Skip to content

A double-entry accounting database with a typed GraphQL API

Notifications You must be signed in to change notification settings

df8org/scaledger

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 

Repository files navigation

Scaledger

pipeline status

A double-entry accounting database with a typed GraphQL API, supporting:

  • Immutable entries
  • An API introspected directly from a PostgreSQL schema
  • GraphQL subscriptions for real-time updates

Basics

Scaledger is designed to be used as a service for recording transactions (called postings) between accounts, and reporting their balances through an entity called ledger. This is particularly useful when you have to track the balances for thousands, or even millions, of individual user accounts such as in the case of marketplaces.

To use it, you deploy it as part of your service architecture and connect over a GraphQL interface. Documentation this interface is available at http://localhost:5000/graphiql

First, create some accounts:

mutation {
  createAccount(input: {account: {type: EQUITY, name: "Y Combinator Seed"}}) {
    account {
      id
    }
  }
}

mutation {
  createAccount(input: {account: {type: ASSET, name: "Cash"}}) {
    account {
      id
    }
  }
}

Next, create a posting between them:

mutation {
  createPosting(input: { posting: {
    creditId: "5c70baa8-f917-4220-afad-1521fdecb5a7",
    debitId: "9c42c59a-7404-47f2-9a63-fb3a8ecab111",
    currency: USD,
    amount: 15000000,
    externalId: "yc-safe-transfer"
  }})
}

You'll notice that the amount field is denominated in the minor value of the currency. This is important - don't use floats for accounting systems! Next, you'll notice currency - Scaledger is natively multi-currency and supports all of the ISO 4217 currency codes. If you're wondering what externalId is, that's required so that each posting from a downstream service is idempotent - as a defense against you sending the same request twice.

Both account and posting also support a metadata field which can be used to store abitrary key/value JSON dictionaries of extra data. Like any good ledger, posting cannot be mutated after it is created - to void a transaction you need to reverse it by creating an inverted one.

The general ledger can be queried after postings are created:

query {
  ledgers {
    nodes {
      accountName
      balance
      currency
    }
  }
}

Lastly, Scaledger also supports WebSockets for newly created postings via the GraphQL Subscription primitive:

subscription {
  postingCreated {
    posting {
      amount
      id
    }
  }
}

Stack

Scaledger uses a purpose-built PostgreSQL schema and provides a GraphQL API.

Services

  • scaledger-db: a PostgreSQL database and ledger schema
  • scaledger-server: a node-based PostGraphile GraphQL API

Development

scaledger includes a docker-compose configuration out of box.

  1. Clone the project
  2. Install Docker Compose
  3. cd docker
  4. Build images with ./scripts/images
  5. docker-compose up

By default, your first initialization of the container will automatically run the schema. Depending on how you've installed docker, you may need to prefix sudo on docker-compose up and on any of the scripts in docker/scripts.

If you'd like to create test data, run ./scripts/seed from inside of docker. You can run ./seed at any point to return the test data and schema to an initial state.

If you wish to recreate the database, or reset it, without any seed data you can run ./schema instead.

To connect to the docker container running the database with psql run ./scripts/psql.

Services In Development

About

A double-entry accounting database with a typed GraphQL API

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published