MongoDB Aggregate Listing by Arrays Value - mongodb

Question

How its making listing order by records array's value with subquery
I want best query for perform in mongo side
models.js
new ms.Schema({
name : {type: String,required: true,unique:true},
display_name: {type: String,required: true,unique:true},
url: {type: String,default:'' },
icon: {type: String,default :'no.png' },
assets : {type:Array,default : ['BTCUSDT']},
active: {type: Boolean, default : true},
})
mongodb's record goes here
{
"_id" : ObjectId("5e9e78c477b1c7a1bfc4978c"),
"url" : "https://bitso.com/",
"active" : false,
"name" : "bitso",
"display_name" : "Bitso",
"icon" : "Bitso.png",
"__v" : 0,
"seq" : 888,
"assets" : [
"BTCUSDT",
"ETHUSDT",
"LTCUSDT"
]
},
{
"_id" : ObjectId("5e9e78c377b1c7a1bfc4978a"),
"url" : "https://www.fybsg.com/",
"active" : false,
"name" : "fybsg",
"display_name" : "FYB-SG",
"icon" : "FYB-SG.png",
"__v" : 0,
"seq" : 888,
"assets" : [
"BTCUSDT",
"ETHUSDT"
]
},
{
"_id" : ObjectId("5e9e78c377b1c7a1bfc49789"),
"url" : "https://hitbtc.com/",
"active" : true,
"name" : "hitbtc",
"display_name" : "Hitbtc",
"icon" : "Hitbtc.png",
"__v" : 0,
"seq" : 99,
"assets" : [
"BCCUSDT"
]
},
{
"_id" : ObjectId("5e9e78c077b1c7a1bfc49787"),
"url" : "https://blockchain.io/",
"active" : false,
"name" : "blockchainio",
"display_name" : "Blockchain.io",
"icon" : "Blockchain.io.png",
"__v" : 0,
"seq" : 999,
"assets" : [
"BTCUSDT",
"ETHUSDT"
]
},
db.markets.aggregate([...])
if possible How can i export result like here from mongo records; Else what algorithm must i use
i need data like this
[
BTCUSDT : { which record assets have BTCUSDT listings },
ETHUSDT : { which record assets have BTCUSDT listings},
...
]

Answer 1

You need to flat assets array of each doc using $unwind stage and after that just group them by this field:
db.markets.aggregate([
{
$unwind: "$assets"
},
{
$group: {
_id: "$assets",
recordId: {
$push: "$_id"
}
}
}
])
You output will be:
[
{
"_id": "BCCUSDT",
"recordId": [
"5e9e78c377b1c7a1bfc49789"
]
},
{
"_id": "BTCUSDT",
"recordId": [
"5e9e78c477b1c7a1bfc4978c",
"5e9e78c377b1c7a1bfc4978a",
"5e9e78c077b1c7a1bfc49787"
]
},
{
"_id": "LTCUSDT",
"recordId": [
"5e9e78c477b1c7a1bfc4978c"
]
},
{
"_id": "ETHUSDT",
"recordId": [
"5e9e78c477b1c7a1bfc4978c",
"5e9e78c377b1c7a1bfc4978a",
"5e9e78c077b1c7a1bfc49787"
]
}
]
If it's more convinient for you to get only one object as the result, you can additionally group them by null id, store all the docs in array with k, v properties and replace the root of the single document:
{
$group: {
_id: null,
result: {
$push: {
k: "$_id",
v: "$recordId"
}
}
}
},
{
$addFields: {
result: {
$arrayToObject: "$result"
}
}
},
{
$replaceRoot: {
newRoot: "$result"
}
}
Result for whole query will be:
[
{
"BCCUSDT": [
"5e9e78c377b1c7a1bfc49789"
],
"BTCUSDT": [
"5e9e78c477b1c7a1bfc4978c",
"5e9e78c377b1c7a1bfc4978a",
"5e9e78c077b1c7a1bfc49787"
],
"ETHUSDT": [
"5e9e78c477b1c7a1bfc4978c",
"5e9e78c377b1c7a1bfc4978a",
"5e9e78c077b1c7a1bfc49787"
],
"LTCUSDT": [
"5e9e78c477b1c7a1bfc4978c"
]
}
]