This guide will teach you:
- How to compile your own copy of Metabase
- How to set up a development environment
- How to run the Metabase Server
- How to contribute back to the Metabase project
- How to add support in Metabase for other languages
In general, we like to have an open issue for every pull request as a place to discuss the nature of any bug or proposed improvement. Each pull request should address a single issue, and contain both the fix as well as a description of how the pull request and tests that validate that the PR fixes the issue in question.
For significant feature additions, it is expected that discussion will have taken place in the attached issue. Any feature that requires a major decision to be reached will need to have an explicit design document written. The goals of this document are to make explicit the assumptions, constraints and tradeoffs any given feature implementation will contain. The point is not to generate documentation but to allow discussion to reference a specific proposed design and to allow others to consider the implications of a given design.
We don't like getting sued, so before merging any pull request, we'll need each person contributing code to sign a Contributor License Agreement here
The development scripts are designed for Linux/Mac environment, so we recommend using the latest Windows 10 version with WSL (Windows Subsystem for Linux) and Ubuntu on Windows. The Ubuntu Bash shell works well for both backend and frontend development.
If you have problems with your development environment, make sure that you are not using any development commands outside the Bash shell. As an example, Node dependencies installed in normal Windows environment will not work inside Ubuntu Bash environment.
These are the set of tools which are required in order to complete any build of the Metabase code. Follow the links to download and install them on your own before continuing.
- Oracle JDK 8 (http://www.oracle.com/technetwork/java/javase/downloads/index.html)
- Node.js (http://nodejs.org/)
- Yarn package manager for Node.js
- Leiningen (http://leiningen.org/)
If you are developing on Windows, make sure to use Ubuntu on Windows and follow instructions for Ubuntu/Linux instead of installing ordinary Windows versions.
The entire Metabase application is compiled and assembled into a single .jar file which can run on any modern JVM. There is a script which will execute all steps in the process and output the final artifact for you.
./bin/build
After running the build script simply look in target/uberjar for the output .jar file and you are ready to go.
See this guide.
If you plan to work on the Metabase code and make changes then you'll need to understand a few more things.
The Metabase application has two basic compnents:
- a backend written in Clojure which contains a REST API as well as all the relevant code for talking to databases and processing queries.
- a frontend written as a Javascript single-page application which provides the web UI.
Both components are built and assembled together into a single jar file which runs the entire application.
Metabase depends on lots of other 3rd party libraries to run, so as you are developing you'll need to keep those up to date. Leiningen will automatically fetch Clojure dependencies when needed, but for JavaScript dependencies you'll need to kick off the installation process manually when needed.
# javascript dependencies
$ yarnRun your backend development server with
lein ring server
Start the frontend build process with
yarn build-hot
We use these technologies for our FE build process to allow us to use modules, es6 syntax, and css variables.
- webpack
- babel
- cssnext
Frontend tasks are executed using yarn. All available tasks can be found in package.json under scripts.
To build the frontend client without watching for changes, you can use:
$ yarn buildIf you're working on the frontend directly, you'll most likely want to reload changes on save, and in the case of React components, do so while maintaining state. To start a build with hot reloading, use:
$ yarn build-hotNote that at this time if you change CSS variables, those changes will only be picked up when a build is restarted.
There is also an option to reload changes on save without hot reloading if you prefer that.
$ yarn build-watchSome systems may have trouble detecting changes to frontend files. You can enable filesystem polling by uncommenting the watchOptions clause in webpack.config.js. If you do this it may be worth making git ignore changes to webpack config, using git update-index --assume-unchanged webpack.config.js
All frontend tests are located in frontend/test directory. Run all frontend tests with
yarn test
which will run unit, end-to-end, and legacy Karma browser tests in sequence.
End-to-end tests simulate realistic sequences of user interactions. They render a complete DOM tree using Enzyme and use temporary backend instances for executing API calls.
End-to-end tests use an enforced file naming convention <test-suite-name>.e2e.spec.js to separate them from unit tests.
Useful commands:
lein run refresh-integration-test-db-metadata # Scan the sample dataset and re-run sync/classification/field values caching
yarn test-e2e-watch # Watches for file changes and runs the tests that have changed
yarn test-e2e-watch TestFileName # Watches the files in paths that match the given (regex) stringThe way integration tests are written is a little unconventional so here is an example that hopefully helps in getting up to speed:
import {
useSharedAdminLogin,
createTestStore,
} from "__support__/e2e";
import {
click
} from "__support__/enzyme"
import { mount } from "enzyme"
import { FETCH_DATABASES } from "metabase/redux/metadata";
import { INITIALIZE_QB } from "metabase/query_builder/actions";
import RunButton from "metabase/query_builder/components/RunButton";
describe("Query builder", () => {
beforeAll(async () => {
// Usually you want to test stuff where user is already logged in
// so it is convenient to login before any test case.
useSharedAdminLogin()
})
it("should let you run a new query", async () => {
// Create a superpowered Redux store.
