Compound Operations
On this page
Overview
Most database requests either read data from a database or write data into a database. However, there are instances where you may require a single operation that reads and writes data.
Compound operations combine read and write operations in a single atomic statement, so there's no chance of data changing in between a read and a subsequent write.
If you execute each operation separately, another request may alter the data between the read and write operations. These data changes may not prevent your operation from succeeding, but they can make error handling more difficult. When your application handles potential errors at any stage of the process, it can become brittle and difficult to test.
Built-in Methods
The Node.js driver provides the following methods to perform compound operations:
These methods accept an optional options
object with
configurable sort and
projection options.
You can also set the includeResultMetadata
option to specify the return type of each
of these methods. To learn more about this option, see the
includeResultMetadata Option
section of this guide.
The findOneAndUpdate()
and findOneAndDelete()
methods take the
returnDocument
setting, which specifies if the method returns the
pre-update or post-update version of the modified document.
includeResultMetadata Option
The includeResultMetadata
option determines the return type of the
compound methods.
This setting defaults to false
, which means that each method returns the matched
document. If no document is matched, each method returns null
. If you set
includeResultMetadata
to true
, the method returns a ModifyResult
type that
contains the found document and metadata.
Suppose a collection contains only the following document:
{ _id: 1, x: "on" }
The following table shows how the value of the
includeResultMetadata
option changes the return type of
the findOneAndDelete()
method:
Option Value | Syntax and Output | ||||
---|---|---|---|---|---|
Default: | Document matched
No document matched
| ||||
|
|