How To Count Values Inside Deeply Nested Array Of Objects in Mongodb Aggregation - mongodb

Question

I want to sum of values inside array of objects which have another array of objects.
In my case; how can I count 'url' values in all documents inside 'urls' array under 'iocs' array;
Mongo playground: open
Here is document example;
[
{
"_id": {
"$oid": "63b4993d0625ebe8b6f5b06e"
},
"iocs": [
{
"urls": [
{
"url": "7.1.5.2",
}
],
},
{
"urls": [
{
"url": "https://l-ink.me/GeheimeBegierde",
},
{
"url": "GeheimeBegierde.ch",
}
],
},
{
"urls": [
{
"url": "https://l-ink.me/GeheimeBegierde",
}
],
}
],
type: "2"
},
{
"_id": {
"$oid": "63b4993d0624ebe8b6f5b06e"
},
"iocs": [
{
"urls": [
{
"url": "7.1.5.2",
}
],
},
{
"urls": [
{
"url": "https://l-ink.me/GeheimeBegierde",
},
{
"url": "GeheimeBegierde.ch",
}
],
},
{
"urls": [
{
"url": "https://l-ink.me/GeheimeBegierde",
}
],
}
],
type: "3"
},
{
"_id": {
"$oid": "63b4993d0615ebe8b6f5b06e"
},
"iocs": [
{
"urls": [
{
"url": "www.google.com",
}
],
},
{
"urls": [
{
"url": "abc.xyz",
},
{
"url": "GeheimeBegierde.ch",
}
],
},
{
"urls": [
{
"url": "https://123.12",
}
],
}
],
type: "1"
}
]
expected output be like;
url: "7.1.5.2",
count:2,
types:[2,3]
url: "https://l-ink.me/GeheimeBegierde",
count:4,
types:[2,3],
url: "abc.xyz",
count:1,
types:[1],
I tried unwind iocs then project urls but can't figure out how to get this output. I think i must use group but how ? Newbie in mongodb.
Any help would be appreciated. Thanks all.
NOTE: All the answers are working. Thank you all for the contributing.

Answer 1

You could do something like this !
db.collection.aggregate([
{
"$unwind": "$iocs"
},
{
"$unwind": "$iocs.urls"
},
{
"$group": {
"_id": "$iocs.urls.url",
"count": {
"$sum": 1
},
"types": {
"$addToSet": "$type"
}
}
},
{
"$project": {
url: "$_id",
_id: 0,
types: 1,
count: 1
}
},
])
https://mongoplayground.net/p/hhMqh2zI_SX

Answer 2

Here's one way you could do it.
db.collection.aggregate([
{"$unwind": "$iocs"},
{"$unwind": "$iocs.urls"},
{
"$group": {
"_id": "$iocs.urls.url",
"count": {"$count": {}},
"types": {"$addToSet": {"$toInt": "$type"}}
}
},
{
"$set": {
"url": "$_id",
"_id": "$$REMOVE"
}
}
])
Try it on mongoplayground.net.

Answer 3

You can try this query:
Double $unwind to deconstruct the nested array.
Then group by url get the count using $sum nad add the types into a set (to avoid duplicates, otherwise you can use simply $push)
db.collection.aggregate([
{
"$unwind": "$iocs"
},
{
"$unwind": "$iocs.urls"
},
{
"$group": {
"_id": "$iocs.urls.url",
"count": {
"$sum": 1
},
"types": {
"$addToSet": "$type"
}
}
}
])
Example here

Answer 4

Since $unwind is considered an inefficient operation, another option is to use $reduce and only $unwind once:
db.collection.aggregate([
{$project: {
type: 1,
urls: {
$reduce: {
input: "$iocs",
initialValue: [],
in: {$concatArrays: ["$$value", "$$this.urls.url"]}
}
}
}},
{$unwind: "$urls"},
{$group: {
_id: "$urls",
type: {$addToSet: "$type"},
count: {$sum: 1}
}},
{$project: {url: "$_id", count: 1, type: 1}}
])
See how it works on the playground example