findAndModify

lastErrorObject

The lastErrorObject embedded document contains the following fields:

value

For remove operations, value contains the removed document if the query matches a document. If the query does not match a document to remove, value contains null.

For update operations, the value embedded document contains the following:

• If the new parameter is not set or is false:

  • the pre-modification document if the query matches a document;

  • otherwise, null.

• If new is true:

  • the updated document if the query returns a match;

  • the inserted document if upsert: true and no document matches the query;

  • otherwise, null.

Upsert with Unique Index

Upserts can create duplicate documents, unless there is a unique index to prevent duplicates.

Consider an example where no document with the name Andy exists and multiple clients issue the following command at roughly the same time:

db.runCommand(
   {
     findAndModify: "people",
     query: { name: "Andy" },
     update: { $inc: { score: 1 } },
     upsert: true
   }
)

If all findAndModify operations finish the query phase before any client successfully inserts data, and there is no unique index on the name field, each findAndModify operation may result in an insert, creating multiple documents with name: Andy.

A unique index on the name field ensures that only one document is created. With a unique index in place, the multiple findAndModify operations now exhibit the following behavior:

• Exactly one findAndModify operation will successfully insert a new document.

• Other findAndModify operations either update the newly-inserted document or fail due to a unique key collision.

  In order for other findAndModify operations to update the newly-inserted document, all of the following conditions must be met:

  • The target collection has a unique index that would cause a duplicate key error.

  • The update operation is not updateMany or multi is false.

  • The update match condition is either:

    • A single equality predicate. For example { "fieldA" : "valueA" }

    • A logical AND of equality predicates. For example { "fieldA" : "valueA", "fieldB" : "valueB" }

  • The fields in the equality predicate match the fields in the unique index key pattern.

  • The update operation does not modify any fields in the unique index key pattern.

The following table shows examples of upsert operations that, when a key collision occurs, either result in an update or fail.

{ name : 1 }

db.people.updateOne(
   { name: "Andy" },
   { $inc: { score: 1 } },
   { upsert: true }
)

{ name : 1 }

db.people.updateOne(
   { name: { $ne: "Joe" } },
   { $set: { name: "Andy" } },
   { upsert: true }
 )

{ name : 1 }

db.people.updateOne(
  { name: "Andy", email: "andy@xyz.com" },
  { $set: { active: false } },
  { upsert: true }
)

Sharded Collections

To use findAndModify on a sharded collection, the query filter must include an equality condition on the shard key.

Starting in version 4.4, documents in a sharded collection can be missing the shard key fields. To target a document that is missing the shard key, you can use the null equality match in conjunction with another filter condition (such as on the _id field). For example:

{ _id: <value>, <shardkeyfield>: null } // _id of the document missing shard key

Document Validation

The findAndModify command adds support for the bypassDocumentValidation option, which lets you bypass document validation when inserting or updating documents in a collection with validation rules.

Comparisons with the update Method

When updating a document, findAndModify and the update() method operate differently:

• By default, both operations modify a single document. However, the update() method with its multi option can modify more than one document.

• If multiple documents match the update criteria, for findAndModify, you can specify a sort to provide some measure of control on which document to update.

  With the default behavior of the update() method, you cannot specify which single document to update when multiple documents match.

• By default, findAndModify returns an object that contains the pre-modified version of the document, as well as the status of the operation. To obtain the updated document, use the new option.

  The update() method returns a WriteResult() object that contains the status of the operation. To return the updated document, use the find() method. However, other updates may have modified the document between your update and the document retrieval. Also, if the update modified only a single document but multiple documents matched, you will need to use additional logic to identify the updated document.

When modifying a single document, both findAndModify and the update() method atomically update the document. See Atomicity and Transactions for more details about interactions and order of operations of these methods.

Transactions

findAndModify can be used inside distributed transactions.

Update and Return

The following command updates an existing document in the people collection where the document matches the query criteria:

db.runCommand(
   {
     findAndModify: "people",
     query: { name: "Tom", state: "active", rating: { $gt: 10 } },
     sort: { rating: 1 },
     update: { $inc: { score: 1 } }
   }
)

This command performs the following actions:

1. The query finds a document in the people collection where the name field has the value Tom, the state field has the value active and the rating field has a value greater than 10.

2. The sort orders the results of the query in ascending order. If multiple documents meet the query condition, the command will select for modification the first document as ordered by this sort.

3. The update increments the value of the score field by 1.

4. The command returns a document with the following fields:

   • The lastErrorObject field that contains the details of the command, including the field updatedExisting which is true, and

   • The value field that contains the original (i.e. pre-modification) document selected for this update:

     {
       "lastErrorObject" : {
          "connectionId" : 1,
          "updatedExisting" : true,
          "n" : 1,
          "syncMillis" : 0,
          "writtenTo" : null,
          "err" : null,
          "ok" : 1
       },
       value" : {
         "_id" : ObjectId("54f62d2885e4be1f982b9c9c"),
         "name" : "Tom",
         "state" : "active",
         "rating" : 100,
         "score" : 5
       },
       "ok" : 1
     }

To return the updated document in the value field, add the new:true option to the command.

