$lookup¶
$lookup performs a left outer join to an unsharded collection in the
same database to filter in documents from the “joined” collection for
processing. For more information, see
$lookup. In
Atlas Data Lake, $lookup can be used to perform a join of collections
from different databases.
Syntax¶
The $lookup
syntax is
described in the MongoDB server manual. In Data Lake, the from
field in $lookup has the following alternate syntax to support specifying
an object containing an optional database name and a required collection name:
from Field Object¶
Field | Type | Description | Necessity |
|---|---|---|---|
db | string | The database name. If the database name is specified, data is read from the collection in the specified database. If you specify a database name that is different from the database upon which the command is operating, all nested $lookup stages must also specify a database name. If the database name is not specified within a $lookup stage, collections in the stage inherit the database name specified in the closest parent $lookup stage if it exists or the database upon which the command is operating. | Conditional |
coll | string | The collection name. | Required |
Examples¶
Suppose there are three databases named sourceDB1, sourceDB2, and
sourceDB3 with the following collections:
The following examples use the $lookup aggregation stage to join documents from one collection with the documents from the collection in the other databases.
Basic Example¶
The following aggregation operation on the sourceDB1.orders collection joins the documents from the orders collection with the documents
from the sourceDB2.catalog collection using the item field from the orders collection and the sku field from the catalog collection:
db.getSiblingDb("sourceDB1").orders.aggregate( { $lookup: { from: { db: "sourceDB2", coll: "catalog" }, localField: "item", foreignField: "sku", as: "inventory_docs" } } )
Nested Example¶
The following aggregation operation on the sourceDB1.orders collection joins the documents from the orders collection with the documents from the sourceDB2.catalog collection and the documents from the sourceDB3.warehouses collection using the item field from the orders collection, the sku field from the catalog collection, and the stock_item and instock fields from the warehouses collection:
db.getSiblingDb(“sourceDB1”).orders.aggregate( [ { $lookup: { from: db: "sourceDB2", coll: "catalog", let: { "order_sku": "$item" }, pipeline: [ { $match: { $expr: { $eq: ["$sku", "$$order_sku"] } } }, { $lookup: { from: db: "sourceDB3", coll: "warehouses", pipeline: [ { $match: { $expr:{ $eq : ["$stock_item", "$$order_sku"] } } }, { $project : { "instock": 1, "_id": 0} } ], as: "wh" } }, { "$unwind": "$wh" }, { $project : { "description": 1, "instock": "$wh.instock", "_id": 0} } ], as: "inventory" }, }, ] )