MongoDB：聚合和展平一个数组字段

Question

After 28 years working with relational databases (SQL Server, MySQL, Oracle, Informix) I have moved to MongoDB. It has been slow going over the last two weeks. I would like to submit a couple of questions to confirm my thoughts.

My document looks like the following (ignore groupings for this question):

{
    "_id": "xyz-800",
    "site": "xyz",
    "user": 800,
    "timepoints": [
        {"timepoint": 0, "a": 1500, "b": 700},
        {"timepoint": 2, "a": 1000, "b": 200},
        {"timepoint": 4, "a": 3500, "b": 1500}
    ],
    "groupings": [
        {"type": "MNO", "group": "<10%", "raw": "1"},
        {"type": "IJK", "group": "Moderate", "raw": "23"}
    ]
}

I would like to flatten the timepoints nested array. The following works, but is there a way to wildcard the attributes in timepoints instead of listing each one? The reason could be if a new attribute (eg, 'c') is added to the subdocument I then have to modify the code or if this subdocument had a lot of attributes I would need to list each one instead of using a wildcard, if possible.

db.records.aggregate( {$unwind : "$timepoints"}, 
                      {$project: {_id: 1, site: 1, user: 1, 
                                  'timepoint': '$timepoints.timepoint', 
                                  'a': '$timepoints.a', 
                                  'b': '$timepoints.b'}})

Result:

{"id":"xyz-800", "site":"xyz", "user":800, "timepoint": 0, "a":1500, "b":700}
{"id":"xyz-800", "site":"xyz", "user":800, "timepoint": 2, "a":1000, "b":200}
{"id":"xyz-800", "site":"xyz", "user":800, "timepoint": 4, "a":3500, "b":1500}

I am currently using MongoDB 3.2

Answer 1

Starting MongoDb 3.4, we can use $ addFields to add the top level fields to the embedded document and use $replaceRoot to promote embedded document to top level.

db.records.aggregate({
    $unwind: "$timepoints"
}, {
    $addFields: {
        "timepoints._id": "$_id",
        "timepoints.site": "$site",
        "timepoints.user": "$user"
    }
}, {
    $replaceRoot: {
        newRoot: "$timepoints"
    }
})

Sample Output

{ "timepoint" : 0, "a" : 1500, "b" : 700, "_id" : "xyz-800", "site" : "xyz", "user" : 800 }
{ "timepoint" : 2, "a" : 1000, "b" : 200, "_id" : "xyz-800", "site" : "xyz", "user" : 800 }
{ "timepoint" : 4, "a" : 3500, "b" : 1500, "_id" : "xyz-800", "site" : "xyz", "user" : 800 }