Skip to main content

Model

Models are fancy constructors compiled from Schema definitions.

An instance of a model is called a document.

Models are responsible for creating and reading documents from the underlying Couchbase database.

Compiling Your First Model

When you call model() function on a schema, Ottoman compiles a model for you.

const schema = new Schema({ name: String, age: Number });
const User = model('User', schema);
caution

The model() function makes a copy of the schema. Make sure that you've added everything you want to the schema, including hooks, before calling model()!

Model Options

You can pass a third argument to model() functions in order to setup your needs. In the next example we will set the collectionName to be users.

const schema = new Schema({ name: String, age: Number });
const User = model('User', schema, { collectionName: 'users' });
Defining Collection Name

Models will be mapped to your Collections, if no Collection name option is provided then the Collection name will be equal to the Model name. There is an exception to this rule:

  • If you provide a collectionName option at Ottoman instance level then the Collection name will be equal to Ottoman collectionName option if it's not explicitly passed as collectionName in model options.

    import { Ottoman } from 'ottoman';

    const ottoman = new Ottoman({ collectionName: '_default' });
    const schema = new Schema({ name: String, age: Number });

    // Collection name for model `Cat` will be `_default`
    const Cat = ottoman.model('Cat', schema);

    // Collection name for model `Dog` will be `dogs`
    const Dog = ottoman.model('Dog', schema, { collectionName: 'dogs' });

    Therefore this is the way to get the Collection name for a Model: Collection Name = Model collectionName Options > Ottoman collectionName Options > Model name :::

The models options are:

interface ModelOptions {
collectionName?: string;
scopeName?: string;
idKey?: string;
modelKey?: string;
maxExpiry?: number;
keyGenerator?: (params: { metadata: ModelMetadata }) => string;
keyGeneratorDelimiter?: string;
}
  • collectionName: define the collection name to be use in the Couchbase Server. The default value will be the Model's name.
  • scopeName: define the scope where the collection will be placed. The default value is _default
  • idKey: it's the value of the key to save your id. The default value is set to 'id'.
  • modelKey: define the key to store the model name into the document. The default value is _type
  • maxExpiry: value used to create a collection for this instance. The default value is 0.
  • keyGenerator: function to generate the key to store documents.
  • keyGeneratorDelimiter: string value used to build the document key. The default value is ::

If you don't provided a keyGenerator or keyGeneratorDelimiter implementation it will be inherited by Ottoman instance options, check this in Ottoman options

Model Id

Ottoman will generate automatically your document's id and will guarantee that each id will be unique.

Each document's id will be included on the document under a property called id by default.

The id property name can be modified using the ModelOptions.idKey

const schema = new Schema({ name: String, age: Number });
const User = model('User', schema, { collectionName: 'users', idKey: '__id' });

The above example will override the default id with __id, now for the User's documents you can get the id value from doc.__id.

You can also get the id value by calling the doc._getId() methods, regardless of the id property name. :::

Constructing Documents

An instance of a model is called a document. Creating and saving them to the database is easy.

const User = model('User', schema);

const user = new User({ name: 'Jane', age: 29 });

user.save();
// saved!

User.create({ name: 'Jane', age: 29 });
// also saved!

Note that no users will be created/removed until the connection that your model uses is open. Every model has an associated connection. When you use model(), your model will use the default Ottoman connection.

Create Many

Also, you can use createMany static function to create multiples documents at once. See the API docs for more detail.

User.createMany([{ name: 'John' }, { name: 'Jane' }]);
tip

The response status will be SUCCESS as long as no error occurs, otherwise it will be FAILURE.

Querying

Finding documents is easy with Ottoman, powered by the built-in Query Builder. Documents can be retrieved using each models find, findById, findOne, defined indexes or where static methods.

