Skip to content

Joi validation for your Mongoose models without the hassle of maintaining two schemas

License

Notifications You must be signed in to change notification settings

yoitsro/joigoose

Repository files navigation

joigoose

npm version Build Status Dependencies Status

Joi and Mongoose sitting in a tree K-I-S-S-I-N-G...

Joi validation for your Mongoose models without the hassle of maintaining two schemas.

Installation

⚠️ WARNING! Use the v7.x.x branch of Joigoose if you're still using the scoped version of Joi.

npm install joigoose

Usage

1. Import Mongoose and Joi

We must pass Mongoose into Joigoose so it knows about ObjectIds and other Mongoose specific stuff:

const Mongoose = require("mongoose");
const Joigoose = require("joigoose")(Mongoose);

With Joi options which apply to all validators:

const Mongoose = require("mongoose");
const Joigoose = require("joigoose")(Mongoose, { convert: false });

With subdocument options to all subdocuments:

const Mongoose = require("mongoose");
const Joigoose = require("joigoose")(Mongoose, null, {
  _id: false,
  timestamps: false,
});

2. Write your Joi schema (see here about how to specify ObjectIds!)

Things to know!

Mongoose specific options can be specified in the meta object (see below).

Arrays with items of different types will end up with the Mongoose type Mixed. Ideally, Joi schemas shouldn't have to contain Mongoose specific types, such as Mongoose.Schema.Types.ObjectId because these Joi schemas may be required to work on frontend clients too.

var joiUserSchema = Joi.object({
  name: Joi.object({
    first: Joi.string().required(),
    last: Joi.string().required(),
  }),
  email: Joi.string().email().required(),
  bestFriend: Joi.string().meta({
    _mongoose: { type: "ObjectId", ref: "User" },
  }),
  metaInfo: Joi.any(),
  addresses: Joi.array()
    .items({
      line1: Joi.string().required(),
      line2: Joi.string(),
    })
    .meta({ _mongoose: { _id: false, timestamps: true } }),
});

3. Convert your Joi schema to a Mongoose-style schema

var mongooseUserSchema = new Mongoose.Schema(
  Joigoose.convert(joiUserSchema, options)
);

Conversion options

Options can be passed to the convert method as an object. Valid options are described below.

Key Type Default Description
typeKey String "type" The name of the key used for specifying the type in the generated schema.
Example:
{
  typeKey: '$type',
}

4. Create your model

User = Mongoose.model("User", mongooseUserSchema);

5. Enjoy!

var aGoodUser = new User({
  name: {
    first: "Barry",
    last: "White",
  },
  email: "[email protected]",
  metaInfo: {
    hobbies: ["cycling", "dancing", "listening to Shpongle"],
  },
});

aGoodUser.save(function (err, result) {
  // -> Success!
});

var aBadUser = new User({
  name: {
    first: "Barry",
    last: "White",
  },
  email: "Im not an email address!",
});

aBadUser.save(function (err, result) {
  // -> Error!
  // {
  //     "message": "User validation failed",
  //     "name": "ValidationError",
  //     "errors": {
  //         "email": {
  //             "properties": {
  //                 "type": "user defined",
  //                 "message": "Validator failed for path `{PATH}` with value `{VALUE}`",
  //                 "path": "email",
  //                 "value": "Im not an email address!"
  //             },
  //             "message": "Validator failed for path `email` with value `Im not an email address!`",
  //             "name": "ValidatorError",
  //             "kind": "user defined",
  //             "path": "email",
  //             "value": "Im not an email address!"
  //         }
  //     }
  // }
});

Known Limitations

I didn't spend long writing this, but I've aimed for 100% code coverage. It can be so much better, so help me out! Submit your issues and pull requests on GitHub.

  • No peer based validation (and, nand, or, xor, with, without, ref, assert, alternatives, when etc)
  • No Joi.binary object type
  • No Joi.func object type
  • No Joi.in object type

About

Joi validation for your Mongoose models without the hassle of maintaining two schemas

Resources

License

Stars

Watchers

Forks

Packages

No packages published