potluckdb

Read and write a local SQLite database. Synchronize it through your own storage service.

Choose your language

PythonJavaScript and TypeScriptGoRustSwiftDart and FlutterJava and KotlinC# and .NETC and C++Objective-C

Each install page ends with a link to the quickstart for that language.

Each device has one writable database. You can synchronize in both directions automatically or with an explicit API call. sync() reports a conflict as a returned value, not as an exception. Every SDK provides the same behavior. Concepts explains databases, remotes, and conflicts.

import potluckdb

remote = potluckdb.Remote.from_directory("shared-remote")
db = potluckdb.open("app.potluckdb", remote)
db.execute("CREATE TABLE IF NOT EXISTS note(id INTEGER PRIMARY KEY, body TEXT)")
db.execute("INSERT INTO note(body) VALUES(?)", ("hello",))
db.close()
remote.close()

These paths persist between runs, as they do in an application. On the second run, the program opens the database that the first run created. It creates the table only if the table does not exist. The runnable examples use temporary directories and remove them after each run. They start with an empty database, so they create their tables directly.

The reference has ten language versions. Compatibility maps fourteen languages to those versions. Each page links to its versions in the other languages.

Page	What it tells you
Install	Requirements and installation steps for your language.
Quickstart	A complete program that synchronizes two devices and resolves a conflict.
Concepts	The database model, in the order that a program uses it.
Open and construct	Every call for opening a database or creating a remote.
Examples	Longer programs that you can run.
Platforms	Requirements and examples for each operating system, including the browser.
Frameworks	Integration with each application framework.
Standard interfaces	DB-API, JDBC, and ODBC access through one connection string.
Features	Every capability and what it means for your application.
Compatibility	Supported platforms, languages, analytical engines, and remotes.
Performance	One measured run for each remote type.
Analytics engine	The DuckDB library that each platform loads for analytical reads.
Errors	Every error code, its category, and the action to take.
Before production	Settings to choose before you release your application.
Licensing and trial	How the trial and license keys work.

The documentation generator reads the code’s public API on every change. The tests run every example against the package that you install. These checks keep the reference and examples current.