User.find({ name: 'Jane' });
// will return a list of all users with the name "Jane"

User.find({ name: 'Jane' }, { limit: 10 });
// will return a list of all users with the name "Jane" and limited to 10 items

User.find({ name: {$eq: 'Jane', $ignoreCase: true }});
// In some cases you need to compare without taking into account case sensitive, for this you can use the $ ignoreCase property: true

User.findOne({ name: 'Jane' });
// will return a document with a User with the name "Jane" or null in case of not finding it
User.findById('userId');
// will return the user document with the current id

User.findById('userId', { select: 'name, cards', populate: 'cards' });
// will return the user document with the current id only with the fields name and cards populated

The find options are: link

export interface IFindOptions {
skip?: number;
limit?: number;
sort?: Record<string, SortType>;
populate?: string | string[];
populateMaxDeep?: number;
select?: ISelectType[] | string | string[];
consistency?: SearchConsistency;
noCollection?: boolean;
lean?: boolean;
ignoreCase?: boolean;
}

Advanced Use of Select Parameter

You can select nested objects using the structure defined in the N1QL Language documentation Link

User.find({name: 'john'}, {select: '{"latLon": {geo.lat, geo.lon}, geo.lat} as geo  }'})

Advanced Use of Filter Parameter

const filter = {
$or: [{ price: { $gt: 'amount_val', $isNotNull: true } }, { auto: { $gt: 10 } }, { amount: 10 }],
$and: [
{ price2: { $gt: 1.99, $isNotNull: true } },
{ $or: [{ price3: { $gt: 1.99, $isNotNull: true } }, { id: '20' }] },
],
};
User.find(filter);
// Returns a list of the elements that match the applied filters

Use of ignoreCase

// Defining `User` schema
const userSchema = {
name: String,
};
// Some test documents
const user1 = { name: 'COUCHBASE' };
const user2 = { name: 'couchbase' };

// Create and save `User` model
const UserModel = model('User', userSchema);

await UserModel.create(user1);
await UserModel.create(user2);
  • Without ignoreCase
const { rows: documents } = await UserModel.find({ name: { $eq: 'Couchbase' } }, { lean: true });
console.log(`Documents: `, documents);

Will get:

$> Documents:  []
  • Using $ignoreCase in filters params
const { rows: documents } = await UserModel.find(
{ name: { $eq: 'Couchbase', $ignoreCase: true } }, // Find filters
{ lean: true } // Find options
);
console.log(`Documents: `, documents);
  • Using ignoreCase in find options
const { rows: documents } = await UserModel.find(
{ name: { $like: 'Couch%' } }, // Find filters
{ lean: true, ignoreCase: true } // Find options
);
// Could also use:
const { rows: documents } = await UserModel.find(
{ name: 'Couchbase' }, // Find filters
{ lean: true, ignoreCase: true } // Find options
);
console.log(`Documents: `, documents);

For the two previous examples will get something like this:

$> Documents: [
{
_type: 'User',
id: 'da520506-11c3-4f66-b36c-6cb38d51fd16',
name: 'COUCHBASE'
},
{
_type: 'User',
id: '97a3b89e-9c2e-4a75-86d0-f83a5b6f3aa3',
name: 'couchbase'
}
]
Note

Using ignoreCase as part of find functions options will always prioritize the value of $ignoreCase defined in the clause

UserModel.find([
{ address: { $like: 'NY-%', $ignoreCase: false } }, // ignoreCase will not be applied
{ name: 'John' } // ignoreCase will be applied
], { ignoreCase: true });

See the chapter on queries for more details on how to use the Query API.

Use of lean

By default, Ottoman queries return an instance of the Ottoman Document class. Documents have a lot of internal state for change tracking. Enabling the lean option tells Ottoman to skip instantiating a full Ottoman Document and just give you the plain old JavaScript object (POJO).

The lean feature is only for the documents (Models instances) query functions like find, findById, findOne, etc.

const UserModel = model('User', schema);
const leanDoc = await UserModel.findById(id, { lean: true });
const normalDoc = await UserModel.findById(id);

// In case you were wondering, the JSON form of a Ottoman document is the same
// as the POJO. The lean option only affects how much memory your
// Node.js process uses, not how much data is sent over the network.
console.log(JSON.stringify(normalDoc).length === JSON.stringify(leanDoc).length); // true

Under the hood, after executing a query, Ottoman converts the query results from POJOs to Ottoman Documents. If you turn on the lean option, Ottoman skips this step.

