从 $unwind mongodb 获取有限的数据

Question

我有一个聊天集合，下面给出了

{
    "_id" : ObjectId("5a44e6545818041cde24aac9"),
    "messages" : [ 
        {
            "createdAt" : ISODate("2017-12-23T07:46:29.201Z"),
            "message" : "Hii",
            "userId" : "000000"
        }, 
        {
            "createdAt" : ISODate("2017-12-23T12:46:29.201Z"),
            "message" : "Bye gtg ttyl",
            "userId" : "111111"
        }, 
        {
            "createdAt" : ISODate("2017-12-24T07:46:29.201Z"),
            "message" : "Bye Take Care",
            "userId" : "000000"
        }
    ],
    "createdAt" : ISODate("2017-10-23T07:46:29.201Z"),
    "users" : [ 
        {
            "facebookId" : "000000",
            "unread" : 0
        }, 
        {
            "facebookId" : "111111",
            "unread" : 0
        }
    ]
}

当我在 nodejs 中执行以下查询时

chats.aggregate( [ { $unwind : "$messages" },{ $unwind : "$users" } ])

我得到一个类似这样的数组，其中 unwind 打开我的数组是这样的

[
    {
        "_id": "5a44e6545818041cde24aac9",
        "messages": {
            "createdAt": "2017-12-23T07:46:29.201Z",
            "message": "Hii",
            "userId": "000000"
        },
        "createdAt": "2017-10-23T07:46:29.201Z",
        "users": {
            "facebookId": "000000",
            "unread": 0
        }
    },
    {
        "_id": "5a44e6545818041cde24aac9",
        "messages": {
            "createdAt": "2017-12-23T07:46:29.201Z",
            "message": "Hii",
            "userId": "000000"
        },
        "createdAt": "2017-10-23T07:46:29.201Z",
        "users": {
            "facebookId": "111111",
            "unread": 0
        },
    }, //And so on.
] 

但是我只能在messages.userId !== users.facebookIdmessages.userId !== users.facebookId展开后获取对象

当我获取数据时，直接从 MongoDB 获取类似于以下对象的内容

[
    {
        "_id": "5a44e6545818041cde24aac9",
        "messages": {
            "createdAt": "2017-12-23T07:46:29.201Z",
            "message": "Hii",
            "userId": "000000"
        },
        "createdAt": "2017-10-23T07:46:29.201Z",
        "users": {
            "facebookId": "111111",
            "unread": 0
        },
    },{
        "_id": "5a44e6545818041cde24aac9",
        "messages": {
            "createdAt": "2017-12-23T12:46:29.201Z",
            "message": "Bye gtg ttyl",
            "userId": "111111"
        },
        "createdAt": "2017-10-23T07:46:29.201Z",
        "users": {
            "facebookId": "000000",
            "unread": 0
        }
    }, {
        "_id": "5a44e6545818041cde24aac9",
        "messages": {
            "createdAt": "2017-12-24T07:46:29.201Z",
            "message": "Bye Take Care",
            "userId": "000000"
        },
        "createdAt": "2017-10-23T07:46:29.201Z",
        "users": {
            "facebookId": "111111",
            "unread": 0
        }
    }
] 

Answer 1

试试下面一个

db.coll3.aggregate([{
            $unwind: '$users'
        }, {
            $project: {
                messages: 1,
                users: 1,
                'messages': {
                    $filter: {
                        input: '$messages',
                        as: 'item',
                        cond: {
                            $ne: ['$$item.userId', '$users.facebookId']
                        }
                    }
                }
            }
        },{ $unwind : "$messages" }
    ])

这是在 mongoshell 中。我注意到您记录中的零数量不相等。即使@veeram 和@Astro 的上述查询为您提供了正确的结果（如果您尝试过），请确保您不会与它们混淆

Answer 2

您可以包含$match 阶段以仅获取符合此条件的那些文档：

db.chats.aggregate( [ { $unwind : "$messages" },{ $unwind : "$users" }, {$match:{ 'messages.userId' :{ $ne: 'users.facebookId'} }])