KeeWee 🥝

Global application state management and recording.
The Keewee library implements an auxiliary class that can be used to track values assigned to class and instance variables at runtime.
One major usecase is to decouple your state-management from your business-logic and keep your code nice and concise.
You can also use it to record statistics about an attribute during runtime.
The library works with regular Python classes or dataclasses but needs little different configuration.

Installing

Install and update using pip

$ pip install -U keewee

A Simple Example

import random

from dataclasses import dataclass, field
from keewee import KeeWee


@dataclass
class PokemonTrainer:
    name: str
    skill_level: int | KeeWee = field(default=KeeWee(), repr=False)


if __name__ == "__main__":
    ash = PokemonTrainer(name="Ash Ketchum", skill_level=0)
    for _ in range(10):
        ash.skill_level = random.randint(1, 10)
    print(KeeWee.dumpd())

{
    'PokemonTrainer': {'skill_level': {"PokemonTrainer(name='Ash Ketchum')": [0, 5, 9, 6, 3, 6, 10, 8, 4, 2, 9]}}
}

Collection Record Modes

When assigning a KeeWee instance to an attribute, the user can customize its internal recording-behavior by providing the mode-option. Currently, there are four different modes, whereas the list-mode is the default setting.

Direct mode

When only the current or resp. the last value is of interest, one can choose the direct mode, where the attribute is directely mapped to its value.

{
    'PokemonTrainer': {'skill_level': {"PokemonTrainer(name='Ash Ketchum')": 3}}
}

List mode

The list-mode keeps all occurring values in an ordered list from the first to the last value this attribute was assigned.
Since this use-case is probably the most common it is also chosen to be the default record-behavior.

{
    'PokemonTrainer': {'skill_level': {"PokemonTrainer(name='Ash Ketchum')": [0, 5, 9, 6, 3, 6, 10, 8, 4, 2, 9]}}
}

Set mode

The set-modes only difference to the list-mode is that duplicates are not tracked.

{
    'PokemonTrainer': {'skill_level': {"PokemonTrainer(name='Ash Ketchum')": {0, 2, 3, 5, 7, 9}}}
}

Datetime to value

If one wants to know exactly at what timestamp the modification took place the dtv (datetime-value)-mode is the best to choose.
Here a new dictionary is created for every attribute and the current time is mapped onto the state change.

{'PokemonTrainer': {'skill_level': {"PokemonTrainer(name='Ash Ketchum')": {
    '15:11:44.976976': 0,
    '15:11:44.976985': 8,
    '15:11:44.976987': 6,
    '15:11:44.976990': 2,
    '15:11:44.976992': 6,
    '15:11:44.976994': 9,
    '15:11:44.976996': 8,
    '15:11:44.976998': 7,
    '15:11:44.977000': 3,
    '15:11:44.977002': 9,
    '15:11:44.977004': 7
}}}}

Numerical Record Modes

The following record modes only work for numerical values, e.g.int or float etc.
The result will look similar to the direct-mode.
Currently there are three numerical record modes

1. sum the sum of all occurring values
2. min the minimal value that has occurred
3. max the maximum value that has occurred

An example usage for taking the sum over all values could look like the following.

@dataclass
class PokemonTrainer:
    name: str
    skill_level: int | KeeWee = field(default=KeeWee(mode='sum'), repr=False)

{
    'PokemonTrainer': {'skill_level': {"PokemonTrainer(name='Ash Ketchum')": 49}}
}

Links

• PyPI Releases: https://pypi.org/project/keewee/