GitLab

整体思路，把结果集抽象成一个 table，然后对这个 table 应用多个 pipeline 函数，最终得到处理后的结果，下面是一些案例~

1. 每个每个用户状态中人的数量

GET /users?$apply=groupby('status', count() as count)

[
	{
		status: 1,
		count: n,
	},
	...
]

GET /orders?$apply=groupby('uid', sum(price) as totalPrice)

[
	{
		uid: 1,
		totalPrice: 10.0
	},
	...
]

3. 在 2 的基础上获取用户信息

GET /orders?$apply=groupby('uid', sum(price) as totalPrice)&$load=users for uid

[
	{
		uid: 1,
		totalPrice: 10.0,
		user: {}
	}
]

4. 行转列？

/selections?$apply=col2row('rid', 'class.schedules.day,class.schedules.lesson')