Committing your settings¶

Stela is better suited for new projects. This is because, traditionally, the .env file is not committed on python projects. So, you need special attention when porting an existing project to Stela.

Changes on .gitignore¶

By default, Stela will add your .env.local and .env.*.local files to project's .gitignore and will comment the .env entry if it exists inside .gitignore too. This is because you shouldn't commit your secret values, even if you want to use them in your local machine. But some values in this file aren't secrets, and you may want to commit these values. That's why you combine both .env and .env.local files.

For example, you may want to commit the API_URL and a fake value, but not the real one for the API_TOKEN:

# .env - can be committed
# the value for API_TOKEN here can be used in your CI/CD
# for running tests, for example
API_URL=https://foo.bar
API_TOKEN=foo

# .env.local - Do not commit! The value for API_TOKEN here is as secret!
# You don't need to repeat the API_URL here, because
# it will be loaded from .env file
API_TOKEN=real_token

The standard python .gitignore need to be updated to reflect this:

# .gitignore

# .env  # Comment or remove this line
.env.local  # Add this
.env.*.local  # Add this

Use the stela init command

The command stela init will make these changes for you on an existing .gitignore file. If you don't have a .gitignore file, the command will create one for you. And if you have a existing .env file in your project, Stela will automatically rename this file to .env.local

Be careful with the original .env file

If you already use a .env file in your project, chances are high this file contains secrets. Please, make sure you rename this file to .env.local (manually or using stela init) or split his contents between these two files (the .env for settings and .env.local for secrets).

I'm not comfortable changing these settings. Can I use Stela?¶

Sure, just don't run the stela init file and create manually the .stela file, renaming the dotenv file to be used (we suggest: .environs), like this:

# A very conservative configuration.
[stela]
environment_variable_name = STELA_ENV
evaluate_data = False
show_logs = False
env_file = .environs
config_file_path = .

In your .gitignore just add:

.environs.local
.environs.*.local

Stela will read the original .env file, and will overwrite the values using the .environs files.

In the next step, we will look closely on how to use Stela in your code.