I\'m currently using aggregate operator to return documents that has an array of embedded (sub) documents. I want to rename the field name for the array and also rename fie
I think I figured it out the following aggregate query. I would be interested in if there is a better way of doing this.
db.friends.aggregate([
{ $unwind: '$friends'},
{ $project: {
_id: 1,
name: 1,
buddy: {
nickName: '$friends.name',
age: '$friends.age',
}
}},
{ $group: {
_id: '$_id',
name: {'$first': '$name'},
buddies: {"$push": "$buddy"}
}}
]);
Below I describe the response of each operator ($unwind, $project, $group)
The $unwind would return something like this. It breaks out each Embedded Document into its own document. So where we had 1 document with 2 friends sub document, now we have two document. Both documents will have the same id and name fields but the friends sub document will be different:
{
"_id" : ObjectId("539c80d43cb9fe99d183a5f7"),
"name" : "Matthew",
"friends" : {"name" : "slim", "age" : "32"}
}, {
"_id" : : ObjectId("539c80d43cb9fe99d183a5f7"),,
"name" : "Matthew",
"friends" : {"name" : "buba", "age" : "36"}
}
The $project would return something like this. Here I'm creating a new Embedded Document called buddy and notice that I assigning "$friends.name" to "nickName":
{
"_id" : ObjectId("539c80d43cb9fe99d183a5f7"),
"name" : "Matthew",
"buddy" : {"nickName" : "slim","age" : "32"}
}, {
"_id" : : ObjectId("539c80d43cb9fe99d183a5f7"),,
"name" : "Matthew",
"friends" : {"nickName" : "buba","age" : "36"}
}
And then the $group would do something like this. Here I'm just group everything back together, but setting the array name to "buddies":
{
"_id" : ObjectId("539c80d43cb9fe99d183a5f7"),
"name" : "Matthew",
"buddies" : [
{"nickName" : "slim","age" : "32"},
{"nickName" : "buba","age" : "36"}
]
}
This seems to work, but if the friends array had duplicates, this process would remove the duplicates. This can be solved if there is a unique id for each friend.