Reputation: 1946
Using MongoDB, How would I write this regular SQL statement?
SELECT * FROM table WHERE (field1+field2+field3) > 1
I've been messing with $group, $project, $add, etc. I feel like I'm dancing all around the solution but can't figure it out.
Upvotes: 6
Views: 7516
Reputation: 91
It's an old post but this might help someone looking for other solutions. I found this to be even simpler than both $where
and .aggregate()
.
> db.foo.insert({"a": 17, "b": 8})
> db.foo.find({$expr: {$gt: [{$add: ["$a", "$b"]}, 25]}}) // no result
> db.foo.find({$expr: {$gt: [{$add: ["$a", "$b"]}, 24]}})
{ "_id" : ObjectId("602acf55a69fb564c11af7db"), "a" : 17, "b" : 8 }
Upvotes: 3
Reputation: 11671
> db.test.drop()
> db.test.insert({ "_id" : 0, "a" : 1, "b" : 1, "c" : 1 })
> db.test.insert({ "_id" : 1, "a" : 1, "b" : 1, "c" : 2 })
> db.test.aggregate([
{ "$project" : {
"sum" : { "$add" : ["$a", "$b", "$c" ] }
} },
{ "$match" : {
"sum" : { "$gte" : 4 }
} }
])
{ "_id" : 1, "sum" : 4 }
Upvotes: 5
Reputation: 222751
The easiest way to do this is by using $where (I am not telling that it is not possible to do this with aggregation)
db.table.find({$where: function() {
return this.field1 + this.field2 + this.field3 > 1
// most probably you have to handle additional cases if some of the fields do not exist.
}}
The pros of it is that it is easy and intuitive, whereas cons:
requires that the database processes the JavaScript expression or function for each document in the collection.
If you need to perform this kind of searches often, I would go ahead and create a new field which will have a sum of 3 fields stored in it and put an index on it. The downside is that you have to increase your app logic.
Upvotes: 5