LooseCoupling Requirement

main文件中，调用底层同功能模块，

一般写法是在 main中，显示引用底层模块。

这构成了 下层对象 直接出现在上层代码中， 上层代码改变，需要考虑是否影响底层代码。

按照依赖倒置原则，需要定义中间抽象层，上层只依赖抽象层， 对底层，上层统一管理，需要时候调用。

IoC Container Demo

https://github.com/fanqingsong/code_snippet/tree/master/python/IoCContainer

main.py -- 负责上层逻辑

service

      hello.py -- 业务

      world.py -- 业务

main.py

'''
https://stackoverflow.com/questions/42401495/how-to-dynamically-import-modules
'''

import importlib
import os

this_file_path = os.path.abspath(__file__)
this_file_dir = os.path.dirname(this_file_path)

service_path = os.path.join(this_file_dir, "service")

service_files = os.listdir(service_path)

services = {}

for one_file in service_files:
    print(one_file)

    file_name = os.path.basename(one_file)
    print(file_name)

    module_name = file_name.replace(".py", "")
    print(module_name)

    module_path = ".".join(["python", "IoCContainer", "service", module_name])
    print(module_path)

    print(os.path.splitext(file_name))

    suffix = os.path.splitext(file_name)[-1]
    if suffix == ".py":
        print("adding-----------------")
        one_module = importlib.import_module(module_path)
        one_service_class = getattr(one_module, "Service")
        services[module_name] = one_service_class()

print("=======================")
services['hello'].run()

hello.py

class Service():
    def run(self):
        print('hello service')

world.py

class Service():
    def run(self):
        print('world service')

参考

模块文件动态加载

https://stackoverflow.com/questions/42401495/how-to-dynamically-import-modules

https://www.tutorialspoint.com/How-I-can-dynamically-import-Python-module