Mongo 具有嵌套嵌入文档和 $project 的聚合管道 $lookup

Question

我有两个集合 users 和 `groups。让我们假装集合中当前包含以下数据。

users 集合有以下文档

{  
   "_id":"57a944390b1acf0d069388c1",
   "first":"John",
   "last":"Smith",
   "email":"John.Smith@gmail.com",
   "groups":[  
      {  
         "id":"57aab09c0b1acf135a6b6856",
         "name":"Group A"
      },
      {  
         "id":"57aab0ed0b1acf135a6b6857",
         "name":"Group B"
      },
      {  
         "id":"57b008be31a5a202ee4ff47b",
         "name":"Group C"
      }
   ]
}

groups 集合有以下两个文档

[  
   {  
      "_id":"57aab0ed0b1acf135a6b6857",
      "name":"Group B",
      "requests":[  
         {  
            "_id":"57b14b1831a5a2756fbc9873",
            "description":"Request A",
            "denied_requests":[ ]
         }
      ]
   }   {  
      "_id":"57b008be31a5a202ee4ff47b",
      "name":"Group C",
      "requests":[  
         {  
            "_id":"57b14c2131a5a2756fbc9874",
            "description":"Request B",
            "denied_requests":[ ]
         },
         {  
            "_id":"57b14e3131a5a2756fbc9875",
            "description":"Request C",
            "denied_requests":[  
               "57a944390b1acf0d069388c1"
            ]
         }
      ]
   }
]

我需要做的是使用聚合管道查询用户的所有请求，不包括用户已拒绝的请求。

我想这将包括以下 psudo 步骤，将 user_id 作为问题的输入。

$match 在用户 _id 字段上
$project 组 id 来自上一步的结果
$lookup 使用 $project 中 users.groups 中 groups 集合的 _id 中的 id
$match 在 requests 上根据要求在 denied_requests 数组中没有当前用户 ID。

样本输入

示例输入为 57a944390b1acf0d069388c1，代表用户 _id

示例输出

示例输出为以下用户未拒绝的请求列表。

[
       {  
          "_id":"57b14b1831a5a2756fbc9873",
          "description":"Request A",
          "denied_requests":[]
       },
       {  
          "_id":"57b14c2131a5a2756fbc9874",
          "description":"Request B",
          "denied_requests":[ ]
       }   
]

Answer 1

运行下面的管道，它应该会给你想要的结果：

var userId = "57a944390b1acf0d069388c1";
db.users.aggregate([
    { "$match": { "_id": userId } },
    { "$unwind": "$groups" },
    {
        "$lookup": {
            "from": "groups",
            "localField": "groups.id",
            "foreignField": "_id",
            "as": "group"
        }
    },
    { "$unwind": "$group" },
    {
        "$project": {
            "requests": {
                "$filter": {
                    "input": "$group.requests",
                    "as": "item",
                    "cond": { "$not": { 
                        "$setIsSubset": [ 
                            [userId], "$$item.denied_requests" 
                        ] 
                    } }
                }
            }
        }
    },
    { "$unwind": "$requests" },
    {
        "$project": {
            "_id": "$requests._id",
            "description": "$requests.description",
            "denied_requests": "$requests.denied_requests"
        }
    }
])

Mongo 具有嵌套嵌入文档和 $project 的聚合管道 $lookup

Mongo Aggregation Pipeline $lookup with nested embedded documents and $project

mongodb

mongodb-query

aggregation-framework

mgo