If no document match the query condition, the command returns a document that contains null in the value field:

{ "value" : null, "ok" : 1 }

The mongo shell and many drivers provide a findAndModify() helper method. Using the shell helper, this previous operation can take the following form:

db.people.findAndModify( {
   query: { name: "Tom", state: "active", rating: { $gt: 10 } },
   sort: { rating: 1 },
   update: { $inc: { score: 1 } }
} );

However, the findAndModify() shell helper method returns only the unmodified document, or if new is true, the updated document.

{
   "_id" : ObjectId("54f62d2885e4be1f982b9c9c"),
   "name" : "Tom",
   "state" : "active",
   "rating" : 100,
   "score" : 5
}

upsert: true

The following findAndModify command includes the upsert: true option for the update operation to either update a matching document or, if no matching document exists, create a new document:

db.runCommand(
               {
                 findAndModify: "people",
                 query: { name: "Gus", state: "active", rating: 100 },
                 sort: { rating: 1 },
                 update: { $inc: { score: 1 } },
                 upsert: true
               }
             )

If the command finds a matching document, the command performs an update.

If the command does not find a matching document, the update with upsert: true operation results in an insertion and returns a document with the following fields:

• The lastErrorObject field that contains the details of the command, including the field upserted that contains the _id value of the newly inserted document, and

• The value field containing null.

{
  "value" : null,
  "lastErrorObject" : {
     "updatedExisting" : false,
     "n" : 1,
     "upserted" : ObjectId("54f62c8bc85d4472eadea26f")
  },
  "ok" : 1
}

Return New Document

The following findAndModify command includes both upsert: true option and the new:true option. The command either updates a matching document and returns the updated document or, if no matching document exists, inserts a document and returns the newly inserted document in the value field.

In the following example, no document in the people collection matches the query condition:

db.runCommand(
   {
     findAndModify: "people",
     query: { name: "Pascal", state: "active", rating: 25 },
     sort: { rating: 1 },
     update: { $inc: { score: 1 } },
     upsert: true,
     new: true
   }
)

The command returns the newly inserted document in the value field:

{
  "lastErrorObject" : {
     "connectionId" : 1,
     "updatedExisting" : false,
     "upserted" : ObjectId("54f62bbfc85d4472eadea26d"),
     "n" : 1,
     "syncMillis" : 0,
     "writtenTo" : null,
     "err" : null,
     "ok" : 1
  },
  "value" : {
     "_id" : ObjectId("54f62bbfc85d4472eadea26d"),
     "name" : "Pascal",
     "rating" : 25,
     "state" : "active",
     "score" : 1
  },
  "ok" : 1
}

Sort and Remove

By including a sort specification on the rating field, the following example removes from the people collection a single document with the state value of active and the lowest rating among the matching documents:

db.runCommand(
   {
     findAndModify: "people",
     query: { state: "active" },
     sort: { rating: 1 },
     remove: true
   }
)

The command returns the deleted document:

{
  "lastErrorObject" : {
     "connectionId" : 1,
     "n" : 1,
     "syncMillis" : 0,
     "writtenTo" : null,
     "err" : null,
     "ok" : 1
  },
  "value" : {
     "_id" : ObjectId("54f62a6785e4be1f982b9c9b"),
     "name" : "XYZ123",
     "score" : 1,
     "state" : "active",
     "rating" : 3
  },
  "ok" : 1
}

Specify Collation

Collation allows users to specify language-specific rules for string comparison, such as rules for lettercase and accent marks.

A collection myColl has the following documents:

{ _id: 1, category: "café", status: "A" }
{ _id: 2, category: "cafe", status: "a" }
{ _id: 3, category: "cafE", status: "a" }

The following operation includes the collation option:

db.runCommand(
   {
     findAndModify: "myColl",
     query: { category: "cafe", status: "a" },
     sort: { category: 1 },
     update: { $set: { status: "Updated" } },
     collation: { locale: "fr", strength: 1 }
   }
)

The operation returns the following document:

{
   "lastErrorObject" : {
      "updatedExisting" : true,
      "n" : 1
   },
   "value" : {
      "_id" : 1,
      "category" : "café",
      "status" : "A"
   },
   "ok" : 1
}

Array Update Operations with arrayFilters

Starting in MongoDB 3.6, when updating an array field, you can specify arrayFilters that determine which array elements to update.

db.students.insert([
   { "_id" : 1, "grades" : [ 95, 92, 90 ] },
   { "_id" : 2, "grades" : [ 98, 100, 102 ] },
   { "_id" : 3, "grades" : [ 95, 110, 100 ] }
])

db.runCommand(
   {
     findAndModify: "students",
     query: { grades: { $gte: 100 } },
     update:  { $set: { "grades.$[element]" : 100 } },
     arrayFilters: [ { "element": { $gte: 100 } } ]
   }
)

