Table of Contents

• Install
• Getting Started
• Documentation
• License

Install

npm install umzug-hyper-storage

Getting Started

umzug-hyper-storage is a zero-dependency Storage Adapter for Umzug that uses a Hyper Data Service to track migrations.

It works great as a tool to track and run migrations of your data stored in a Hyper Data Service. It's also a great generalized script runner, not just for running migrations.

To instantiate, provide an instance of hyper-connect that connects to your data service you would like to use to track your scripts.

import { Umzug } from 'umzug'
import { HyperStorage } from 'umzug-storage-adapter'
import { connect } from 'hyper-connect'

const hyper = connect(process.env.HYPER)

const umzug = new Umzug({
  storage: new HyperStorage({ hyper }),
  // optionally pass a hyper-connect instance to each of your scripts
  context: {
    hyper
  }
  ...
})

await umzug.up() // run migrations

This will create a document in your Hyper Data Service in the follow shape:

{
  _id: 'hyper-scripts-meta',
  type: '__scripts',
  migrations: [
    // array of scripts that have been ran
  ],
  createdAt: '2022-04-27T20:30:40.688Z',
  updatedAt: '2022-04-27T20:30:40.688Z'
}

As scripts are ran, their names are appended to the migrations array in this document. If scripts are rolled back, their names are removed from the migrations array, following normal Umzug rules.

This is how Umzug will track which scripts to run, rollback, or have been previously executed.

Documentation

See Umzug Docs for Umzug usage.

A hyper-connect instance connecting to your Hyper Data Service is required upon instantiation.

You may also pass a doc argument which dictates the shape of the document added to hyper data service:

interface HyperStorageDocArgs {
  // the _id of the meta document. Defaults to 'hyper-scripts-meta'
  id?: string;
  // the type of the meta document. Defaults to '__scripts'
  type?: string;
  /**
   * the field on the meta document to use to store the document type
   * ie. 'docType'. Defaults to 'type'
   */
  typeField?: string;
  /**
   * the field on the meta document to store the time
   * the meta document was created. Defaults to 'createdAt'
   */
  createdField?: string;
  /**
   * the field on the meta document to store the time
   * the meta document was last updated. Defaults to 'updatedAt'
   *
   * This field is updated each time migrations are ran or rolled back
   */
  updatedField?: string;
}

Example using doc

import { Umzug } from 'umzug'
import { HyperStorage } from 'umzug-storage-adapter'
import { connect } from 'hyper-connect'

const hyper = connect(process.env.HYPER)

const umzug = new Umzug({
  storage: new HyperStorage({
    hyper,
    doc: {
      type: 'umzug-scripts',
      createdField: 'created_at',
      updatedField: 'updated_at'
    }
  }),
  ...
})

Passing hyper-connect to your Umzug scripts

Chances are you are running scripts against a Hyper Data Service or Hyper Cloud Application Services. You can use Umzug context to pass an instance of hyper-connect to each script ran.

By using context to inject hyper-connect, this will make your scripts easier to unit test.

import { Umzug } from 'umzug'
import { HyperStorage } from 'umzug-storage-adapter'
import { connect } from 'hyper-connect'

const hyper = connect(process.env.HYPER)

const umzug = new Umzug({
  storage: new HyperStorage({ hyper }),
  context: { hyper }
  ...
})

// then in your script:
const migration = {
  name: '01-awesome-migration',
  async up ({ context: { hyper }}) {
    ... // use hyper-connect to perform migration
  },
  async down ({ context: { hyper }}) {
    ... // use hyper-connect to perform rollback
  }
}

License

Apache 2.0