// with lean:true
console.log(leanDoc instanceof UserModel); // false
console.log(leanDoc instanceof Model); // false
console.log(leanDoc instanceof Document); // false
console.log(leanDoc instanceof Object); // true
console.log(leanDoc.constructor.name === 'Object'); // true

// with lean:false
console.log(normalDoc instanceof UserModel); // true
console.log(normalDoc instanceof Model); // true
console.log(normalDoc instanceof Document); // true
console.log(normalDoc instanceof Object); // true
console.log(normalDoc.constructor.name === '_Model'); // true
NOTE

The downside of enabling lean is that lean docs don't have:

  • Change tracking
  • Casting and validations
  • Hooks
  • save(), remove() and others model's methods

This is the main lean feature difference when is applied over an Ottoman document

caution

ManyQueryResponse is an util class and doesn't have this lean feature.

Use lean and populate

If you use both populate and lean, the lean option propagates to the populated documents as well. In the below example you can see it:

// Define schemas
const IssueSchema = new Schema({
title: String,
description: String,
});
const CardSchema = new Schema({
cardNumber: String,
zipCode: String,
issues: [{ type: IssueSchema, ref: 'Issue' }],
});
const CatSchema = new Schema({
name: String,
age: Number,
});
const UserSchema = new Schema({
type: String,
isActive: Boolean,
name: String,
card: { type: CardSchema, ref: 'Card' },
cats: [{ type: CatSchema, ref: 'Cat' }],
});

// Create models
const Issue = model('Issue', IssueSchema);
const Card = model('Card', CardSchema);
const Cat = model('Cat', CatSchema);
const User = model('User', UserSchema);

// Start Ottoman instance
const ottoman = getDefaultInstance();
await ottoman.start();

// Initialize data
const issueDoc = await Issue.create({ title: 'stolen card' });
const cardDoc = await Card.create({
cardNumber: '4242 4242 4242 4242',
zipCode: '42424',
issues: [issueDoc.id],
});
const cat1Doc = await Cat.create({ name: 'Figaro', age: 6 });
const cat2Doc = await Cat.create({ name: 'Garfield', age: 27 });
const userDoc = new User({
type: 'userPopulate',
isActive: false,
name: 'John Torvald',
card: cardDoc.id,
cats: [cat1Doc.id, cat2Doc.id]
});
const saved = await userDoc.save();

// Define query options
const options = { select: 'card, cats, name', populate: '*', lean: true };

// Execute a lean=true query
const userWithLean = await User.findById(saved.id, options);

// Execute a lean=false query
options.lean = false;
const userWithoutLean = await User.findById(saved.id, options);

console.log(userWithLean);
console.log(userWithoutLean);

ottoman.close();

With lean will get a POJO output:

{
card: {
cardNumber: '4242 4242 4242 4242',
zipCode: '42424',
issues: [ '794f6771-6f8b-417d-a814-1b535176824f' ],
id: 'c725c1f5-68dc-4e3a-9b89-58ce52185f24',
_type: 'Card'
},
cats: [
{
name: 'Figaro',
age: 6,
id: '84fc7fca-6099-4299-a03a-286f7464457e',
_type: 'Cat'
},
{
name: 'Garfield',
age: 27,
id: '0acbb8f7-6771-4468-bd21-8de7739cadcb',
_type: 'Cat'
}
],
name: 'John Torvald'
}

Without lean output:

_Model {
card: _Model {
cardNumber: '4242 4242 4242 4242',
zipCode: '42424',
issues: [ '794f6771-6f8b-417d-a814-1b535176824f' ],
id: 'c725c1f5-68dc-4e3a-9b89-58ce52185f24',
_type: 'Card'
},
cats: [
_Model {
name: 'Figaro',
age: 6,
id: '84fc7fca-6099-4299-a03a-286f7464457e',
_type: 'Cat'
},
_Model {
name: 'Garfield',
age: 27,
id: '0acbb8f7-6771-4468-bd21-8de7739cadcb',
_type: 'Cat'
}
],
name: 'John Torvald'
}

If we try to call userWithLean.toJSON() will get TypeError: userWithLean.toJSON is not a function

NOTE

When document to populate does not exist, the corresponding id is kept:

// Update user from above example
await User.updateById(saved.id, {
card: 'dummy-user-card-ID', // Reference card ID that does't not exist
cats: [
'dummy-user-cat-ID', // Reference cat ID that does't exist
...saved.cats
]
});

// Retrieving data
userWithLean = await User.findById(saved.id, options);

console.log(userWithLean.card); // 'dummy-user-card-ID'
console.log(userWithLean.cats[0]) ; // 'dummy-user-cat-ID'

When to Use lean

You should use lean when:

  • You're executing a query and sending the results without modification to, say, an Express response.
  • If you do not modify the query results and do not use custom getters.

