I have data that has user primary keys in "user.pk", and "usertags.user.pk".
[
{
_id: ObjectId("62015250cb18d3cadb32c38e"),
user: { pk: '7485070525' },
usertags: [ { user: { pk: '334777488' } } ]
},
{
_id: ObjectId("62015250cb18d3cadb32c38f"),
user: { pk: '334777488' },
usertags: [
{ user: { pk: '7485070525' } },
{ user: { pk: '271738610' } },
{ user: { pk: '31961021' } }
]
}
]
I am looking to aggregate all this data into a single list of all distinct user primary keys, so the result for the data above would be something like { _id: null, user_pks: [7485070525, 334777488, 271738610, 31961021] }
I tried using the $group aggregation, but I can't figure out how to merge "user.pk", and "usertags.user.pk". Can anyone help me with that?
Edit: This is the closest I've gotten, which puts users and usertags in separate lists.
db.collection.aggregate([
{ $unwind: "$usertags" },
{ $group: { _id: null, users: { $push: "$user.pk" }, usertags: {$push: "$usertags.user.pk" } } },
])
you can do this:
test it here mongodb playground
db.collection.aggregate([
{
$group: {
_id: null,
user_pk: {
$push: "$user.pk"
},
user_sub_pk: {
$push: "$usertags"
}
},
},
// flattern array in user_sub_pk
{
$set: {
user_sub_pk: {
$reduce: {
input: "$user_sub_pk",
initialValue: [],
in: {
$concatArrays: [
"$$value",
"$$this"
],
},
},
},
},
},
{
$project: {
user_pks: {
$concatArrays: [
"$user_pk",
"$user_sub_pk.user.pk"
],
},
},
// remove duplicate items
{
"$project": {
"user_pks": {
"$setUnion": [
"$user_pks",
[]
]
}
}
}
}
])
update: refactored by @zrbecker
db.collection.aggregate([
{
$group: {
_id: null,
users: {
$push: "$user.pk"
},
usertags: {
$push: "$usertags.user.pk"
}
}
},
{
$project: {
users: {
$setUnion: [
"$users",
{
$reduce: {
input: "$usertags",
initialValue: [],
in: {
$concatArrays: [
"$$value",
"$$this"
]
}
}
}
]
}
}
}
])