Fast SQLite library with optional full db encryption, simple query builder, and a host of utility features, all in one neat package.
super-sqlite3
is a thin wrapper around better-sqlite3-multiple-ciphers
, which extends better-sqlite3
(the fastest SQLite library for node.js) with full database encryption using SQLite3MultipleCiphers
. super-sqlite3
then adds its own query builder and other convenience features.
- fast synchronous API (yes, it is)
- optional full database encryption
- full transaction support
- simple query builder inspired by
knex.js
- db utility and convenience functions
- nice API
npm install super-sqlite3
const Database = require('super-sqlite3')
const db = new Database('/path/to/file', options?)
Options:
options.readonly
open the database connection in readonly mode (default:false
).options.fileMustExist
if the database does not exist, an Error will be thrown instead of creating a new file. This option is ignored for in-memory, temporary, or readonly database connections (default:false
).options.timeout
the number of milliseconds to wait when executing queries on a locked database, before throwing aSQLITE_BUSY
error (default:5000
).options.verbose
provide a function that gets called with every SQL string executed by the database connection (default:null
).options.nativeBinding
if you're using a complicated build system that moves, transforms, or concatenates your JS files,better-sqlite3-multiple-ciphers
might have trouble locating its native C++ addon (better_sqlite3.node
). If you get an error that looks like this, you can solve it by using this option to provide the file path ofbetter_sqlite3.node
(relative to the current working directory).
Note: not all functionality of
better-sqlite3
is directly exposed yet, but you can access the wrapped object instance atdb.$
.
db.query
(aliased db.prepare
) returns Statement
object. Check better-sqlite3
docs for more details.
db.query(sql).run(bindParams?) // exec query
db.query(sql).get(bindParams?) // get single record
db.query(sql).all(bindParams?) // get array of records
db.query(sql).raw(bindParams?) // return data as value arrays, instead of objects
You can also use shorthands, but multiple args for binding parameters are not supported, use array or object:
db.run(sql, bindParams?) // exec query
db.get(sql, bindParams?) // get single record
db.all(sql, bindParams?) // get array of records
db.raw(sql, bindParams?) // get array of value arrays
Handlers accept optional binding parameters, that are used to prevent SQL injection.
Anonymous parameters use ?
sign for a placeholder:
db.query('INSERT INTO table VALUES (?, ?, ?)').run(1, 2, 3)
// all these are equivalent
.run(1, 2, 3) // multiple args
.run([1, 2, 3]) // array
.run([1], [2, 3]) // mixed
Named parameters use @
, $
or :
followed by a name, and require an object:
db.query('INSERT INTO table VALUES (:foo, @bar, $baz)').run({foo:1, bar:2, baz:3})
// exec multiple statements
db.exec(queries)
// exec multiple statements from file
db.execFile(pathToFile)
Note: bulk insert mode is experimental, be cautious
In bulk insert mode, super-sqlite3
temporarily sets custom PRAGMA
values, which trade safety for speed. You can view, change, remove, or add your own values, by directly manipulating the object:
console.log(db._bulkInsertPragma)
.bulkInsert
takes one parameter, a function that performs the insertion. A transaction is automatically started before execution and committed immediately afterwards.
db.bulkInsert( db => {...insert...} )
.pragma
allows you to query and set PRAGMA values.
db.pragma('auto_vacuum')
// [ { auto_vacuum: 0 } ]
To get bare value, use .pragmaValue
:
db.pragmaValue('busy_timeout')
// 5000
To set the value:
db.pragma('busy_timeout=3000')
Full PRAGMA list can be found here.
Returns the number of rows in the specified table, or all tables if omitted. Returns object.
db.count('table') // single
// { table: 42 }
db.count(['table1', 'table2']) // multiple
// { table1: 42, table2: 420 }
db.count() // all
super-sqlite3
provides convenient functions for querying schema.
// get names of all tables
db.tables() // array
// check if table exists
db.hasTable(tableName) // bool
// get column names of a table
db.columns(tableName) // array
// get extended column information
db.columnsExt(tableName) // object
// get table schema sql
// tables - string or array of table names, returns all if omitted
db.schema(tables) // string
// get table indexes
db.indexes(tableName) // object
db.altergen
generates a set of queries that help to add/move/update columns while maintaining the intended column order.
db.altergen(tableName) // string
super-sqlite3
uses ChaCha20-Poly1305
cipher algorithms, the default for SQLite3MultipleCiphers
. See docs for more details.
