/ /

Query for and Modify Valid or Invalid Documents

If you add validation to your collection after you create it, or modify an existing validation schema, you may have invalid documents in your collection. Similarly, if your schema's validationAction is warn, your collection is allowed to contain invalid documents. You can query for invalid documents to potentially update or delete them from your collection.

To find documents that either match or don't match a specified schema, use $jsonSchema with query operators. Similarly, you can update or delete documents based on a schema by using $jsonSchema in query conditions for write operations.

Examples

Create a sample collection inventory with the following documents:

db.inventory.insertMany( [
 { item: "journal", qty: 25, size: { h: 14, w: 21, unit: "cm" }, instock: true },
 { item: "notebook", qty: 50, size: { h: 8.5, w: 11, unit: "in" }, instock: true },
 { item: "paper", qty: 100, size: { h: 8.5, w: 11, unit: "in" }, instock: 1 },
 { item: "apple", qty: 45, status: "A", instock: true },
 { item: "pears", qty: 50, status: "A", instock: true }
] )

Define a Schema Object

Define a sample schema object and store it in a variable called myschema:

let myschema =
{
 $jsonSchema: {
 required: [ "item", "qty", "instock" ],
 properties: {
 item: { bsonType: "string" },
 qty: { bsonType: "int" },
 size: {
 bsonType: "object",
 required: [ "unit" ],
 properties: {
 unit: { bsonType: "string" },
 h: { bsonType: "double" },
 w: { bsonType: "double" }
 }
 },
 instock: { bsonType: "bool" }
 }
 }
}

The schema enforces the following validation:

Required fields:
- item must be BSON type string.
- qty must be BSON type integer.
- instock must be BSON type boolean.
size, if present:
- Must be BSON type object.
- Must include unit as a required string field.
- If the embedded h and w fields are present, they must be type double.

Find Documents that Match the Schema

These commands return all documents that match the schema:

db.inventory.find(myschema)
db.inventory.aggregate( [ { $match: myschema } ] )

Both commands return the same result:

[
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a3"),
 item: 'apple',
 qty: 45,
 status: 'A',
 instock: true
 },
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a4"),
 item: 'pears',
 qty: 50,
 status: 'A',
 instock: true
 }
]

Find Documents that Don't Match the Schema

To find documents in a collection that don't match the schema validation rules, use $jsonSchema with the $nor operator. For example:

db.inventory.find( { $nor: [ myschema ] } )

Output:

[
 // Neither size.h nor size.w are type double
 {
 _id: ObjectId("62b5cd5a14b92d148400f79e"),
 item: 'journal',
 qty: 25,
 size: { h: 14, w: 21, unit: 'cm' },
 instock: true
 },
 // size.w is not a double
 {
 _id: ObjectId("62b5cd5a14b92d148400f79f"),
 item: 'notebook',
 qty: 50,
 size: { h: 8.5, w: 11, unit: 'in' },
 instock: true
 },
 // size.w is not a double and instock is not a boolean
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a0"),
 item: 'paper',
 qty: 100,
 size: { h: 8.5, w: 11, unit: 'in' },
 instock: 1
 }
]

Update Documents that Don't Match the Schema

This command updates all documents that don't match the schema and sets the documents' isValid field to false:

db.inventory.updateMany(
 {
 $nor: [ myschema ]
 },
 {
 $set: { isValid: false }
 }
)

To verify the update, query the collection:

db.inventory.find()

Output:

[
 {
 _id: ObjectId("62b5cd5a14b92d148400f79e"),
 item: 'journal',
 qty: 25,
 size: { h: 14, w: 21, unit: 'cm' },
 instock: true,
 isValid: false
 },
 {
 _id: ObjectId("62b5cd5a14b92d148400f79f"),
 item: 'notebook',
 qty: 50,
 size: { h: 8.5, w: 11, unit: 'in' },
 instock: true,
 isValid: false
 },
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a0"),
 item: 'paper',
 qty: 100,
 size: { h: 8.5, w: 11, unit: 'in' },
 instock: 1,
 isValid: false
 },
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a3"),
 item: 'apple',
 qty: 45,
 status: 'A',
 instock: true
 },
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a4"),
 item: 'pears',
 qty: 50,
 status: 'A',
 instock: true
 }
]

Delete Documents that Don't Match the Schema

This command deletes all documents that don't match the schema:

db.inventory.deleteMany( { $nor: [ myschema ] } )

To verify the update, query the collection:

db.inventory.find()

Output:

[
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a3"),
 item: 'apple',
 qty: 45,
 status: 'A',
 instock: true
 },
 {
 _id: ObjectId("62b5cd5a14b92d148400f7a4"),
 item: 'pears',
 qty: 50,
 status: 'A',
 instock: true
 }
]

Learn More

Back

Modify Rules

Specify Validation for Polymorphic Collections