Sometimes, you just need to save and retain a few settings for your desktop program, like a theme preference, or last viewed directory. This is a library intended to easily load and save simple configuration data to and from disk through a type-checked data class.
First, a basic use case:
pip install simple-file-settings
import os
from simplefilesettings.json import JSONClass
class _Settings(JSONClass):
class Config:
json_file = os.path.join(os.path.expanduser("~"), "config.json")
mqtt_host: str = "mqtt"
mqtt_port: int = 1883
serial_port: str = "COM1"
serial_baud_rate: int = 115200
log_file_directory: str = os.path.join(os.getcwd(), "logs")
force_light_mode: bool = False
joystick_inverted: bool = False
max_moving_map_tracks: int = 5000
takeoff_height: float = 3
Settings = _Settings()
# this will attempt to load the value from the file on disk, or revert to the default
print(Settings.serial_port)
# this will save the change to the config file
Settings.serial_port = "/dev/tty1"
Inherit simplefilesettings.json.JSONClass
and add class attributes with
type hints and optionally default values. Attributes without type hints will
not be loaded or saved. Attributes starting with an underscore will cause an error.
If a default is not provided, None
is assumed.
from simplefilesettings.json import JSONClass
class _Settings(JSONClass):
name: str = "John" # valid
age = 26 # invalid
_height_cm: int # invalid
By default, a JSON file called settings.json
in the current working directory
is used. To change this, add a nested class called Config
with an attribute
json_file
. This accepts any path-like variable.
import os
from simplefilesettings.json import JSONClass
class _Settings(JSONClass):
class Config:
json_file = os.path.join(os.path.expanduser("~"), "config.json")
name: str = "John"
Data types need to serializable for the selected file format (JSON, TOML, YAML, see below). The following additional types are also supported:
datetime.datetime
datetime.date
datetime.time
datetime.timedelta
enum.Enum
pathlib.Path
By default, when any attribute is accessed, the configured file will be read.
If the file does not exist, the default value will be used.
If the file has a parse error, it will be deleted automatically.
To only read the file one time, set the Config
value always_read
to False
.
When any attribute has its value set, that will be written to the configured file.
from simplefilesettings.json import JSONClass
class _Settings(JSONClass):
name: str = "John"
Settings = _Settings()
print(Settings.name)
Settings.name = "Bob"
Running this twice will print John
the first time and Bob
the second time.
If pure JSON isn't your thing, TOML, YAML, and JSON5 are available with the
[toml]
, [yaml]
, [json5]
extras, respectively.
from simplefilesettings.toml import TOMLClass
from simplefilesettings.yaml import YAMLClass
from simplefilesettings.json5 import JSON5Class
class _TSettings(TOMLClass):
name: str = "Tom"
class Config:
toml_file = os.path.join(os.path.expanduser("~"), "config.toml")
class _YSettings(YAMLClass):
name: str = "Ingy"
class Config:
yaml_file = os.path.join(os.path.expanduser("~"), "config.yaml")
class _JSettings(JSON5Class):
name: str = "Douglas"
class Config:
json5_file = os.path.join(os.path.expanduser("~"), "config.jsonc")
Use the provided devcontainer or run the following for local development:
# Install uv
# https://docs.astral.sh/uv/getting-started/installation/
uv tool install vscode-task-runner
vtr install