classconfig

Package for creating (yaml) configuration files automatically and loading objects from those configuration files.

Installation

You can install it using pip:

pip install classconfig

Usage

At first we need a class that is configurable it means that it has ConfigurableAttribute class members. Such as ConfigurableValue or ConfigurableFactory. Let's create two simple classes where one of them will have the other instance as a member:

from classconfig import ConfigurableValue, ConfigurableFactory, ConfigurableMixin


class Inventory(ConfigurableMixin):
    size: int = ConfigurableValue(desc="Size of an inventory", user_default=10, validator=lambda x: x > 0)


class Character(ConfigurableMixin):
    lvl: int = ConfigurableValue(desc="Level of a character", user_default=1, validator=lambda x: x > 0)
    name: str = ConfigurableValue(desc="Name of a character")
    inventory: Inventory = ConfigurableFactory(desc="Character's inventory", cls_type=Inventory)

You can see that the usage is similar to dataclasses as it also uses descriptors. You can omit the ConfigurableMixin inheritance but then you will have to write your own __init__ method e.g.:

class Inventory:
    size: int = ConfigurableValue(desc="Size of an inventory", user_default=10, validator=lambda x: x > 0)

    def __init__(self, size: int):
        self.size = size

Creating configuration file

Now let's create a configuration file for our Character class. You can do it by calling save method of Config class:

from classconfig import Config

Config(Character).save("config.yaml")

You will get a file with the following content:

lvl: 1  # Level of a character
name: # Name of a character
inventory: # Character's inventory
  size: 10  # Size of an inventory

Validation

As you have seen in the previous example, it is possible to add a validator. A validator could be any callable that takes one argument and return True when valid. You can also raise an exception if the argument is invalid to specify the reason for the failure.

There are various predefined validators in classconfig.validators module.

Transformation

It is possible to specify a transformation (transform attribute) that will transform user input. The transformation is done before the validation. Thus, it can be used to transform input into valid form.

It can be any callable that takes one argument and returns the transformed value, but there also exist some predefined transformations in classconfig.transforms module.

Loading

Now let's load the configuration file we just created and create an instance of Character class:

from classconfig import Config, ConfigurableFactory

config = Config(Character).load(path)   # load configuration file
loaded_obj = ConfigurableFactory(Character).create(config)  # create an instance of Character class

Why YAML?

YAML is a human-readable data serialization language. It is easy to read and write. It is also easy to parse and generate.

It supports hierarchical data structures, which are very useful when you need to represent configuration of a class that has other configurable classes as members.

It supports comments, unlike e.g. json, which is also a big advantage.