{% if not project_name %}

Project template

You can create a new project out of this template by using a gist.

Requirements:

• python3
• bash
• wget
• docker and docker-compose in version 1.10.0+

Do the following, replace <project_name> with your desired project name:

wget -O - https://gist.githubusercontent.com/mjainta/2e57c4b04d1bb9c526dc0843b4353aa5/raw/fbf135ae7b0935581c101136f2fd802747b21cf8/start-django-project.sh | bash -s <project_name>

This will create a folder called project_name in the current directory.

{% endif %}# {{ project_name|title }}

Setup working environment

$ make

This will build all necessary docker images and start them up.

Executing tests

(myenv) $ make coverage

This will execute all unittests it finds in the {{ project_name }}.tests module. Also coverage is enabled which ends up in a file called .coverage in the root directory. Coverage also generates a html report in coverage_html_report and after prints a summary to stdout.

By configuration, we aim to get a minimum code coverage of 95% Remember to use coverage before committing so the build will not fail after pushing!

If you do not want coverage to be executed, use make test instead

(myenv) $ make test

Both commands also work with a subset of the tests:

(myenv) $ make coverage TESTMODULE={{ project_name }}.tests.models

Settings for the coverage tool can be found in the .coveragerc file.

Use the Linter

We use pylint for linting functionality.

First, install the pre-commit hook which will check your files with the linter before you commit.

Remember, since it creates a pre-commit hook you need to init the git repository first.

(myenv) $ make install_pre_commit

This will create the file .git/hooks/pre-commit with the desired hook for the linter.

You can also use pylint directly if you want to check every file (except a few which get ignored).

(myenv) $ make lint

Settings for the linter can be found in the .pylintrc file.

The pylint manual can be found here.

Using postgresql

Built-in the template will use a postgresql database, instead of the sqlite default.

If you want to use sqlite instead, delete the according lines from the files:

• [base.py]({{ project_name }}/settings/base.py)
• [base.py]({{ project_name }}/settings/test.py)
• In base.txt remove the psycopg2 library.
• In docker-compose.yml remove the postgres container and the link for it.

Connect to it via make psql.

Data is not preserved by default! Modify docker-compose.yml to use volumes if you need to.

Documentation on the postgresql image is found on hub.docker.com.

Starting an interactive python shell

make shell

Running custom commands in the python container

$ docker-compose run --rm py <your command>

Deployment

The requirements the loaded in the docker build process during make images.

Requirements are held in their files, respectively: test / development / staging / production

base should be included for every stage.

GitLab

You can use GitLab-CI for deployment, the .gitlab-ci.yml may help in setting up the build and deployment system.

Out of the box it has to stages:

• build: Creates a new docker container and saves it into the gitlab Registry
• test: Uses the created container for executing tests and linting