Mongodb aggregation nested lookup - mongodb

Question

I have 3 collections as follows:
GroupRoles Collection:
//1
{
"_id": ObjectId("62a384ee0c4dbafc64000fba"),
"name": "GroupRole template 1",
"groupRoles": [
{
"_id": ObjectId("6298503f8a31000024002107"),
"members": [
ObjectId("629e1bb117366c39bc7d78e1")
]
},
{
"_id": ObjectId("629850368a31000024002106"),
"members": [
ObjectId("629ee6d502877d0f93f5dabe"),
ObjectId("629ee6d002877d0f93f5dab8")
]
},
{
"_id": ObjectId("6298502f8a31000024002105"),
"members": [ ]
},
{
"_id": ObjectId("629850288a31000024002104"),
"members": [ ]
},
{
"_id": ObjectId("6298501f8a31000024002103"),
"members": [ ]
},
{
"_id": ObjectId("629850128a31000024002102"),
"members": [ ]
}
]
}
Role Collections:
// 1 {"_id": ObjectId("629850128a31000024002102"), "type": 1, "name": "Role 1",}
// 2 {"_id": ObjectId("6298501f8a31000024002103"), "type": 1, "name": "Role 2",}
// 3 {"_id": ObjectId("629850288a31000024002104"), "type": 1, "name": "Role 3",}
// 4 {"_id": ObjectId("6298502f8a31000024002105"), "type": 1, "name": "Role 4",}
// 5 {"_id": ObjectId("629850368a31000024002106"), "type": 1, "name": "Role 5",}
// 6 {"_id": ObjectId("6298503f8a31000024002107"), "type": 1, "name": "Role 6",}
and User Collection:
// 1 {"_id": ObjectId("629e1bb117366c39bc7d78e1"), "email": "abc1#gmail.com", "name": "user 01"}
// 2 {"_id": ObjectId("629ee6d502877d0f93f5dabe"), "email": "abc2#gmail.com", "name": "user 02"}
// 3 {"_id": ObjectId("629ee6d002877d0f93f5dab8"), "email": "abc3#gmail.com", "name": "user 03"}
How can select GroupRoles with output format like this:
{
"_id": ObjectId("62a384ee0c4dbafc64000fba"),
"name": "GroupRole template 1",
"groupRoles": [
{
"_id": ObjectId("6298503f8a31000024002107"),
"type": 1,
"name": "Role 6",
"members": [
{
"_id": ObjectId("629e1bb117366c39bc7d78e1"),
"email": "abc1#gmail.com",
"name": "user 01"
}
]
},
{
"_id": ObjectId("629850368a31000024002106"),
"type": 1,
"name": "Role 5",
"members": [
{
"_id": ObjectId("629ee6d502877d0f93f5dabe"),
"email": "abc2#gmail.com",
"name": "user 02"
},
{
"_id": ObjectId("629ee6d002877d0f93f5dab8"),
"email": "abc3#gmail.com",
"name": "user 03"
}
]
},
{
"_id": ObjectId("6298502f8a31000024002105"),
"type": 1,
"name": "Role 4",
"members": [ ]
},
{
"_id": ObjectId("629850288a31000024002104"),
"type": 1,
"name": "Role 3",
"members": [ ]
},
{
"_id": ObjectId("6298501f8a31000024002103"),
"type": 1,
"name": "Role 2",
"members": [ ]
},
{
"_id": ObjectId("629850128a31000024002102"),
"type": 1,
"name": "Role 1",
"members": [ ]
}
]
}
I've tried aggregation and lookup, but it didn't get the results I wanted.
here is my aggregation query
The list of members is not in the correct index of groupRoles
[
{
'$lookup': {
'from': 'roles',
'localField': 'groupRoles._id',
'foreignField': '_id',
'as': 'temp_roles'
}
}, {
'$lookup': {
'from': 'users',
'localField': 'groupRoles.members',
'foreignField': '_id',
'as': 'temp_users'
}
}, {
'$addFields': {
'groupRoles': {
'$map': {
'input': '$groupRoles.members',
'as': 'mems',
'in': {
'members': {
'$filter': {
'input': '$temp_users',
'as': 'temu',
'cond': {
'$in': [
'$$temu._id', '$$mems'
]
}
}
}
}
}
}
}
}, {
'$addFields': {
'groupRoles': {
'$map': {
'input': {
'$zip': {
'inputs': [
'$groupRoles', '$temp_roles'
]
}
},
'in': {
'$mergeObjects': '$$this'
}
}
}
}
}, {
'$unset': [
'temp_roles', 'temp_users'
]
}, {
'$facet': {
'metadata': [
{
'$group': {
'_id': null,
'total': {
'$sum': 1
}
}
}
],
'data': []
}
}, {
'$project': {
'data': {
'_id': 1,
'name': 1,
'groupRoles': {
'_id': 1,
'type': 1,
'name': 1,
'members': {
'_id': 1,
'name': 1,
'email': 1
}
}
},
'total': {
'$arrayElemAt': [
'$metadata.total', 0
]
}
}
}
]
please help me find the solution.

Answer 1

Revise your query and thanks for the hints.
The issue is on $zip, which I think this operator is not suitable. You can check the behavior of $zip.
Combine both $addFields stages into one.
Iterate each value in groupRoles and with $mergeObjects for
Current iterate value.
The first match document from the temp_roles array by matching _id.
The document with members array.
db.groupRoles.aggregate([
{
"$lookup": {
"from": "roles",
"localField": "groupRoles._id",
"foreignField": "_id",
"as": "temp_roles"
}
},
{
"$lookup": {
"from": "users",
"localField": "groupRoles.members",
"foreignField": "_id",
"as": "temp_users"
}
},
{
"$addFields": {
"groupRoles": {
"$map": {
"input": "$groupRoles",
"as": "gr",
"in": {
"$mergeObjects": [
"$$gr",
{
$first: {
"$filter": {
"input": "$temp_roles",
"as": "r",
"cond": {
$eq: [
"$$gr._id",
"$$r._id"
]
}
}
}
},
{
"members": {
"$filter": {
"input": "$temp_users",
"as": "mem",
"cond": {
"$in": [
"$$mem._id",
"$$gr.members"
]
}
}
}
}
]
}
}
}
}
},
{
"$unset": [
"temp_roles",
"temp_users"
]
},
])
Sample Mongo Playground