Skip to content

TurboBus is an opinionated implementation of Command Responsibility Segregation pattern in python.

Notifications You must be signed in to change notification settings

cafadev/turbobus

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TurboBus

TurboBus is a powerful Python package designed to streamline the development of software applications adhering to the Command Responsibility Segregation (CRS) pattern.

Installation

pip install turbobus

Simple usage

Let's see an example using python typings. You can omit all the typing stuffs if you want to.

God Mode ⚡

from datetime import date
from turbobus.command import Command, CommandBus, CommandHandler, kw_only_frozen
from turbobus.constants import Provider

# We need to create a Command class that receives the values that the handler will use
# to execute the command. The Command class is a generic class that receives the return

# @kw_only_frozen: is a shortcut decorator for @dataclass(kw_only=True, frozen=True)

# Command[int]: is a generic class that receives a return_type.
# This is useful to check if the handler is returning the correct type
# And allow the CommandBus to know the return type of the command

@kw_only_frozen 
class CalculateAgeCommand(Command[int]):
    birthdate: str | date


# We need to create a CommandHandler class that will receive the Command class.
# The handler class must implement the execute method
    
# CommandHandler[CalculateAgeCommand]: is a generic class that receives the Command class
# this is useful to check if the handler is implementing the correct command class
class CalculateAgeHandler(CommandHandler[CalculateAgeCommand]):

    # The execute method must receive the Command class and return
    # the same type as in the Command class return_type
    def execute(self, cmd: CalculateAgeCommand) -> int:
        birthdate: date = cmd.birthdate if isinstance(cmd.birthdate, date) else date.fromisoformat(cmd.birthdate)

        today = date.today()
        age = today.year - birthdate.year - ((today.month, today.day) < (birthdate.month, birthdate.day))
        return age


# We need to register the Command and Handler in the Provider
# This is necessary to allow the CommandBus to find the correct handler
# to execute the command
Provider.set(CalculateAgeCommand, CalculateAgeHandler)


if __name__ == '__main__':
    # We need to create a CommandBus instance to execute the command
    bus = CommandBus()

    # Here we are executing the CalculateAgeCommand
    # if you're using an IDE that supports type hinting
    # you'll see that the result variable is inferred as int
    # because the CalculateAgeCommand is a generic class
    # that receives int as return_type
    result = bus.execute(
        CalculateAgeCommand(birthdate='1994-03-09')
    )

    print(f'You are {result} years old')

Human Mode (No types, obviously 🙄)

Here's the same example, but without types

from datetime import date
from turbobus.command import Command, CommandBus, CommandHandler, kw_only_frozen
from turbobus.constants import Provider


class CalculateAgeCommand(Command):

    def __init__(self, birthdate):
        self.birthdate = birthdate


class CalculateAgeHandler(CommandHandler):

    def execute(self, cmd: CalculateAgeCommand):
        birthdate = cmd.birthdate if isinstance(cmd.birthdate, date) else date.fromisoformat(cmd.birthdate)

        today = date.today()
        age = today.year - birthdate.year - ((today.month, today.day) < (birthdate.month, birthdate.day))
        return age

Provider.set(CalculateAgeCommand, CalculateAgeHandler)

if __name__ == '__main__':
    bus = CommandBus()

    result = bus.execute(
        CalculateAgeCommand(birthdate='1994-03-09')
    )

    print(f'You are {result} years old')

Dependency injection

In many cases we're going to need to inject dependencies to our command handler. To accomplish that we have the @inject decorator. For example:

from abc import ABC, abstractmethod
from dataclasses import field
import uuid
from turbobus.command import Command, CommandBus, CommandHandler, kw_only_frozen
from turbobus.constants import Provider
from turbobus.injection import inject


# This is a simple Entity to represent a User
@kw_only_frozen
class UserEntity:
    id: uuid.UUID = field(default_factory=uuid.uuid4)
    name: str
    email: str


# We need to define the repository interface
# to save and retrieve users
class UserRepository(ABC):

    @abstractmethod
    def get_by_id(self, id: uuid.UUID) -> UserEntity | None:
        """Get user by id"""

    @abstractmethod
    def save(self, user: UserEntity) -> None:
        """Save user"""


# This is an in-memory implementation of the UserRepository
class UserRepositoryInMemory(UserRepository):

    def __init__(self):
        self._users: dict[uuid.UUID, UserEntity] = {}

    def get_by_id(self, id: uuid.UUID) -> UserEntity | None:
        return self._users.get(id)
    
    def save(self, user: UserEntity) -> None:
        self._users[user.id] = user


# Let's create a command to create a user account
@kw_only_frozen
class CreateUserAccount(Command[None]):
    name: str
    email: str


#  @inject is used to inject the dependencies
@inject
@kw_only_frozen
class CreateUserAccountHandler(CommandHandler[CreateUserAccount]):

    user_repository: UserRepository

    def execute(self, cmd: CreateUserAccount) -> None:
        user = UserEntity(name=cmd.name, email=cmd.email)

        # It's unnecessary to retrieve the user from the repository
        # this is just to demonstrate that the user was saved
        self.user_repository.save(user)
        user = self.user_repository.get_by_id(user.id)

        if user is None:
            raise Exception('User not found')
        
        print(f'Welcome {user.name}!')


Provider.set(UserRepository, UserRepositoryInMemory)
Provider.set(CreateUserAccount, CreateUserAccountHandler)


if __name__ == '__main__':
    bus = CommandBus()

    bus.execute(
        CreateUserAccount(name='Christopher Flores', email='[email protected]')
    )

The @inject decorator also accepts the next parameters:

Alias

The @inject decorator use the typing to resolve the required dependency. With the alias: dict[str, Callable[..., Any]] parameter you can specify a different implementation for the same interface. For example, let's say we have a UserRepository interface and then two different implementations; UserRepositoryInMemory and UserRepositorySQL.

Provider.set(UserRepository, UserRepositoryInMemory)
Provider.set('UserRepositorySQL', UserRepositorySQL)

@inject
class CreateUserAccount:

    user: UserRepository

By default, the @inject will use the UserRepositoryInMemory to provide the dependency. Let's specify the UserRepositorySQL as the provider. To accomplish that we just need to specify the parameter name that we want to override, and then the Provider Key:

Provider.set(UserRepository, UserRepositoryInMemory)
Provider.set(UserRepositorySQL, UserRepositorySQL)

@inject(alias={ 'user': 'UserRepositorySQL' })
class CreateUserAccount:

    user: UserRepository

Only and Exclude

The only and exclude parameters in the @inject decorator allow you to fine-tune which dependencies should be resolved based on their names.

The only parameter specifies a list of dependency names that should exclusively be resolved. All other dependencies will be ignored.

@inject(only=['dependency1', 'dependency2'])
def my_function(dep1: Dependency1, dep2: Dependency2, dep3: Dependency3):
    # Only dep1 and dep2 will be injected
    pass

In this example, only dep1 and dep2 will be injected into the function because they are specified in the only list.

The exclude parameter specifies a list of dependency names that should be excluded from injection.

@inject(exclude=['dependency3'])
def my_function(dep1: Dependency1, dep2: Dependency2, dep3: Dependency3):
    # Dep3 will not be injected
    pass

In this example, dep3 will not be injected into the function because it is specified in the exclude list.

These parameters provide flexibility in controlling which dependencies are resolved, allowing you to customize the injection behavior according to your specific needs.