Dymmond Lazy Settings

🚀 Generic settings system applied to any application. 🚀

---

Documentation: https://settings.dymmond.com 📚

Source Code: https://github.com/dymmond/dymmond-settings

The official supported version is always the latest released.

---

Motivation

Not overcomplicating what it should be easy by default.

In the current technological environment almost for certain, every system needs some sort of settings. These settings are often inside a settings.py file or similar and then internally is called everywhere needed.

The problem encountered is the fact that those are not considered lazy.

What does that mean? Means that you often face issues when trying to get your application up and running and those settings need be loaded beforehand but not evaluated until the time "comes".

The dymmond-settings was created with the purpose of addressing those same lazy settings problems.

The current Dymmond ecosystem uses a similar system an it was where this package came from.

No hard dependencies, just pure Python using dataclasses. Yes, you read properly, the settings system is built on top of the default Python dataclasses, which means, you can take full advantage of the built-in in your favour.

Of course some defaults are provided and some extra functionality on top to make sure you can use it without too much hassle.

Installation

To install, you can simply run.

$ pip install dymmond-settings

How to use it

This is the best part. Inspired by the way the Dymmond ecosystem (Esmerald, Edgy, Saffier, Mongoz) operates, these settings aim to mimic exactly the same behaviour.

You can simply import the global Settings object and extend and then call the settings lazy object anywhere in your code without the risk of blowing up beforehand.

Defaults

The settings system provides some defaults for you that usually are common in every application but you are not entitled to use them.

These defaults can be overridden at any given time, of course.

Attributes

• debug - Boolean flag generally used to indicate if the application should be in debug mode or not. This is particularly useful if you want to separate some settings by environment.

  ^{Default: False}

• environment - String value indicating the the environment of the settings. This can be particularly useful if you want to show specific settings by the given environment. You can use the provided values from dymmond_settings.enums.EnvironmentType or create your own.

  ^{Default: production}

• version - A value indicating a possible the version of your application running the settings. It defaults to the current version of dymmond_settings if nothing is provided.

  ^{Default: dymmond_settings.__version__}

Functions

Although Python dataclasses also provides that same functionality for you, dymmond_settings makes it easier for you to use it.

All dymmond_settings objects can be used with any of the normal dataclasses functionalities and applications.

dict()

This simply converts your settings module into a python dictionary. This provides the same level of functionality of asdict from the Python dataclasses module when exclude_none is set to False (default).

Parameters:

• exclude_none - Excludes the values containing None.
• upper - Boolean flag indicating if the keys should be in upper case.

from dataclasses import dataclass, field

from dymmond_settings import Settings


@dataclass
class AppSettings(Settings):
    ...


my_settings = AppSettings()
my_settings.dict()

Or if you want to exclude the None values.

from dataclasses import dataclass, field

from dymmond_settings import Settings


@dataclass
class AppSettings(Settings):
    ...


my_settings = AppSettings()
my_settings.dict(exclude_none=True)

Or if you want the keys to be upper case.

from dataclasses import dataclass, field

from dymmond_settings import Settings


@dataclass
class AppSettings(Settings):
    ...


my_settings = AppSettings()
my_settings.dict(upper=True)

tuple()

This simply converts your settings module into a python tuple but is slightly different from the astuple from Python dataclasses module.

The default from dataclasses provides a tuple with all the values of the object provided astuple whereas dymmond_settings tuple function provides a list of tuples key/pair valued.

As per dict functionality, the tuple() also provides a exclude_none in case you want a list attributes with the values set.

Parameters:

• exclude_none - Excludes the values containing None.
• upper - Boolean flag indicating if the keys should be in upper case.

from dataclasses import dataclass, field

from dymmond_settings import Settings


@dataclass
class AppSettings(Settings):
    ...


my_settings = AppSettings()
my_settings.tuple()

Or if you want to exclude the None values.

from dataclasses import dataclass, field

from dymmond_settings import Settings


@dataclass
class AppSettings(Settings):
    ...


my_settings = AppSettings()
my_settings.tuple(exclude_none=True)

Or if you want the tuple to contain the keys in upper case.

from dataclasses import dataclass, field

from dymmond_settings import Settings


@dataclass
class AppSettings(Settings):
    ...


my_settings = AppSettings()
my_settings.tuple(upper=True)

How to use it

There are many ways you can use the settings and all of them valid. Here we show just a few examples how to use it.

The auto detection

When using the settings system, you kinda need to be able to tell your application where to look up for the settings when loading, right?

Well yes and no, if you don't want to rely of the auto-load of the settings for you and you want to take charge of if, you can simply skip this but if you don't, well, this is the way.

As mentioned before, to make sure you use the automatic detection of the settings in your application you need to set an environment variable, called SETTINGS_MODULE.

When the SETTINGS_MODULE is set and pointing to your settings, this will make the settings object from dymmond_settings aware of it and make it available globally and in lazy mode.

Confusing? Well, let us see an example.

Let us imagine we have a settings.py file somewhere in the code, example: myapp/config/settings.py.

Let us also create our custom application settings and provide some values, like this:

from dataclasses import dataclass, field

from dymmond_settings import Settings


@dataclass
class AppSettings(Settings):
    debug: bool = False
    my_setting: str = field(default="works")

Simple so far, right? Ok, no we need to make sure our system is aware of these new settings and for that we will set the SETTINGS_MODULE to point it out to our custom settings. Like this:

$ SETTINGS_MODULE='myapp.config.settings.AppSettings' python ...

And that is it, really! Although looking very simple, there is a lot of magic happening behind the scenes that you don't need to worry about.

Now, because we made the application aware, we are also able to use the settings object provided by the dymmond_settings anywhere in your code base by simply importing it.

Like this:

from dymmond_settings import settings


def get_my_settings() -> str:
    return settings.my_setting

This simple settings is very powerful and lazy, which makes it perfect if you want to use it anywhere in your codebase without those annoying imports happening all over the place.