{ "_id" : 1, "grades" : [ 95, 92, 90 ] }
{ "_id" : 2, "grades" : [ 98, 100, 100 ] }
{ "_id" : 3, "grades" : [ 95, 110, 100 ] }

db.students2.insert([
   {
      "_id" : 1,
      "grades" : [
         { "grade" : 80, "mean" : 75, "std" : 6 },
         { "grade" : 85, "mean" : 90, "std" : 4 },
         { "grade" : 85, "mean" : 85, "std" : 6 }
      ]
   },
   {
      "_id" : 2,
      "grades" : [
         { "grade" : 90, "mean" : 75, "std" : 6 },
         { "grade" : 87, "mean" : 90, "std" : 3 },
         { "grade" : 85, "mean" : 85, "std" : 4 }
      ]
   }
])

db.runCommand(
   {
     findAndModify: "students2",
     query: { _id : 1 },
     update: { $set: { "grades.$[elem].mean" : 100 } },
     arrayFilters: [ { "elem.grade": { $gte: 85 } } ]
   }
)

{
   "_id" : 1,
   "grades" : [
      { "grade" : 80, "mean" : 75, "std" : 6 },
      { "grade" : 85, "mean" : 100, "std" : 4 },
      { "grade" : 85, "mean" : 100, "std" : 6 }
    ]
}
{
   "_id" : 2,
   "grades" : [
      { "grade" : 90, "mean" : 75, "std" : 6 },
      { "grade" : 87, "mean" : 90, "std" : 3 },
      { "grade" : 85, "mean" : 85, "std" : 4 }
   ]
}

Use an Aggregation Pipeline for Updates

Starting in MongoDB 4.2, findAndModify can accept an aggregation pipeline for the update. The pipeline can consist of the following stages:

• $addFields and its alias $set

• $project and its alias $unset

• $replaceRoot and its alias $replaceWith.

Using the aggregation pipeline allows for a more expressive update statement, such as expressing conditional updates based on current field values or updating one field using the value of another field(s).

For example, create a collection students2 with the following documents:

db.students2.insert([
   {
      "_id" : 1,
      "grades" : [
         { "grade" : 80, "mean" : 75, "std" : 6 },
         { "grade" : 85, "mean" : 90, "std" : 4 },
         { "grade" : 85, "mean" : 85, "std" : 6 }
      ]
   },
   {
      "_id" : 2,
      "grades" : [
         { "grade" : 90, "mean" : 75, "std" : 6 },
         { "grade" : 87, "mean" : 90, "std" : 3 },
         { "grade" : 85, "mean" : 85, "std" : 4 }
      ]
   }
])

The following operation finds a document where the _id field equals 1 and uses an aggregation pipeline to calculate a new field total from the grades field:

db.runCommand(
   {
     findAndModify: "students2",
     query: {  "_id" : 1 },
     update: [ { $set: { "total" : { $sum: "$grades.grade" } } } ],
     new: true
   }
)

After the operation, the collection has the following documents:

{
  "_id" : 1,
  "grades" : [ { "grade" : 80, "mean" : 75, "std" : 6 }, { "grade" : 85, "mean" : 90, "std" : 4 }, { "grade" : 85, "mean" :85, "std" : 6 } ],
  "total" : 250
}
{
   "_id" : 2,
   "grades" : [ { "grade" : 90, "mean" : 75, "std" : 6 }, { "grade" : 87, "mean" : 90, "std" : 3 }, { "grade" : 85, "mean" : 85,"std" : 4 } ]
}

Specify hint for findAndModify Operations

From the mongo shell, create a members collection with the following documents:

db.members.insertMany([
   { "_id" : 1, "member" : "abc123", "status" : "P", "points" :  0,  "misc1" : null, "misc2" : null },
   { "_id" : 2, "member" : "xyz123", "status" : "A", "points" : 60,  "misc1" : "reminder: ping me at 100pts", "misc2" : "Some random comment" },
   { "_id" : 3, "member" : "lmn123", "status" : "P", "points" :  0,  "misc1" : null, "misc2" : null },
   { "_id" : 4, "member" : "pqr123", "status" : "D", "points" : 20,  "misc1" : "Deactivated", "misc2" : null },
   { "_id" : 5, "member" : "ijk123", "status" : "P", "points" :  0,  "misc1" : null, "misc2" : null },
   { "_id" : 6, "member" : "cde123", "status" : "A", "points" : 86,  "misc1" : "reminder: ping me at 100pts", "misc2" : "Some random comment" }
])

Create the following indexes on the collection:

db.members.createIndex( { status: 1 } )
db.members.createIndex( { points: 1 } )

The following operation explicitly hints to use the index { status: 1 }:

db.runCommand({
   findAndModify: "members",
   query: { "points": { $lte: 20 }, "status": "P" },
   remove: true,
   hint: { status: 1 }
})

To see the index used, run explain on the operation:

db.runCommand(
   {
     explain: {
       findAndModify: "members",
       query: { "points": { $lte: 20 }, "status": "P" },
       remove: true,
       hint: { status: 1 }
     },
     verbosity: "queryPlanner"
   }
)