You should not use lean when:

  • Need modify the query results or rely on features like getters or transforms.

Below is an example of an Express route that is a good candidate for lean. This route does not modify the person document and doesn't rely on any Ottoman-specific functionality.

app.get('/person/:id', function(req, res) {
Person.findById(req.params.id, { lean: true })
.then(person => res.json(person))
.catch(error => res.json({ error: error.message }));
});

Below is an example of an Express route that should not use lean. As a general rule of thumb, GET routes are good candidates for lean in a RESTful API. On the other hand, PUT, POST, etc. routes generally should not use lean.

// This route should **not** use `lean()`, because lean means no `save()`.
app.put('/person/:id', function (req, res) {
Person.findOne(req.params.id).then(person => {
assert.ok(person);
Object.assign(person, req.body);
return person.save();
})
.then(person => res.json(person))
.catch(error => res.json({ error: error.message }));
});

Deleting

Models have static removeById() function to remove documents matching the given id value. See the API docs for more detail.

User.removeById('userId');

Models have static removeMany() function to remove all documents matching the given condition. See the API docs for more detail.

User.removeMany({ name: { $like: '%JohnDoe%' } });
tip

The response status will be SUCCESS as long as no error occurs, otherwise it will be FAILURE.

Updating

Each model has its own updateById method for modifying documents in the database without returning them to your application. See the API docs for more detail.

User.updateById('userId', { age: 30 });
// update document with id equal to 'userId' with age 30

Models have static method replaceById which has the same behavior as updateById, except that the replaceById replaces the existing document with the given document. See the API docs for more detail.

User.replaceById('userId', { age: 30, name: 'John' });
caution

The replaceById method completely replaces the existing document as long as the new document complies with the schema rules.

Models have static updateMany function to update all documents matching the given condition. See the API docs for more detail.

User.updateMany({ name: { $like: '%JohnDoe%' } }, { name: 'John' });
tip

The response status will be SUCCESS as long as no error occurs, otherwise it will be FAILURE.

Models have static findOneAndUpdate function to finds a document that matches the conditions of the collection and updates it. See the API docs for more detail.

User.findOneAndUpdate({ name: { $like: '%John Doe%' } }, { name: 'John' }, { new: true, upsert: true });
tip

By default the option new and upsert are false

If options.new is true return the document after update otherwise by default return the document before update.

If options.upsert is true insert a document if the document does not exist.

Handling Multilpe Models

When you create a new Model Ottoman will register it by name.

const User = model('User', userSchema);

// Ottoman under the hood will register in a dictionary object with a key set to model name
const models = {
User: UserModel,
};
caution

Duplicate Model's name will throw an exception notifying about the register model duplication.

Getting Existing Models

You can retrieve a registered Model using the getModel function.

import { getModel, model } from 'ottoman';

const User = model('User', { name: string });

// anywhere else in the app
const User = getModel('User');

If the name provided doesn't match any registered model undefined value will be returned.

tip

Maybe you want to get an existing model and if it's don't exist then attempt to create, the next example could be helpful.

import { getModel, model } from 'ottoman';

const User = getModel('User') || model('User', userSchema);

Drop Collection

Ottoman's Models provide a dropCollection static method to remove a collection.

...
const User = model('User', schema, {scopeName: 'scopeA'});

// dropCollection without parameter will drop it's own collection
// This case User Collection in the scopeA will be removed
User.dropCollection()

// dropCollection with collectionName parameter will drop the collection in the same scope
// This case Cat Collection in the scopeA will be removed
User.dropCollection('Cat')

// dropCollection can even drop a collection from another scope, if it's provide explicitly
// This case Cat Collection in the scopeB will be removed
User.dropCollection('Cat', 'scopeB')

To check the dropCollection API click here

Next Up

Now that we've covered Models, let's take a look at Documents.