CODEWITHSUNDEEP
mongodb
trope
trope

Reputation: 537

MongoDB query help: $elemMatch in nested objects

> db.test.insert({"a" : { "b" : { "c" : { "d1" : [ "e1" ],
                                          "d2" : [ "e2" ], 
                                          "d3" : [ "e3", "e4" ], 
                                          "d4" : [ "e5", "e6" ] } } } })
> db.test.find({'a.b.c' : {$exists : true}})
{ "_id" : ObjectId("4daf2ccd697ebaacb10976ec"), "a" : { "b" : { "c" : { "d1" : [ "e1" ], "d2" : [ "e2" ], "d3" : [ "e3", "e4" ], "d4" : [ "e5", "e6" ] } } } }

But none of these work:

> db.test.find({'a.b': "c"})
> db.test.find({'a.b': {$elemMatch : {"c" : {$exists: true}}}})
> db.test.find({'a.b': {$elemMatch : {$elemMatch : {$all : ["e1"] }}}})

Suppose I don't know what the values of c and d1...d4 are. Is there a generic way to search the nested-objects' structure for particular values?

I thought that was what $elemMatch was for.

Thank you.

Upvotes: 15

Views: 17360

Answers (2)

Matt Diamond
Matt Diamond

Reputation: 11696

I believe the query you're looking for is

db.test.find({ 'a.b.c': { '$exists': true } });

To your credit, you were very close!

Upvotes: 2

Gates VP
Gates VP

Reputation: 45277

I thought that was what $elemMatch was for...

From the docs: Using the $elemMatch query operator, you can match an entire document within an array.

This does not sounds like what you're looking for.

Is there a generic way to search the nested-objects' structure for particular values?

It sounds like you want to search "everything in object 'c' for an instance of 'e1'".

MongoDB supports two related features, but the features not quite what you're looking for.

  • Reach into objects, dot notation: db.test.find({'a.b.c.d1' : 'e1'})
  • Read through arrays: `db.test.find({'a.b.c.d4' : 'e5'})

It sounds like you're looking for the ability to do both at the same time. You want to "reach into objects" and "read through arrays" in the same query.

Unfortunately, I do not know of such a feature. You may want to file a feature request for this.

Upvotes: 7

Related Questions