I'm kinda new to no-sql databases, so I have a question on subqueries here.
Let's imagine the following structure:
Type (_id, offerId)
Offer (_id, typeId, productId)
Product (_id, subId)
I need to find all types by subId.
I have no idea on how does it work for MongoDB, in SQL I'd do something like:
select * from Type where offerId in
(select _id from Offer where productId in
(select _id from Product where subId = 'test'));
For MongoDB I tried to create some kind of aggregation query, but it doesn't work:
{
"aggregate": "Type",
"pipeline": [
{
"$lookup": {
"from": "Offer",
"localField": "_id",
"foreignField": "typeId",
"as": "subOffer"
}
},
{
"$lookup": {
"from": "Product",
"localField": "_id",
"foreignField": "subOffer.productId",
"as": "subProduct"
}
},
{
"$match": {
"subProduct.subId": "test"
}
},
{
"$unwind": "$subProduct"
},
{
"$unwind": "$subOffer"
}
]
}
Any suggestions here?
You can try,
$lookup
on offer
collection using pipeline$match
type id$lookup
on product
collection using pipeline$match
fields subId
and productId
$match
product is not []
empty$match
offer is not []
empty$project
remove offer fielddb.type.aggregate([
{
$lookup: {
from: "offer",
let: { tid: "$_id" },
as: "offer",
pipeline: [
{ $match: { $expr: { $eq: ["$$tid", "$typeId"] } } },
{
$lookup: {
from: "product",
as: "product",
let: { pid: "$productId" },
pipeline: [
{
$match: {
$and: [
{ subId: "test" },
{ $expr: { $eq: ["$_id", "$$pid"] } }
]
}
}
]
}
},
{ $match: { product: { $ne: [] } } }
]
}
},
{ $match: { offer: { $ne: [] } } },
{ $project: { offer: 0 } }
])