centralized-data 0.0.4

Package for cantralized-data, such as singletons, bindable singletons, global collections, dynamically loaded objects in Python.

Singleton objects similar to Angular Services

Installation

pip install centralized_data

Usage 1 - binding to properties

The singleton object is first initialized whenever it is bound to a property or when the first constructor is called.

from typing import override
from centralized_data import Bindable

class MyBindable(Bindable):
    @override
    def constructor(self) -> None:
        self.value = 3

    def method(self) -> str:
        return "Success"

class MyUsingClass:
    @MyBindable.bind
    def my_binding(self) -> MyBindable: ... # This code is overridden by the binding function. Just use ... or pass.

    def some_method(self):
        print(self.my_binding.value)
        print(self.my_binding.method())

Usage 2 - Singleton

The singleton object is first initialized whenever the first constructor is called.

from bindable import Bindable

class MyBindable(Bindable):
    @override
    def constructor(self) -> None:
        self.value = 3

    def method(self) -> str:
        return "Success"

def do_something() -> None:
    my_bindable = MyBindable()
    print(my_bindable.value)
    print(my_bindable.method())
    print(MyBindable().value) # always returns the same object.
    print(MyBindable().method()) # always returns the same object.

Usage 3 - Global Collections

The singleton objects are initialized when first accessed.

from bindable import GlobalCollection

class Customer(GlobalCollection[int]):
    @override
    def constructor(self, key: int) -> None:
        super().constructor(key)
        self.load_from_database()

    def load_from_database(self) -> None:
        record = database.sql('select name, address from customers')
        self.name = record['name']
        self.address = record['address']

Customer(1)
Customer(2).name = 'john'
Customer(2).name # already exists and is therefore not going to call the constructor again
Customer.clear()

Asset Loader

• You have multiple Python classes deriving from another python class that extend it, but aren't directly referenced by your code.

/project
|- task.py
|- task_list.py
/assets
  /tasks
  |- write_hello_world_every_minute.py # is descendant of a class from task.py
  |- daily_plan_for_world_dominance.py # is descendant of a class from task.py
  |- monthly_defeat_the_demon_king.py # is descendant of a class from task.py

• You have multiple yaml/json-Files that describe a workflow or a class in your code

/project
|- table.py
/assets
  /db_tables
  |- customer.yaml
  |- article.yaml
  |- shop.yaml

Usage 1 - use the existing templates for Yaml/JSON/Py Files

Yaml/JSON

• Override YamlAsset/JSONAsset.
  • you can access the self.source object for information retrieval.
  • I recommend you to make @properties for retrieving the information for cleaner code.
• Override YamlAssetLoader[T]/JSONAssetLoader[T] with your YamlAsset/JSONAsset derivative
  • define the asset_folder_name() and asset_class() methods.
• Create your yaml/json files inside the folder */assets/<asset_folder_name()>/

class MyTable(YamlAsset):
    @property
    def some_field(self):
        return self.source.get("some_field")

class MyTables(YamlAssetLoader[MyTable]):
    @override
    def asset_class(self) -> Type[MyTable]: return MyTable
    @override
    def asset_folder_name(self) -> str: return 'db_tables'

Python

• Override PythonAsset with your base Python class.
  • define the classmethod base_asset_class_name to return the class name of your main derivative.
  • define your properties and methods that you want to override in your asseted classes.
• Override PythonAssetLoader[T] with your PythonAsset derivative and define the asset_folder_name() method.
• Create your python files inside the folder */assets/<asset_folder_name()>/
  • These modules will only be loaded regardless of wheter they contain a derivative of your PythonAsset derivative or not.
  • The derivatives are automatically instantiated and added into your Loader's loaded_assets.

class Task(PythonAsset):
    @abstractmethod
    def type(self) -> Enum: ...
    @abstractmethod
    async def execute(self) -> any:

class Tasks(PythonAssetLoader[Task]):
    @override
    def asset_folder_name(self) -> str: return 'tasks'

    def get_task(self, type: Enum) -> Task:
        return next([task for task in self.loaded_assets if task.type() == type])

assets/tasks/daily_plan_for_world_dominance.py
...

class PlanForWorldDominance(Task):
    @override
    def type(self) -> Enum: return ...
    @override
    async def execute(self) -> any: print('veni vidi vici')

Usage 2 - implement your own loader

• Simply override the LoadedAsset and the AssetLoader in a similar manner to the existing derivatives.