Currently, the $slice
operator is unavailable in the the $project
operation, of the aggregation pipeline.
So what you could do is,
First $unwind
, the my_field
array, and then group them together and take the $first
element of the group.
db.my_collection.aggregate([
{$unwind:"$my_field"},
{$group:{"_id":"$_id","resp":{$first:"$my_field"}}},
{$project:{"_id":0,"resp":1}}
])
Or using the find()
command, where you could make use of the $slice operator in the projection
part.
db.my_collection.find({},{"my_field":{$slice:1}})
Update: based on your comments, Say you want only the second
item in an array, for the record with an id, id
.
var field = 2;
var id = ObjectId("...");
Then, the below aggregation command gives you the 2nd item in the my_field
array of the record with the _id
, id
.
db.my_collection.aggregate([
{$match:{"_id":id}},
{$unwind:"$my_field"},
{$skip:field-1},
{$limit:1}
])
The above logic cannot be applied for more a record, since it would involve a $group
, operator after $unwind
. The $group
operator produces a single record for all the records in that particular group making the $limit
or $skip
operators applied in the later stages to be ineffective.
A small variation on the find()
query above would yield you the expected result as well.
db.my_collection.find({},{"my_field":{$slice:[field-1,1]}})
Apart from these, there is always a way to do it in the client side, though a bit costly if the number of records is very large:
var field = 2;
db.my_collection.find().map(function(doc){
return doc.my_field[field-1];
})
Choosing from the above options depends upon your data size and app design.