Bun SQLite

Use this guide when your app runs on Bun and uses bun:sqlite.

The sqlfu project loop stays the same as Getting Started: write schema SQL, draft migrations, write query SQL, generate wrappers. The runtime adapter changes to createBunClient().

Config

The CLI can still use a local SQLite file path:

import {defineConfig} from 'sqlfu';

export default defineConfig({
  db: './db/app.sqlite',
  definitions: './definitions.sql',
  migrations: './migrations',
  queries: './sql',
  generate: {
    sync: true,
  },
});

bun:sqlite is synchronous, so generate.sync: true keeps generated wrappers synchronous too.

Schema and query

create table jobs (
  id integer primary key,
  name text not null,
  status text not null
);

Put the query in sql/queries.sql:

/** @name listJobsByStatus */
select id, name, status
from jobs
where status = :status
order by id;

Generate the migration and wrappers:

npx sqlfu draft
npx sqlfu migrate
npx sqlfu generate

Bun runtime

import {Database} from 'bun:sqlite';
import {createBunClient} from 'sqlfu';

import {listJobsByStatus} from './sql/.generated/queries.sql.ts';

const db = new Database('./db/app.sqlite');
const client = createBunClient(db);

const pendingJobs = listJobsByStatus(client, {status: 'pending'});

The wrapper is the same wrapper you would call with node:sqlite or better-sqlite3 when generate.sync: true is enabled.

Read next

• Getting Started for the base SQL workflow.
• Adapters for the bun:sqlite adapter snippet.
• Type generation from SQL for params, list params, and generated row types.