How to find mongo ObjectId from Map? - mongodb

Question

I am facing one issue where I need to find ObjectId which is basically a user Id and need to perform a lookup from other collection called "tb_user" from that user Id.
I tried many times but not found a optimised way.
below I is the detail what I tried.
first match the conditions.
perform projection which matches only one object from the list.
unwind subtask array
remove unwanted(only in this operation) keys form object.
again convert to array of object which contain key value pairs.
remove the values from object.
unwind again subtask.
In project take userId from keys.
convert it to objectId.
perform the lookup from tb_user collection.
at last get user information.
db.getCollection('tb_task').aggregate([
{
"$match": { "$and": [{ "_id" : ObjectId("5d766ac6f3195ed8e7361b62")},
{'sub_tasks.name': "task_1_19Z50DZYDV"}] },
},
{
"$project": {
"sub_tasks": {
"$filter": {
"input": "$sub_tasks",
"as": "sub_task",
"cond": {
"$and":{ "$eq": [
"$$sub_task.name","task_1_19Z50DZYDV"
]
}
}
}
}
}
},
{
"$unwind": {
"path": "$sub_tasks",
"preserveNullAndEmptyArrays": true
}
},
{
"$project": {
"_id":0,
"sub_tasks.name":0,
"sub_tasks.total_queries":0,
"sub_tasks.total_query_result_pairs":0,
"sub_tasks.total_assigned_to":0
}
},
{ $project: { "sub_tasks" : { $objectToArray: "$sub_tasks" } } },
{ $project: { "sub_tasks.v" : 0} },
{
"$unwind": {
"path": "$sub_tasks",
"preserveNullAndEmptyArrays": true
}
},
{ $project: { "user_id" : "$sub_tasks.k"}},
{
"$addFields": {
"user_id": {
"$toObjectId": "$user_id"
}
}
},
{ "$lookup": {
"from": "tb_user",
"localField": "user_id",
"foreignField": "_id",
"as": "user"
}
},
{
"$unwind": {
"path": "$user",
preserveNullAndEmptyArrays": true
}
},
{ $project: { "user_name" : "$user.first_name",
"is_active" : "$user.is_active",
"user_id" : "$user._id",
}
}
])
Actual collection schema
{
"_id" : ObjectId("5d766ac6f3195ed8e7361b62"),
"name" : "task_1",
"sub_tasks" : [
{
"name" : "task_1_19Z50DZYDV",
"total_queries" : 10,
"total_query_result_pairs" : 20,
"total_assigned_to" : 2,
"5d6e387524d8bd317909806a" : {
"assigned_time" : ISODate("2019-09-09T15:41:26.000Z"),
"finished_time" : null,
"is_finished" : false,
"total_queries_rated" : 0
},
"5d6e387524d8bd31790ab4" : {
"assigned_time" : ISODate("2019-09-09T15:41:26.000Z"),
"finished_time" : null,
"is_finished" : false,
"total_queries_rated" : 0
}
}
]
}
Excepted output
{
"user_name" : "Heey",
"is_active" : true,
"user_id" : ObjectId("5d6e387524d8bd317909806a")
},
{
"user_name" : "ram",
"is_active" : true,
"user_id" : ObjectId("5d6e387524d8bd31790ab4")
}

Answer 1

Since _id is unique for each document, you can apply just
{
"$match": {
"_id" : ObjectId("5d766ac6f3195ed8e7361b62")
}
}
And also it will return only one document, no need to use filter in $project instead, use :
{
$project:{
"sub_tasks":1
}
}
and after unwind match need to be added so that it will send you the documents which contain the "sub_task.name" = "task_1_19Z50DZYDV".
{
$match:{
"sub_task.name":"task_1_19Z50DZYDV"
}
}
Rest of the stages seem good to me.
So, the final query will look like this:
db.getCollection('tb_task').aggregate([
{
"$match": {
"_id" : ObjectId("5d766ac6f3195ed8e7361b62")
}
},
{
$project:{
"sub_tasks":1
}
},
{
"$unwind": {
"path": "$sub_tasks",
"preserveNullAndEmptyArrays": true
}
},
{
$match:{
"sub_task.name":"task_1_19Z50DZYDV"
}
},
{
"$project": {
"_id":0,
"sub_tasks.name":0,
"sub_tasks.total_queries":0,
"sub_tasks.total_query_result_pairs":0,
"sub_tasks.total_assigned_to":0
}
},
{ $project: { "sub_tasks" : { $objectToArray: "$sub_tasks" } } },
{ $project: { "sub_tasks.v" : 0} },
{
"$unwind": {
"path": "$sub_tasks",
"preserveNullAndEmptyArrays": true
}
},
{ $project: { "user_id" : "$sub_tasks.k"}},
{
"$addFields": {
"user_id": {
"$toObjectId": "$user_id"
}
}
},
{ "$lookup": {
"from": "tb_user",
"localField": "user_id",
"foreignField": "_id",
"as": "user"
}
},
{
"$unwind": {
"path": "$user",
preserveNullAndEmptyArrays": true
}
},
{ $project: { "user_name" : "$user.first_name",
"is_active" : "$user.is_active",
"user_id" : "$user._id",
}
}
])
And for understanding more tips of optimizing the aggregation pipelines refer here