@ladjs/mongoose-unique-validator

Mongoose plugin which adds pre-save validation for unique fields within a Mongoose schema. This makes error handling much easier, since you will get a Mongoose validation error when you attempt to violate a unique constraint, rather than an E11000 error from MongoDB. Fork of the original unmaintained package.

NOTE: As of v5.0.0+ if the only unique index is _id, then E11000 will be thrown by default. This prevents an unnecessary call to countDocuments and is a major optimization.

Table of Contents

• Install
• Usage
• Example
• Find + Updates
• Custom Error Types
• Custom Error Messages
• Case Insensitive
• Additional Conditions
• Caveats
• Contributors
• License

Install

npm:

npm install @ladjs/mongoose-unique-validator

Usage

const mongoose = require('mongoose');
const uniqueValidator = require('@ladjs/mongoose-unique-validator');

const mySchema = mongoose.Schema({
  // TODO: put your schema definition here
});

// NOTE: this should come after any indexes are added to your schema (incl from other plugins)
mySchema.plugin(uniqueValidator);

Example

Let's say you have a user schema. You can easily add validation for the unique constraints in this schema by applying the uniqueValidator plugin to your user schema:

const mongoose = require('mongoose');
const uniqueValidator = require('@ladjs/mongoose-unique-validator');

// Define your schema as normal.
const userSchema = mongoose.Schema({
  username: { type: String, required: true, unique: true },
  email: { type: String, index: true, unique: true, required: true },
  password: { type: String, required: true }
});

// Apply the uniqueValidator plugin to userSchema.
userSchema.plugin(uniqueValidator);

Now when you try to save a user, the unique validator will check for duplicate database entries and report them just like any other validation error:

const user = new User({ username: 'JohnSmith', email: '[e-mail]', password: 'j0hnNYb0i' });
await user.save();

{
  message: 'Validation failed',
  name: 'ValidationError',
  errors: {
    username: {
      message: 'Error, expected `username` to be unique. Value: `JohnSmith`',
      name: 'ValidatorError',
      kind: 'unique',
      path: 'username',
      value: 'JohnSmith'
    }
  }
}

Find + Updates

When using findOneAndUpdate and related methods, mongoose doesn't automatically run validation. To trigger this, you need to pass a configuration object. For technical reasons, this plugin requires that you also set the context option to query.

{ runValidators: true, context: 'query' }

A full example:

await User.findOneAndUpdate(
  { email: '[e-mail]' },
  { email: '[e-mail]' },
  { runValidators: true, context: 'query' }
);

Custom Error Types

You can pass through a custom error type as part of the optional options argument:

userSchema.plugin(uniqueValidator, { type: 'mongoose-unique-validator' });

After running the above example the output will be:

{
  message: 'Validation failed',
  name: 'ValidationError',
  errors: {
    username: {
      message: 'Error, expected `username` to be unique. Value: `JohnSmith`',
      name: 'ValidatorError',
      kind: 'mongoose-unique-validator',
      path: 'username',
      value: 'JohnSmith'
    }
  }
}

You can also specify a default custom error type by overriding the plugin defaults.type variable:

uniqueValidator.defaults.type = 'mongoose-unique-validator'

Custom Error Messages

You can pass through a custom error message as part of the optional options argument:

userSchema.plugin(uniqueValidator, { message: 'Error, expected {PATH} to be unique.' });

You have access to all of the standard Mongoose error message templating:

• {PATH}
• {VALUE}
• {TYPE}

You can also specify a default custom error message by overriding the plugin defaults.message variable:

uniqueValidator.defaults.message = 'Error, expected {PATH} to be unique.'

Case Insensitive

For case-insensitive matches, include the uniqueCaseInsensitive option in your schema. Queries will treat [e-mail] and [e-mail] as duplicates.

const userSchema = mongoose.Schema({
  username: { type: String, required: true, unique: true },
  email: { type: String, index: true, unique: true, required: true, uniqueCaseInsensitive: true },
  password: { type: String, required: true }
});

Additional Conditions

For additional unique-constraint conditions (ex: only enforce unique constraint on non soft-deleted records), the MongoDB option partialFilterExpression can be used.

Note: the option index must be passed as an object containing unique: true, or else partialFilterExpression will be ignored.

const userSchema = mongoose.Schema({
  username: { type: String, required: true, unique: true },
  email: {
    type: String,
    required: true,
    index: {
      unique: true,
      partialFilterExpression: { deleted: false }
    }
  },
  password: { type: String, required: true }
});

Caveats

Because we rely on async operations to verify whether a document exists in the database, it's possible for two queries to execute at the same time, both get 0 back, and then both insert into MongoDB.

Outside of automatically locking the collection or forcing a single connection, there's no real solution.

For most of our users this won't be a problem, but is an edge case to be aware of.

Contributors

Name
Mike Botsko

License

MIT © Blake Haswell