如何在Mongodb中将文档与对象属性的值匹配

Question

I'm trying to match some document in mongoDB:

My Document model :

profile: {
  languages: [
    {"name": "French", "level": 1},
    {"name": "English", "level": 2},
    {"name": "Spanish", "level": 4}
  ]
}

What I (can) have to search my result set:

lang = ["French", "English"];
objLang = [
  {name: "French"},
  {name: "English"}
];

What I need is to db.find() all documents that match at least one of the languages, for example :

profile.languages.name = "French"

or

profile.languages.name = "English"

WHat I mean is that if I have French or English in my option set, I need to get all the users who have a element of their languages array where name match one of my options, no matter the level of the languages.

So, unless I'm wrong, I can't do

db.find({"profile.languages": {$in: [{name: "French"}, {name: "English"}]});

How would you proceed ?

Thanks a lot.

David

Answer 1

You could try using the dot notation to access both the elements of an array and to access the fields of an embedded document:

db.find({
    "profile.languages.name": {
        "$in": ["French", "English"]
    }
});

Answer 2

Actually, you were almost right:

db.collection.find({"profile.languages.name":{$in: ["French","English"]} })

Since profile.languages is an array of subdocuments, you can call in for one of the subdocument's keys and the subsequent query gets mapped to all subdocuments containing that key.

However, without proper indexing, an added .explain() shows something pretty ugly:

{
  "queryPlanner" : {
    "plannerVersion" : 1,
    "namespace" : "test.languages",
    "indexFilterSet" : false,
    "parsedQuery" : {
        "profile.languages.name" : {
            "$in" : [
                "English",
                "French"
            ]
        }
    },
    "winningPlan" : {
        "stage" : "COLLSCAN",
        "filter" : {
            "profile.languages.name" : {
                "$in" : [
                    "English",
                    "French"
                ]
            }
        },
        "direction" : "forward"
    },
    "rejectedPlans" : [ ]
  }
}

( serverInfo was ommited).

So in order to make this query efficient, you need to create an index over the field you want to query:

db.languages.ensureIndex({"profile.languages.name":1})

An added explain now tells us that the matching documents are identified via the index:

{
"queryPlanner" : {
    "plannerVersion" : 1,
    "namespace" : "test.languages",
    "indexFilterSet" : false,
    "parsedQuery" : {
        "profile.languages.name" : {
            "$in" : [
                "English",
                "French"
            ]
        }
    },
    "winningPlan" : {
        "stage" : "FETCH",
        "inputStage" : {
            "stage" : "IXSCAN",
            "keyPattern" : {
                "profile.languages.name" : 1
            },
            "indexName" : "profile.languages.name_1",
            "isMultiKey" : true,
            "direction" : "forward",
            "indexBounds" : {
                "profile.languages.name" : [
                    "[\"English\", \"English\"]",
                    "[\"French\", \"French\"]"
                ]
            }
        }
    },
    "rejectedPlans" : [ ]
  },
  "ok" : 1
}

Answer 3

Please try the below query :

Solution 1 :

db.collection.aggregate([
     {
       $unwind:"$profile.languages"
     },
     {
       $match:{"profile.languages.name":{"$in":["French", "English"]}}
     },
     {
       $group:{_id: "$_id", profile:{"$push":"$profile.languages"}}
     }
  ])

Solution 2 :

db.collection.find(
   {
     "profile.languages.name": 
       {
        "$in": [ "English", "French"]
       }
   }
 );