Fluree DB

Usage documentation is located at https://docs.flur.ee.

Overview

Fluree is an immutable, temporal, ledger-backed semantic graph database that has a cloud-native architecture.

This repository is a stateless database as a library and designed to be utilized in conjunction with the Fluree Ledger which maintains all state. This database can be run in containers and dynamically scale to any desired load, can be embedded inside of your applications (Clojure, NodeJS for now) or can run as a stand-alone JVM service.

This database can also be built as a web-worker, and be embedded inside the browser. Thus far, a React Wrapper (Beta) has been developed that allows you to create real-time apps by wrapping your React components with queries (GraphQL or FlureeQL).

It is also possible to run Fluree in a "serverless" manner, where by utilizing Fluree SmartFunctions to embed data security along side your data (Data Defending Itself), you can have a permissioned application with just a single-page application (i.e. React) and Fluree Ledgers, but no application server.

Fluree includes time travel, allowing you to instantly query as of any historical moment in time, and even allows the abilty to stage proposed transactions to time travel into the future, to a hypothesized version of your data.

The best way to get started with Fluree is to go to the Getting Started page at https://flur.ee/getstarted/.

Development

Contributing

All contributors must complete a Contributor License Agreement.

Prerequisites

1. Install clojure tools-deps (version 1.10.3.933 or later).
   1. macOS: brew install clojure/tools/clojure
   2. Arch Linux: pacman -S clojure
2. Install Node & NPM
   1. macOS: brew install node
   2. Arch Linux: pacman -S nodejs
3. Install babashka
   1. macOS: brew install borkdude/brew/babashka

Building

NOTE: use make -j to run tasks in parallel.

• make deps - install all local dependencies

• make - make everything below

• make jar - make Java JAR file

• make nodejs - make JavaScript Fluree DB for Node

• make browser - make JavaScript Fluree DB for browsers

• make webworker - make JavaScript Fluree DB for web worker

• make install - install jar file into local .m2/maven

• make clean - clean all build directories/files

Tests

• make test - run all automated tests below
• make cljtest - run all CLJ tests
• make cljstest - run CLJS tests in headless Chrome & NodeJS
• make cljs-browser-test - run CLJS tests in headless Chrome
  • Needs karma-cli installed globally: npm install -g karma-cli
  • Needs Google Chrome browser installed
• make cljs-node-test - run CLJS tests in NodeJS
• make nodejs-test - run node package in node
• make browser-test - run browser package in headless Chrome

Running specific tests

This applies to CLJ tests only, not CLJS.

clojure -X:cljtest :kaocha.filter/focus [focus-spec]

...where focus-spec can be a test namespace or a fully-qualified deftest var. Note that the square brackets around the focus-spec must be present in the command, they are NOT there to indicate "optional" or "placeholder" in the example.

This feature comes from the test runner kaocha which has additional features.

CLJS REPL

In order to get a Node or web browser CLJS REPL running, you need to do the following:

1. Run npx shadow-cljs watch flureenjs (or flureedb for a browser REPL)
2. (Node only) Run node out/nodejs/flureenjs.js in a separate shell
3. (browser only) Connect to http://localhost:9630/ in your browser
   1. This seems currently broken though. It first gets stuck at "shadow-cljs Loading..." with an "unknown route" error in the JS console. If you reload that seems to resolve. But even once the dashboard loads and you connect a REPL, it says "No available JS runtime."
4. Connect an nREPL to the port specified in .shadow-cljs/nrepl.port
5. Inside that REPL run (shadow/repl :flureenjs) (or :flureedb for a browser REPL)
6. Try running something CLJS-specific to ensure you've got a working CLJS REPL
   1. For example: (js/parseInt "42")