Skip to content

Latest commit

 

History

History
366 lines (215 loc) · 9.94 KB

CONTRIBUTING.rst

File metadata and controls

366 lines (215 loc) · 9.94 KB
Raw

Contributing

Bugs

All bugs are tracked in https://bugzilla.mozilla.org/.

Write up a new bug:

https://bugzilla.mozilla.org/enter_bug.cgi?product=Webtools&component=Pontoon

Docker

Everything runs in a Docker container. Thus Pontoon requires fewer things to get started and you're guaranteed to have the same setup as everyone else and it solves some other problems, too.

If you're not familiar with Docker and docker-compose, it's worth reading up on.

Database

By default, you will have default data loaded for only the Pontoon Intro project. If you have a database dump, you can load it into your PostgreSQL database by running:

$ make loaddb DB_DUMP_FILE=path/to/my/dump

Note that your database container needs to be running while you do that. You can start just the postgresql container by running:

$ docker-compose run postgresql -d

Interactive shell

If you need to run specific commands, that are not covered by our Makefile, you can start an interactive shell inside a Pontoon container:

$ make shell

Browser Support

The following is the browser support matrix of Pontoon:

Firefox >= 52,
FirefoxAndroid >= 52,
Chrome >= 57,
ChromeAndroid >= 57,
Edge >= 16,
Safari >= 10.1,
iOS >= 10.3

Python code conventions

Our Python code is automatically formatted using black. We enforce that in our Continuous Integration tool (travis), so you will need to run black on your code before sending it for review.

You can run black locally either as an add-on in your code editor, or as a git pre-hook commit. Alternatively, you can format your code using:

$ make black

Note

Using black on all Python code means that we cannot fight over code style anymore. You are free to write code however you like, because in the end black is the one that will format it. We thus don't need to pay any more attention to style during code reviews, and are free from those never-ending code style discussions.

Additionally, we use a linter to verify that imports are correct. You can run it with:

$ make flake8

In the rare case when you cannot fix a flake8 error, use # noqa to make the linter ignore that error. Note that in most cases, it is better to fix the issues than ignoring them.

Javascript code conventions

Outside the frontend folder, we don't follow strict rules other than using 2-space indentation.

Inside frontend (which contains the Translate app), we use 4-space indentation, as well as a lot of other rules that are defined in our .eslintrc.js file.

To run the linter, do:

$ make lint-frontend

For more specifics about the `frontend folder, look at the README.md file there.

Git conventions

The first line is a summary of the commit. It should start with one of the following:

Fix bug XXXXXXX

or:

Bug XXXXXXX

The first, when it lands, will cause the bug to be closed. The second one does not.

After that, the commit should explain why the changes are being made and any notes that future readers should know for context or be aware of.

We follow The seven rules of a great Git commit message:

  1. Separate subject from body with a blank line
  2. Limit the subject line to 50 characters
  3. Capitalize the subject line
  4. Do not end the subject line with a period
  5. Use the imperative mood in the subject line
  6. Wrap the body at 72 characters
  7. Use the body to explain what and why vs. how

Pull requests

Pull request summary should indicate the bug the pull request addresses.

Pull request descriptions should cover at least some of the following:

  1. what is the issue the pull request is addressing?
  2. why does this pull request fix the issue?
  3. how should a reviewer review the pull request?
  4. what did you do to test the changes?
  5. any steps-to-reproduce for the reviewer to use to test the changes

Code reviews

Pull requests should be reviewed before merging.

Style nits should be covered by linting as much as possible.

Code reviews should review the changes in the context of the rest of the system.

Dependencies

Dependencies for production Pontoon are in requirements/default.txt. Development dependencies are in requirements/dev.txt. They need to be pinned and hashed, and we use hashin for that.

Note that we use a specific format for our dependencies, in order to make them more maintainable. When adding a new requirement, you should add it to the appropriate section and put its sub-dependencies in requirements/contraints.txt if applicable. For example, to add foobar version 5:

$ hashin -r requirements/default.txt foobar==5

Then open requirements/default.txt and move the added dependencies to:

  • the first section if it has no other requirements
  • the requirements/constraints.txt if they are sub-dependencies, and add all their dependencies there as well.

That format is documented more extensively inside the requirements/default.txt file.

Once you are done adding or updating requirements, rebuild your docker environment:

$ make build

If there are problems, it'll tell you.

Documentation

Documentation for Pontoon is built with Sphinx and is available on ReadTheDocs.

Building docs is not covered with docker yet, so you will have to do it on your host. To make a virtualenv to build docs, do this:

$ cd docs/
$ virtualenv venv
$ source venv/bin/activate
$ pip install --require-hashes -r requirements/default.txt

Then, to build the docs, run this:

$ make html

The HTML documentation will be in docs/_build/html/. Try to open docs/_build/html/index.html for example.

Note

Pontoon uses GraphViz as part of the documentation generation, so you'll need to install it to generate graphs that use it. Most package managers, including Homebrew, have a package available for install.

Running tests

To run the entire test suite, do:

$ make test

To run only the frontend tests:

$ make jest

To run only the Python tests:

$ make pytest

To run specific tests or specify arguments, you'll want to start a shell in the test container:

$ make shell

Then you can run tests as you like.

Running all the unittests (make sure you run ./manage.py collectstatic first):

app@...:/app$ pytest

Running a directory of tests:

app@...:/app$ pytest pontoon/base/

Running a file of tests:

app@...:/app$ pytest pontoon/base/tests/test_views.py

Writing tests

Put your tests in the tests/ directory of the appropriate app in pontoon/.

Mock usage

Mock is a python library for mocks objects. This allows us to write isolated tests by simulating services besides using the real ones. Best examples are existing tests which admittedly do mocking different depending on the context.

Tip! Try to mock in limited context so that individual tests don't affect other tests. Use context managers instead of monkey patching imported modules.

Updating Your Local Instance

When changes are merged into the main Pontoon repository, you'll want to update your local development instance to reflect the latest version of the site. You can use Git as normal to pull the latest changes, but if the changes add any new dependencies or alter the database, you'll want to install any new libraries and run any new migrations.

If you're unsure what needs to be run, it's safe to just perform all of these steps, as they don't affect your setup if nothing has changed:

# Pull the latest code (assuming you've already checked out master).
git pull origin master

# Install new dependencies or update existing ones.
pip2 install -U --force --require-hashes -r requirements/default.txt

# Run database migrations.
python manage.py migrate

Building front-end resources

We use webpack to build our JavaScript files for some pages. While make build will build those files for you, you might want to rebuild them while programming on the front. To build the files just once, run:

$ make build-frontend

If you want to have those files be built automatically when you make changes, you can run:

$ make build-frontend-w

Integration with fluent

Pontoon is able to synchronize translations produced by libraries provided by Project Fluent and provides an advanced editor for translators.

Because of our very close integration, we'll need to compile the fresh versions of javascript/python libraries in order to provide new features.

It's important to remember to update both packages:

  • python-fluent (responsible for e.g. server-side sync process)
  • fluent-syntax (required by the fluent editor)