Database Manual / Introduction / BSON Types

Migrate Undefined Data and Queries

Starting in MongoDB 8.0, comparisons to null in equality match expressions don't match undefined values.

For example, consider these documents and query:

// create the people collection

db.people.insertMany( [
   { _id: 1, name: null },
   { _id: 2, name: undefined },
   { _id: 3, name: [ "Gabriel", undefined ] },
   { _id: 4, names: [ "Alice", "Charu" ] }
] )
db.people.find( { name: null } )

Prior to MongoDB 8.0, the preceding query would match documents where:

Starting in MongoDB 8.0, the preceding query does not match documents where the name field is undefined or contains undefined array elements. The query only matches documents where:

This query behavior change also impacts these operations:

To account for this behavior change, you can:

Note

undefined is a deprecated BSON type. Recent versions of the MongoDB Shell and drivers automatically convert undefined values to null when performing inserts and updates. The guidance on this page applies to deployments that have undefined values from older driver versions or legacy mongo shell.

Remove Undefined Fields

If you don't need to keep fields with undefined values in your documents, you can remove those fields. MongoDB's flexible data model means your collection's document fields do not need to be consistent, so you can remove a particular field from a subset of documents.

How to remove undefined fields from your documents depends on whether you know the field name to remove. If you know the field name, the operation is more performant because it can use an index.

See either:

Remove Field with Known Name

If you know the name of the field that contains undefined values that you want to remove, use the following example. The example updates the people collection to remove:

  • The name field if its value is the scalar value undefined.
  • undefined array elements in the name field.
db.people.updateMany(
   { name: { $type: "undefined" } },
   [ {
      $set: {
         "name": {
            $cond: {
               // When "name" is an array, convert { name: [ "Alice", undefined ] }
               // to { name: [ "Alice" ] }
               if: {
                  $eq: [ { $type: "$name" }, "array" ]
               },
               then: {
                  $filter: {
                     input: "$name",
                     cond: {
                        $not: { $eq: [ { $type: "$$this" }, "undefined" ] }
                     }
                  },
               },
               // When "name" is scalar undefined, remove it
               else: "$$REMOVE"
            }
         }
      }
   } ]
)

After you run the operation, the people collection contains these documents:

[
   { _id: 1, name: null },
   { _id: 2 },
   { _id: 3, name: [ "Gabriel" ] }
   { _id: 4, names: [ "Alice", "Charu" ] }
]

Remove Fields with Unknown Names

If you don't know which fields contain undefined values, use the following example to remove all undefined top-level fields.

Note

When you don't specify a field name for the update, the operation is not performant because the query can't use an index. If you run the following example on a large collection, the query might be slow and resource-intensive.

The following example removes top-level document fields from the people collection where the value is undefined:

db.people.updateMany(
   { },
   [ {
      $replaceWith: {
         // Detect undefined top-level fields under the root and remove them
         $arrayToObject: {
            $filter: {
               input: { $objectToArray: "$$ROOT" },
               cond: {
                  $not: { $eq: [ { $type: "$$this.v" }, "undefined" ] }
               }
            }
         }
      }
   } ]
)

After you run the operation, the people collection contains these documents:

[
   { _id: 1, name: null },
   { _id: 2 },
   { _id: 3, name: [ "Gabriel", undefined ] }
   { _id: 4, names: [ "Alice", "Charu" ] }
]

Note

The preceding approach only modifies top-level fields. The document with _id: 3 still contains an undefined value because the value appears in an array.

Update Undefined Values to Null

You can update undefined data values to the null data type. Use this approach to migrate your data off of the deprecated undefined data type while retaining your document fields.

How to update undefined fields depends on whether you know the field name to update. If you know the field name, the operation is more performant because it can use an index.

See either:

Update Field with Known Name

If you know the name of the field that contains undefined values that you want to set to null, use the following example. The example updates the people collection to set the following values to null:

  • The name field if its value is the scalar value undefined.
  • undefined array elements that appear in the name field.
db.people.updateMany(
   { name: { $type: "undefined" } },
   [ {
      $set: {
         "name": {
            $cond: {
               // When "name" is an array, convert { name: [ "Alice", undefined ] }
               // to { name: [ "Alice", null ] }
               if: {
                  $eq: [ { $type: "$name" }, "array" ]
               },
               then: {
                  $map: {
                     input: "$name",
                     in: {
                        $cond: {
                           if: { $eq: [ { $type: "$$this" }, "undefined" ] },
                           then: null,
                           else: "$$this"
                        }
                     }
                  },
               },
               // When "name" is the scalar undefined, convert to null
               else: null
            }
         }
      }
   } ]
)

After you run the operation, the people collection contains these documents:

[
   { _id: 1, name: null },
   { _id: 2, name: null },
   { _id: 3, name: [ "Gabriel", null ] }
   { _id: 4, names: [ "Alice", "Charu" ] }
]

Update Fields with Unknown Names

If you don't know which fields contain undefined values, use the following example to set all undefined top-level fields to null.

Note

When you don't specify a field name for the update, the operation is not performant because the query can't use an index. If you run the following example on a large collection, the query might be slow and resource-intensive.

The following example updates the people collection to set undefined top-level document fields to null:

db.people.updateMany(
   { },
   [ {
      $replaceWith: {
         // Detect undefined top-level fields under the root and replace them with null
         $arrayToObject: {
            $map: {
               input: { $objectToArray: "$$ROOT" },
               in: {
                  $cond: {
                     if: { $eq: [ { $type: "$$this.v" }, "undefined" ] },
                     then: { k: "$$this.k", v: null },
                     else: "$$this"
                  }
               }
            }
         }
      }
   } ]
)

After you run the operation, the people collection contains these documents:

[
   { _id: 1, name: null },
   { _id: 2, name: null },
   { _id: 3, name: [ "Gabriel", undefined ] }
   { _id: 4, names: [ "Alice", "Charu" ] }
]

Note

The preceding approach only modifies top-level fields. The document with _id: 3 still contains an undefined value because the value appears in an array.

Update Queries to Match Undefined Values

If you can't migrate your data types from null to undefined, you can rewrite your queries to match undefined values. If you use this approach, your data will still contain the deprecated undefined BSON type.

To have queries for null match undefined values, add a query predicate that explicitly matches the undefined type. For example, the following query matches documents where name is undefined, null, or missing:

db.people.find( {
   $or: [
      { name: null },
      { name: { $type: "undefined" } }
   ]
} )

The query returns all documents in the people collection:

[
   { _id: 1, name: null },
   { _id: 2, name: undefined },
   { _id: 3, name: [ "Gabriel", undefined ],
   { _id: 4, names: [ "Alice", "Charu" ] }
]

Learn More