Skip to content

A very small library for running sql migrations with postgres.

License

Notifications You must be signed in to change notification settings

ovotech/pg-sql-migrate

Repository files navigation

Postgres migration tool with plain sql

A very small library for running sql migrations with postgres. It differs from the numerous other libs in this domain by being very minimal, using only raw timestamped sql files. No "down" migrations are provided by design, as that is usually a bad idea in production anyway.

Using with CLI

yarn add @ovotech/pg-sql-migrate

add a configuration file, which by default is ./pg-sql-migrate.config.json to configure the connection:

{
  "client": "postgresql://postgres:dev-pass@localhost:5432/postgres",
  "directory": "migrations",
  "table": "migrations"
}

The default values for "directory" and "table" configuration is migrations but you can override that if you need to.

Instead of a string you can use an object. This is passed directly to pg https://node-postgres.com/features/connecting

{
  "client": {
    "user": "postgres",
    "password": "dev-pass",
    "host": "localhost",
    "database": "postgres",
    "port": 5432
  }
}

To create new migrations in the designated directory you can run:

yarn migrate create my_migration

This will create a file migrations/<timestamp>_my_migration.pgsql that you can place raw sql into. After that, you can run the migration(s) by calling

yarn migrate execute

You can also specify the configuration as cli options

yarn migrate execute --config-directory ~/my/dir/migrations --config-table my-table --config-client postgresql://localhost:5432/my-database

In you code you can run it as a library

import { migrate } from '@ovotech/pg-sql-migrate';

await migrate();

Environment variables

In your config file you can use environment variables.

For example, if you have the env var PG_USER_PASS setup, you can access it with:

{
  "client": "postgresql://postgres:${PG_USER_PASS}@localhost:5432/postgres",
  "directory": "migrations"
  "table" "migrations"
}

Using the library

You can choose a different location for the config file, or to just input its contents directly:

import { migrate } from '@ovotech/pg-sql-migrate';
import { createLogger } from 'winston';

await migrate();

await migrate({ config: 'custom-config.json' });

await migrate({
  config: {
    client: 'postgresql://postgres:dev-pass@localhost:5432/postgres',
    // Custom table location
    table: 'my_table',
    // Custom directory for migration files
    directory: 'migrations_dir',
  },
});

// Custom logger
const logger = createLogger();
await migrate({ logger });

// Dry run
await migrate({ dryRun: true });

Distabling transactions for some migrations

By default migrations are wrapped in transactions, but there are some PG operations that cannot be performed inside a transaction. You can disable this for a specific transaction by adding a prefix

-- pg-sql-migrate: DISABLE TRANSACTION
ALTER TYPE my_type ADD VALUE 'VAL2' AFTER 'VAL1';

Running the tests

You can run the tests with:

yarn test

Coding style (linting, etc) tests

Style is maintained with prettier and eslint

yarn lint

Deployment

Deployment is preferment by circleci automatically on merge / push to master, but you'll need to bump the package version numbers yourself.

Contributing

Have a bug? File an issue with a simple example that reproduces this so we can take a look & confirm.

Want to make a change? Submit a PR, explain why it's useful, and make sure you've updated the docs (this file) and the tests (see test folder).

License

This project is licensed under Apache 2 - see the LICENSE file for details