nubby 0.0.2

Config loader that injects configs using Bevy.

Nubby

A simple config loader for Python using Bevy for dependency injection. You just need to create a model, declare what file name it should look for, and mark that model as a dependency wherever you need to have it made available.

Installation

pip install nubby

You can optionally install the toml or yaml extras to support those file formats.

pip install nubby[toml]
pip install nubby[yaml]

Out of the box Nubby will use tomllib to read TOML files but the toml extra installs the tomlkit package which allows writing to TOML files.

Usage

Nubby is designed to have the smallest possible API surface area. You just need to create a model type that implements the ConfigModel interface and declare the file name it should look for. Then you can inject that model wherever you need it using Bevy.

The ConfigModel type provides some helper functionality for setting the file name as a class keyword argument. Alternatively you can set the __config_filename__ attribute on the class directly.

The file name shouldn't include the file extension. Nubby's file handlers look for supported file extensions in the search paths. By default, the current working directory is the only search path. You can add more paths using the nubby.controllers.get_active_controller().add_path method. It is also possible to add more file handlers using the nubby.controllers.get_active_controller().add_handler method.

Here's a basic example of loading a model from a file.

from dataclasses import dataclass
from bevy import inject, dependency
from nubby import ConfigModel

@dataclass
class Person(ConfigModel, filename="person_info"):
    name: str
    age: int

    def to_dict(self):
        return {"name": self.name, "age": self.age}

@inject
def print_person_details(person: Person = dependency()):
    print(f"{person.name} is {person.age} years old")

Running print_person_details() will print the name and age of the person loaded from the person_info file. Depending on the available file handlers it could be a json, toml, or yaml file.

Note that it is not necessary to pass anything to the print_person_details function. Bevy automatically injects the Person model and loads the appropriate config file.

To save any modifications to a model just call the active controller's save method, passing it the updated model.

from dataclasses import dataclass
from nubby import ConfigModel
from nubby.controllers import get_active_controller

@dataclass
class Person(ConfigModel, filename="person_info"):
    name: str
    age: int

    def to_dict(self):
        return {"name": self.name, "age": self.age}

person = Person("Bob", 31)
get_active_controller().save(person)

The save method will update the in-memory cache as well as the file on disk. It will not update any models that have already been created.