// encrypt currently opened database
db.encrypt(key)
// PERMANENTLY decrypt the database (once unlocked)
db.decrypt()
// unlock database for a session
db.unlock(key)
// check if db is unlocked
db.isUnlocked()
Create backup of the database. Returns promise resolving to the path of the backup file created.
await db.backup(options?)
Options:
options.file
absolute path to backup fileoptions.dir
path to backup dir (defaults to the same dir as the current database). The filename will be derived from the current name, with an added datestamp.options.time
add time to datestamp, defaults totrue
.options.progress
seebetter-sqlite3
documentation.
If options are omitted, the backup file will be stored alongside the current database, using the same name with an added timestamp, so: database.sqlite
=> database--2027-09-14-09-50-36.sqlite
.
Gracefully shut down the database connection.
db.close()
To use query builder, invoke db
object with table name:
db(table).select(fields).where(conditions).get()
Note: query builder is only intended to help with basic queries, it is not meant to replace SQL.
Select is the default mode, and can be omitted. Accepts arguments in several formats:
select('field1', 'field2') // multiple args
select(['field1', 'field2']) // array
select({field1:'alias', field2:true}) // object
// to alias field:
select('*', {field:'alias'})
Select type query must be executed by one of the following:
.get() // single record
.all() // array of records
.count() // record count
.exists() // boolean
// where
where('field', 3)
where('field', '>', 3)
where([ ['field', 3], ['field2', '>', 15] ]) // note single top arg
where({field:3, field2:'value'})
// where not
whereNot('field', 3)
whereNot({field: 3})
// where in
whereIn('field', [5, 8])
// where not in
whereNotIn('field', [5, 8])
// where null
whereNull('field', 'field2')
whereNull(['field', 'field2'])
whereNull({field:1, field2:true, field3:false}) // field3 not included
// where not null
whereNotNull('field', 'field2')
If the table uses id
as its primary key, a shorthand can be used:
// .id(n) is a shorthand to .where({id:n})
db(table).id(10).get()
// SELECT * FROM table WHERE id=10;
// when id is the only selector in select type query:
db(table).getId(10)
db.order('field') // asc
db.order('field', 'asc')
db.order('field', 'desc')
db.limit(100)
db.offset(1000)
db(table).join(table2, field)
db(table).innerJoin(table2, field) // alias
// SELECT * FROM table INNER JOIN table2 USING (field);
db(table).join(table2, [field, field2])
// SELECT * FROM table INNER JOIN table2 USING (field, field2);
db(table).join(table2, {'table.id':'table2.tab_id'})
// SELECT * FROM table INNER JOIN table2 ON table.id=table2.tab_id;
db(table).leftJoin(table2, field)
db(table).leftOuterJoin(table2, field) // alias
// SELECT * FROM table LEFT OUTER JOIN table2 USING (field);
db(table).leftJoin(table2, [field, field2])
// SELECT * FROM table LEFT OUTER JOIN table2 USING (field, field2);
db(table).leftJoin(table2, {'table.id':'table2.tab_id'})
// SELECT * FROM table LEFT OUTER JOIN table2 ON table.id=table2.tab_id;
.insert
returns rowId on success.
db(table).insert({foo:'bar'})
.upsert
tries to update an existing record, and falls back to insert if it is not found.
db(table).upsert(data, conflict?)
// upsert
db(table).upsert({foo:'bar'})
// INSERT INTO table (foo) VALUES (?)
// ON CONFLICT DO UPDATE SET foo=excluded.foo;
db(table).upsert({foo:'bar', baz:3}, 'foo')
// INSERT INTO table (foo, baz) VALUES (?,?)
// ON CONFLICT (foo)
// DO UPDATE SET baz=excluded.baz;
db(table).upsert({one:'1st', two:2, three:'tre'}, ['one', 'two'])
// INSERT INTO table (one, two, three) VALUES (?,?,?)
// ON CONFLICT (one, two)
// DO UPDATE SET three=excluded.three;
.upsert
returns result object:
{changes:1, lastInsertRowid:0} // update
{changes:1, lastInsertRowid:42} // insert
.update
returns the number of affected rows.
db(table).where(cond).update(obj) // int
.delete
returns the number of affected rows.
db(table).where(cond).delete() // int
MIT