A wrapper library written in Typescript with ZERO dependencies that adds ES6 promises and SQL-based migrations API to sqlite3 (docs).
note v4 of sqlite
has breaking changes compared to v3! Please see CHANGELOG.md
for more details.
- Installation
- Usage
- Typescript tricks
- API Documentation
- Management Tools
- Alternative SQLite libraries
- References
- License
Most people who use this library will use sqlite3 as the database driver.
Any library that conforms to the sqlite3
(API)
should also work.
$ npm install sqlite3 --save
# v4 of sqlite is targeted for nodejs 10 and on.
$ npm install sqlite --save
# If you need a legacy version for an older version of nodejs
# install v3 instead, and look at the v3 branch readme for usage details
$ npm install sqlite@3 --save
This module has the same API as the original sqlite3
library (docs),
except that all its API methods return ES6 Promises and do not accept callback arguments (with the exception of each()
).
import sqlite3 from 'sqlite3'
import { open } from 'sqlite'
// this is a top-level await
(async () => {
// open the database
const db = await open({
filename: '/tmp/database.db',
driver: sqlite3.Database
})
})()
or
import sqlite3 from 'sqlite3'
import { open } from 'sqlite'
open({
filename: '/tmp/database.db',
driver: sqlite3.Database
}).then((db) => {
// do your thing
})
or
import sqlite3 from 'sqlite3'
import { open } from 'sqlite'
// you would have to import / invoke this in another file
export async function openDb () {
return open({
filename: '/tmp/database.db',
driver: sqlite3.Database
})
}
If you want to enable the database object cache
import sqlite3 from 'sqlite3'
import { open } from 'sqlite'
(async () => {
const db = await open({
filename: '/tmp/database.db',
driver: sqlite3.cached.Database
})
})()
import sqlite3 from 'sqlite3'
sqlite3.verbose()
For more info, see this doc.
db.on('trace', (data) => {
})
You can use an alternative library to sqlite3
as long as it conforms to the sqlite3
API.
For example, using sqlite3-offline-next
:
import sqlite3Offline from 'sqlite3-offline-next'
import { open } from 'sqlite'
(async () => {
const db = await open({
filename: '/tmp/database.db',
driver: sqlite3Offline.Database
})
})()
import sqlite3 from 'sqlite3'
import { open } from 'sqlite'
(async () => {
const [db1, db2] = await Promise.all([
open({
filename: '/tmp/database.db',
driver: sqlite3.Database
}),
open({
filename: '/tmp/database2.db',
driver: sqlite3.Database
}),
])
await db1.migrate({
migrationsPath: '...'
})
await db2.migrate({
migrationsPath: '...'
})
})()
// db is an instance of `sqlite#Database`
// which is a wrapper around `sqlite3#Database`
const db = await open({
/**
* Valid values are filenames, ":memory:" for an anonymous in-memory
* database and an empty string for an anonymous disk-based database.
* Anonymous databases are not persisted and when closing the database
* handle, their contents are lost.
*/
filename: string
/**
* One or more of sqlite3.OPEN_READONLY, sqlite3.OPEN_READWRITE and
* sqlite3.OPEN_CREATE. The default value is OPEN_READWRITE | OPEN_CREATE.
*/
mode?: number
/**
* The database driver. Most will install `sqlite3` and use the `Database` class from it.
* As long as the library you are using conforms to the `sqlite3` API, you can use it as
* the driver.
*
* @example
*
* ```
* import sqlite from 'sqlite3'
*
* const driver = sqlite.Database
* ```
*/
driver: any
})
- See the
src/**/__tests__
directory for more example usages - See the
docs/
directory for full documentation. - Also visit the
sqlite3
library API docs
await db.exec('CREATE TABLE tbl (col TEXT)')
await db.exec('INSERT INTO tbl VALUES ("test")')
const result = await db.get('SELECT col FROM tbl WHERE col = ?', 'test')
// { col: 'test' }
const result = await db.get('SELECT col FROM tbl WHERE col = ?', ['test'])
// { col: 'test' }
const result = await db.get('SELECT col FROM tbl WHERE col = :test', {
':test': 'test'
})
// { col: 'test' }
const result = await db.all('SELECT col FROM tbl')
// [{ col: 'test' }]
const result = await db.run(
'INSERT INTO tbl (col) VALUES (?)',
'foo'
)
/*
{
// row ID of the inserted row
lastID: 1,
// instance of `sqlite#Statement`
// which is a wrapper around `sqlite3#Statement`
stmt: <Statement>
}
*/
const result = await db.run('INSERT INTO tbl(col) VALUES (:col)', {
':col': 'something'
})
const result = await db.run(
'UPDATE tbl SET col = ? WHERE col = ?',
'foo',
'test'
)
/*
{
// number of rows changed
changes: 1,
// instance of `sqlite#Statement`
// which is a wrapper around `sqlite3#Statement`
stmt: <Statement>
}
*/
// stmt is an instance of `sqlite#Statement`
// which is a wrapper around `sqlite3#Statement`
const stmt = await db.prepare('SELECT col FROM tbl WHERE 1 = ? AND 5 = ?5')
await stmt.bind({ 1: 1, 5: 5 })
let result = await stmt.get()
// { col: 'some text' }
const stmt = await db.prepare(
'SELECT col FROM tbl WHERE 13 = @thirteen ORDER BY col DESC'
)
const result = await stmt.all({ '@thirteen': 13 })
each()
is a bit different compared to the other operations due to its underlying implementation.
