Formn

Formn, pronounced foreman, is an object-relational mapper for Node.js. What's the "F" stand for? It depends on your mood.

It's a fast, database-first ORM for projects written in TypeScript, and it uses the data-mapper pattern.

Check out the documentation. It has a getting started guide, tutorials, and more.

What does it do?

It takes queries that look like this:

SELECT  p.personID, p.firstName, p.lastName,
        pn.phoneNumberID, pn.phoneNumber
FROM    people p
INNER JOIN phone_numbers pn ON p.personID = pn.personID
ORDER BY p.firstName, p.lastName

and makes them look like this:

const query: Select<Person> = dataContext
  .from(Person, 'p')
  .innerJoin(PhoneNumber, 'pn', 'p.phoneNumbers')
  .select(
    'p.id', 'p.firstName', 'p.lastName',
    'pn.id', 'pn.phoneNumber')
  .orderBy('p.firstName', 'p.lastName')

const people: Person[] = await query
  .execute();

It maps tabular, relational data that look like this:

personID	firstName	lastName	phoneNumberID	phoneNumber
1	Joe	Shmo	1	530-307-8810
1	Joe	Shmo	2	916-200-1440
1	Joe	Shmo	3	916-293-4667
2	Rand	AlThore	4	666-451-4412

to normalized entities that look like this:

[
  Person {
    id: 1,
    firstName: 'Joe',
    lastName: 'Shmo',
    phoneNumbers: [
      PhoneNumber {
        id: 1,
        phoneNumber: '530-307-8810'
      },
      PhoneNumber {
        id: 2,
        phoneNumber: '916-200-1440'
      },
      PhoneNumber {
        id: 3,
        phoneNumber: '916-293-4667'
      }
    ]
  },
  Person {
    id: 2,
    firstName: 'Rand',
    lastName: 'AlThore',
    phoneNumbers: [
      PhoneNumber {
        id: 4,
        phoneNumber: '666-451-4412'
      }
    ]
  }
]

Why should I use it?

• It's fast.
• The code is well documented and thoroughly tested.
• Tutorials and documentation help you to get started quickly.
• It's database first rather than entity first, so it works well with existing projects and databases.
• There's a built-in model generator to generate entity classes from a database.
• The CLI handles database migrations.
• The query interface is intuitive and closely resembles SQL.
• It's type safe.
• Security concerns like SQL injection are covered.
• CRUD operations can be reused. One complex query with joins and filters can be used to select, update, and delete in a similar manner to SQL views.
• It lets you easily create queries that can be filtered and ordered dynamically.
• There are hooks for global conversions and transformations, like normalizing dates and formatting phone numbers.

Check out the docs!