如何从mongodb获得匹配的子文档？

Question

I have a collection like this

{
    "_id" : ObjectId("545dad3562fa028fb48832f0"),
    "key" : "123456",
    "members" : [
            {
                    "name" : "Shiva",
                    "country" : "India",
                    "profession" : "Software"
            },
            {
                    "name" : "Neil",
                    "country" : "Australia",
                    "profession" : "Software"
            },
            {
                    "name" : "anil",
                    "country" : "India",
                    "profession" : "Software"
            }

    ]
}

Now i want to retrieve the records with country India. When i tried like this

db.user_details.find({ 'key' :'123456','members':{$elemMatch:{'country': "India"}}}, {'members.$': 1}).pretty()

When executed the above code i am getting only first occurrence, How to get all the matched sub documents(like name:Shiva and name:Anil ) from that document

Answer 1

The $elemMatch operator along the the counterpart positional $ operator will only currently match the first element that meets the conditions specified.

In order to get "more than one" match, the best method is to use the aggregation framework:

 db.user_details.aggregate([
     # Always match first to at least find the "documents" with matching elements
     { "$match": {
         "members.country": "India"
     }},

     # Use $unwind to "de-normalize" the array as individual documents
     { "$unwind": "$members" },

     # Then match again to "filter" the content down to matches
     { "$match": {
         "members.country": "India"
     }},

     # Group back to keep an array
     { "$group": {
         "_id": "$_id",
         "key": { "$first": "$key" },
         "members": { "$push": "$members" }
     }}
 ])

That is the basic process for "filtering" more than one match from an array. The basic projection cannot currently do it.

Which returns this:

{
    "_id" : ObjectId("545dad3562fa028fb48832f0"),
    "key" : "123456",
    "members" : [
            {
                    "name" : "Shiva",
                    "country" : "India",
                    "profession" : "Software"
            },
            {
                    "name" : "anil",
                    "country" : "India",
                    "profession" : "Software"
            }
    ]
}