// Remember `await` here!
const store = await createTestStore()
// Go to a desired path in the app. This is safest to do before mounting the app.
store.pushPath('/question')
// Get React container for the whole app and mount it using Enzyme
const app = mount(store.getAppContainer())
// Usually you want to wait until the page has completely loaded, and our way to do that is to
// wait until the completion of specified Redux actions. `waitForActions` is also useful for verifying that
// specific operations are properly executed after user interactions.
// Remember `await` here!
await store.waitForActions([FETCH_DATABASES, INITIALIZE_QB])
// You can use `enzymeWrapper.debug()` to see what is the state of DOM tree at the moment
console.log(app.debug())
// You can use `testStore.debug()` method to see which Redux actions have been dispatched so far.
// Note that as opposed to Enzyme's debugging method, you don't need to wrap the call to `console.log()`.
store.debug();
// For simulating user interactions like clicks and input events you should use methods defined
// in `enzyme.js` as they abstract away some React/Redux complexities.
click(app.find(RunButton))
// Note: In pretty rare cases where rendering the whole app is problematic or slow, you can just render a single
// React container instead with `testStore.connectContainer(container)`. In that case you are not able
// to click links that lead to other router paths.
});
})
You can also skim through __support__/e2e.js and __support__/enzyme.js to see all available methods.
Unit tests are focused around isolated parts of business logic.
Unit tests use an enforced file naming convention <test-suite-name>.unit.spec.js to separate them from end-to-end and integration tests.
yarn test-unit # Run all tests at once
yarn test-unit-watch # Watch for file changes
If you need to test code which uses browser APIs that are only available in real browsers, you can add a Karma test to frontend/test/legacy-karma directory.
yarn test-karma # Run all tests once
yarn test-karma-watch # Watch for file changes
Leiningen and your REPL are the main development tools for the backend. There are some directions below on how to setup your REPL for easier development.
And of course your Jetty development server is available via
lein run
To automatically load backend namespaces when files are changed, you can instead run with
lein ring server
lein ring server takes significantly longer to launch than lein run, so if you aren't working on backend code we'd recommend sticking to launching with lein run.
Most of the drivers Metabase uses to connect to external data warehouse databases are separate Leiningen projects under the modules/ subdirectory. When running Metabase via lein, you'll
need to build these drivers in order to have access to them. You can build drivers as follows:
# Build the 'mongo' driver
./bin/build-driver.sh mongo
(or)
# Build all drivers
./bin/build-drivers.sh
The first time you build a driver, it will be a bit slow, because Metabase needs to build the core project a couple of times so the driver can use it as a dependency; you can take comfort in the
fact that you won't need to build the driver again after that. Alternatively, running Metabase 1.0+ from the uberjar will unpack all of the pre-built drivers into your plugins directory; you can
do this instead if you already have a Metabase uberjar (just make sure plugins is in the root directory of the Metabase source, i.e. the same directory as project.clj).
For development when running various Leiningen tasks you can add the include-all-drivers profile to merge the drivers' dependencies and source paths into the Metabase
project:
# Install dependencies
lein with-profiles +include-all-drivers deps
This profile is added by default when running lein repl, tests, and linters.
Run unit tests with
lein test
or a specific test with
lein test metabase.api.session-test
By default, the tests only run against the h2 driver. You can specify which drivers to run tests against with the env var DRIVERS:
DRIVERS=h2,postgres,mysql,mongo lein test
Some drivers require additional environment variables when testing since they are impossible to run locally (such as Redshift and Bigquery). The tests will fail on launch and let you know what parameters to supply if needed.
lein eastwood && lein bikeshed && lein docstring-checker && lein check-namespace-decls && ./bin/reflection-linter
.dir-locals.el contains some Emacs Lisp that tells clojure-mode how to indent Metabase macros and which arguments are docstrings. Whenever this file is updated,
Emacs will ask you if the code is safe to load. You can answer ! to save it as safe.
By default, Emacs will insert this code as a customization at the bottom of your init.el.
You'll probably want to tell Emacs to store customizations in a different file. Add the following to your init.el:
(setq custom-file (concat user-emacs-directory ".custom.el")) ; tell Customize to save customizations to ~/.emacs.d/.custom.el
(ignore-errors ; load customizations from ~/.emacs.d/.custom.el
(load-file custom-file))Start up an instant cheatsheet for the project + dependencies by running
lein instant-cheatsheet
We are an application with lots of users all over the world. To help them use Metabase in their own language, we mark all of our strings as i18n.
If you need to add new strings (try to be judicious about adding copy) do the following:
- Tag strings in the frontend using
tandjtES6 template literals (see more details in https://ttag.js.org/):
const someString = t`Hello ${name}!`;
const someJSX = <div>{jt`Hello ${name}`}</div>;and in the backend using trs and related macros (see more details in https://github.com/puppetlabs/clj-i18n):
(trs "Hello {0}!" name)If you see incorrect or missing strings for your langauge, please visit our POEditor project and submit your fixes there.
Copyright © 2017 Metabase, Inc
Distributed under the terms of the GNU Affero General Public License (AGPL) except as otherwise noted. See individual files for details.