The function signature looks like this:
async each (sql, [...params], callback)
callback(err, row)
is triggered when the database has a row to return- The promise resolves when all rows have returned with the number of rows returned.
try {
// You need to wrap this in a try / catch for SQL parse / connection errors
const rowsCount = await db.each(
'SELECT col FROM tbl WHERE ROWID = ?',
[2],
(err, row) => {
if (err) {
// This would be if there is an error specific to the row result
throw err
}
// row = { col: 'other thing' }
}
)
} catch (e) {
throw e
}
// rowsCount = 1
Useful if you need to call methods that are not supported yet.
const rawDb = db.getDatabaseInstance()
const rawStatement = stmt.getStatementInstance()
await db.close()
This module is compatible with sql-template-strings.
import SQL from 'sql-template-strings'
const book = 'harry potter';
const author = 'J. K. Rowling';
const data = await db.all(SQL`SELECT author FROM books WHERE name = ${book} AND author = ${author}`);
This module comes with a lightweight migrations API that works with SQL-based migration files
With default configuration, you can create a migrations/
directory in your project with SQL files,
and call the migrate()
method to run the SQL in the directory against the database.
See this project's migrations/
folder for examples.
await db.migrate({
/**
* If true, will force the migration API to rollback and re-apply the latest migration over
* again each time when Node.js app launches.
*/
force?: boolean
/**
* Migrations table name. Default is 'migrations'
*/
table?: string
/**
* Path to the migrations folder. Default is `path.join(process.cwd(), 'migrations')`
*/
migrationsPath?: string
})
import { ISqlite, IMigrate } from 'sqlite'
See the definitions for more details.
// Assuming you have @types/sqlite3 installed
import sqlite3 from 'sqlite3'
// sqlite3.Database, sqlite3.Statement is the default if no explicit generic is specified
await open<sqlite3.Database, sqlite3.Statement>({
filename: ':memory'
})
Most methods allow for the use of generics to specify the data type of your returned data. This allows your IDE to perform better autocomplete and the typescript compiler to perform better static type analysis.
interface Row {
col: string
}
// result will be of type Row, allowing Typescript supported IDEs to autocomplete on the properties!
const result = await db.get<Row>('SELECT col FROM tbl WHERE col = ?', 'test')
interface Row {
col: string
}
// Result is an array of rows, you can now have array-autocompletion data
const result = await db.all<Row[]>('SELECT col FROM tbl')
result.each((row) => {
// row should have type information now!
})
See the docs
directory for full documentation.
- Beekeeper Studio: Open Source SQL Editor and Database Manager
- DB Browser for SQLite: Desktop-based browser.
- datasette: Datasette is a tool for exploring and publishing data. Starts up a server that provides a web interface to your SQLite data.
- SQLite Studio: A free, open source, multi-platform SQLite database manager written in C++, with use of Qt framework.
- HeidiSQL: Full-featured database editor.
- DBeaver: Full-featured multi-platform database tool and designer.
This library and the library it primarily supports, sqlite3
, may not be the best library that
fits your use-case. You might want to try these other SQLite libraries:
- better-sqlite3: Totes itself as the fastest and simplest library for SQLite3 in Node.js.
- Bun sqlite3:
bun:sqlite
is a high-performance builtin SQLite3 module forbun.js
. - sql.js: SQLite compiled to Webassembly.
- sqlite3-offline-next: Offers pre-compiled
sqlite3
binaries if your machine cannot compile it. Should be mostly compatible with this library.
If you know of any others, feel free to open a PR to add them to the list.
- Using SQLite with Node.js for Rapid Prototyping on Medium.com
- SQLite Documentation, e.g. SQL Syntax, Data Types etc. on SQLite.org
- ES6 tagged sql-template-strings.
The MIT License © 2020-present Kriasoft / Theo Gravity. All rights reserved.
Made with ♥ by Konstantin Tarkus (@koistya), Theo Gravity and contributors