1

I want to execute Document Nested in an Array queries with the Doctrine mongodb odm query builder

I have basically a structure similar to :

db.inventory.insertMany( [

{ item: "journal", instock: [ { warehouse: "A", qty: 5 }, { warehouse: "C", qty: 15 } ] },

{ item: "notebook", instock: [ { warehouse: "C", qty: 5 } ] },

{ item: "postcard", instock: [ { warehouse: "B", qty: 15 }, { warehouse: "C", qty: 35 } ] }

]);

I want to execute this shell query within the query builder

db.inventory.find( { "instock": { warehouse: "A", qty: 5 } } )

I have tried this, but this won't work as it ignores the field 'instock'.

$qb->field('instock')->field('warehouse')->equals('A')->field('qty')->equals(5)

Keep in mind that doing the following

$qb->field('instock.warehouse')->equals('A')->field('instock.qty')->equals(5)

is a different logic (check second example), because it matches not necessarily the same embedded document.

I can implement it using $where and a loop, but it is ugly and slow. any better ideas ?

louffi
  • 215
  • 5
  • 19

2 Answers2

3

using $elemMatch, this is the way to solve this issue : 

$qb->expr()->field('instock')->elemMatch(
        $qb->expr()
            ->field('warehouse')->equals('A')
            ->field('qty')->equals(5)
);
louffi
  • 215
  • 5
  • 19
1

ODM supports embedded documents via query builder so you can do the following:

$qb->field('instock')->equals(new InStock('A', 5));

If using objects is not what you want I think you can do away with associative array instead of new InStock().

malarzm
  • 2,831
  • 2
  • 15
  • 25
  • that way i would need the whole object data, i can't look for a specific field :( – louffi Mar 27 '18 at 15:15
  • @volkovmqx sorry then, was mislead by query using all fields that were in the data. Have you tried something like `$qb->field('instock')->equals(['warehouse' => 'A', 'qty' => 5]);`? – malarzm Mar 27 '18 at 19:28
  • yes, it will match if a whole ['warehouse' => 'A', 'qty' => 5] exists, and in the same order, so ['warehouse' => 'A'] won't work – louffi Mar 28 '18 at 10:23
  • yes, that's how i solved this issue. i will share the answer – louffi Mar 29 '18 at 16:42