Mongodb processing in stored procedures - BSON to JSON in javascript

Question

I've a got a mongodb database, which contains records of domains, and the integer number of 'hits' each has been seen.

record =
{
"dom" : "Somedomain.com",
"hits" : 23
}

I want to write some mongodb stored procedures to process these records on the server. MongoDb can store procedures in javascript.

I want to (to give a trivial example) get the aggregate of the number of hits for two domains. In Python/pymongo I can do this easily - I can get the db.find({"dom": "domainname.com"}) the records for the two domains, parse the returned record as JSON, extract the hit numbers, and add them.

However, I can't seem to do the same in javascript. Part of the problem seems to be that find return BSON, not JSON, and there seems to be no native way to convert it.

> This is in the Mongo Shell, which is mostly javascript.

> var r = db.test_collection.find({dom:"somedomain.com"},{hits:true,_id:false})
> r
{ "hits" : 1043 }
> var j = JSON.parse(r)
2016-03-03T17:27:52.046-0500 SyntaxError: Unexpected token D

> var r = db.test_collection.find({dom:"somedomain.com"},{hits:true,_id:false}).valueOf() 
> r
{ "hits" : 1043 }
> var j = JSON.parse(r)
2016-03-03T17:27:52.046-0500 SyntaxError: Unexpected token D

The source of this 'SyntaxError' seems to be BSON vs JSON

> var r = db.test_collection.find({dom:"somedomain.com"},{hits:true,_id:false}).valueOf()
> r.hits
>

I can't get the 'hits' value out where I can use it.

I can do this trivially in Python with Pymongo, but not in mongo shell, nor in javascript.

I want the stored javascript procedures to be able to retrieve values from the db, and process them, finally returning a result. But I can't get to the actual value (hit number, in this case) to do so. Anyone?

Answer 1

I think MongoDB aggregation might be what you're looking for.

You'd do something like this:

db.test_collection.aggregate( [
   { $group: { _id: "$dom", total: { $sum: "$hits" } } },
   { $match: { dom: { $or: ['somedomain.com', 'anotherdomain.com'] } } }
] )

The above is just a simple example, and I don't know the exact structure of all your data, so you may have to tweak it a bit. The MongoDB docs provide a great aggregation example related to zipcodes and city population:

db.zipcodes.aggregate( [
   { $group: { _id: { state: "$state", city: "$city" }, pop: { $sum: "$pop" } } },
   { $group: { _id: "$_id.state", avgCityPop: { $avg: "$pop" } } }
] )

Also, keep in mind, you shouldn't have to do var j = JSON.parse(r). MongoDB queries return plain old JavaScript objects.