Query on all nested docs inside a nested doc in MongoDB - mongodb

Question

I have a collection with documents that look like this:
{ keyA1: "stringVal",
keyA2: "stringVal",
keyA3: { keyB1: { feild1: intVal,
feild2: intVal}
keyB2: { feild1: intVal,
feild2: intVal}
}
}
Currently the [keyB1, keyB2, ...] set is 7 keys, same for all documents in the collection. I want to query the intVals on specific fields for all keyB's. So, for example, I might want to find all documents where field2 has value greater than 100 regardless of whcih keyB it falls in.
For any one specific keyB, I simply use the dot notation: {"keyA3.keyB2.field2": {$gte: 100}}. Right now, I have the option of looping over all keyB's, but this may not be the case in the future where more keyB values can be added. I don't want to have to modify the code then, and would like to avoid harcoding those values in anyway. I also need the solution to be fairly fast, as the final deployment is expected to have over 20M documents.
How can I write a query that can "skip" the keyB field in the dot notation and just go through all the embedded docs?
FWIW, I'm implementing this in python using pymongo. Thanks.

Answer 1

first convert keyA3 object to array and add new field with $addFields
then filter the new array to match field2 value is greater than 100
then query the doc that size of matched array is greater than 0 , then remove extra field we add
db.collection.aggregate([
{
"$addFields": {
"arr": {
"$objectToArray": "$keyA3"
}
}
},
{
"$addFields": {
"matchArrSize": {
$size: {
"$filter": {
"input": "$arr",
"as": "z",
"cond": {
$gt: [
"$$z.v.feild2",
100
]
}
}
}
}
}
},
{
$match: {
matchArrSize: {
$gt: 0
}
}
},
{
$unset: [
"arr",
"matchArrSize"
]
}
])
https://mongoplayground.net/p/VumwL9y7Km1