写了。

1
2
3
4
5
6
7
8
9
10
11
// GetArticle 和 GetAuthor 都能用上

func (u *article) GetArticle(ctx context.Context, id int) (*domain.Article, error) {
// 这里可能有其他业务逻辑...
return u.repo.GetArticle(ctx, u.repo.WithByID(uint(id)))
}

func (u *article) GetAuthor(ctx context.Context, id int) (*domain.Author, error) {
// 这里可能有其他业务逻辑...
return u.repo.GetAuthor(ctx, u.repo.WithByID(uint(id)))
}

最小化: 这么修改了之后,拆分组合更加方便了,很多查询条件都可以最小化,例如我们可以添加一个 WithSelects 的方法,我们在 usecase 调用的时候就可以传入当前场景只需要关注的字段就可以了

1
2
3
4
5
6
7
8
9
10
// GetArticle 返回文章的同时也需要返回作者的名字
func (u *article) GetArticle(ctx context.Context, id int) (*domain.Article, error) {
article, err := u.repo.GetArticle(ctx, u.repo.WithByID(uint(id)))
if err != nil {
return err
}

article.Author, err = u.repo.GetAuthor(ctx, u.repo.WithByArticleID(id), u.repo.WithBySelects("id", "name"))
return article, err
}

可测性: repo 层的测试会变得更加方便,这样修改之后我们可以将查询条件拆分出来进行测试,会比之前耦合在一起测试简单很多。

抽象: 这种方式可以让我们抽象 CURD 接口更加方便,在 repo 层实现的时候,我们可以直接把 curd 的方法都给抽象出来

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
// 这里以创建为例

func (r *userRepo) optionDB(ctx context.Context, opts ...model.DBOption) *gorm.DB {
db := r.db.WithContext(ctx)
for _, opt := range opts {
db = opt(db)
}
return db
}

func (r *userRepo) create(ctx context.Context, data any, opts ...model.DBOption) error {
db := r.optionDB(ctx, opts...)
err := db.Create(data).Error
if err != nil {
return pb.ErrorDbCreateFailf("err: %+v", err)
}
return nil
}

总结

今天给大家介绍了使用 Function Option 的方式来写 repo 层的代码,接下来我们就简单总结一下

1
2
3
4
5
6
7
8
type DBOption func(*gorm.DB) *gorm.DB

// IArticleRepo IArticleRepo
type IArticleRepo interface {
WithByID(id uint) DBOption
WithByTitle(title string) DBOption
GetArticle(ctx context.Context, opts ...DBOption) (*Article, error)
}

优点

  • 复用: 虽然看上去我们只是把 GetArticleByXXX 换成了 WithByXXX 该有的方法并没有变少,但是我们拆分之后会发现很多可以复用的方法,例如 WithByID 这种几乎是每个实体都会有的方法,我们就不用重复写了。
  • 最小化: 这么修改了之后,拆分组合更加方便了,很多查询条件都可以最小化,例如我们可以添加一个 WithSelects 的方法,我们在 usecase 调用的时候就可以传入当前场景只需要关注的字段就可以了
  • 可测性: repo 层的测试会变得更加方便,这样修改之后我们可以将查询条件拆分出来进行测试,会比之前耦合在一起测试简单很多。
  • 抽象: 这种方式可以让我们抽象 CURD 接口更加方便

缺点

  • 最大的缺点就是有的问题在单测可能测试不出来了,usecase 的测试中,repo 层被 mock 掉了,repo 在测试的时候大部分我们只会测试当前的方法,所以 usecase 有使用比较复杂的查询语句的时候,repo 测试最好测一测真实的使用场景,不要仅测试单个 Option 方法

今天的文章就到这里,下篇文章给大家介绍一下 API 定义上的一点小技巧

关注我获取更新

wechat
知乎
github

猜你喜欢