First Class Objects

First-class functions：在程式語言中，functions可以被當作是一個物件或變數。

因此，Python中的functions是first-class objects。並具有下列幾項功能：

1. 指派到一個變數：

   def hello():
       print("Hello World")
   
   greet = hello
   greet()
   
   --------------------------------------------------
   # Hello World
   

2. 可以將一個function當作另一個functions的參數(argument)：

   (The same idea from Calculus or discrete math - composite function.)

   def hello():
       print("Hello World")
   
   def call_another_function(other_func):
       # other_func = hello
       other_func()
   
   call_another_function(hello)
   
   --------------------------------------------------
   # Hello World
   

   def square(num):
       return num ** 2
   
   my_list = [1, 2, 3, 4]
   result = map(square, my_list)
   for i in result:
   		print(i)
   
   --------------------------------------------------
   # 1
   # 4
   # 9
   # 16
   

3. functions可以回傳一個function。

   def hello():
       def greet():
           print("greet!!")
       return greet
   
   welcome = hello()
   welcome()
   
   --------------------------------------------------
   # greet!!
   

   def hello(name):
       def greet(another_name):
           print("hello, " + another_name)
   
       def bye(another_name):
           print("bye, " + another_name)
   
       if name == "greet":
           return greet
       else:
           return bye
   
   welcome = hello("greet")
   goodbye = hello("something else")
   welcome("Wilson")
   goodbye("Grace")
   
   --------------------------------------------------
   # hello, Wilson
   # bye, grace
   

4. 可以放入list、tuple或dictionary。

   def hello():
       print("hello 1")
   
   def hello2():
       print("hello 2")
   
   for func in [hello, hello2]:
       func()
   
   --------------------------------------------------
   # hello 1
   # hello 2
   

• 在JavaScript、Python和其他某些程式語言中，functions被當作「一等公民(first-class citizens)」。因此，functions與strings、numbers和其他物件具有相同的privilege。
• 在Java、C++中，functions是second-class objects，所以不適用上面四點。
• higher-order function：一個function將另一個function當作input。

Decorators

• 在Python中，常會使用一個higher-order function將其他function包在裡面。使我們在執行該function的前後執行其他程式碼。

  Example:

  def new_decorator(original_func):
      def wrap_func():
          print("Here is some codes before the original function.")
          original_func()
          print("Here is some codes after the original function.")
      return wrap_func
  
  def func_needs_decorator():
      print("I am a function that needs decorator.")
  
  decorated_function = new_decorator(func_needs_decorator) # wrap_func(func_needs_decorator)
  decorated_function()
  
  --------------------------------------------------
  # Here is some codes before the original function.
  # I am a function that needs decorator.
  # Here is some codes after the original function.
  

• 因為此作法太常見，因此Python提供一個名為decorator的語法糖來簡化這個過程。

  Example:

  def new_decorator(original_func):
      def wrap_func():
          print("Here is some codes before the original function.")
          original_func()
          print("Here is some codes after the original function.")
      return wrap_func
  
  @new_decorator # @ 後面接想要被哪個higher-order function包起來
  def func_needs_decorator():
      print("I am a function that need decorator.")
  
  func_needs_decorator()
  
  --------------------------------------------------
  # Here is some codes before the original function.
  # I am a function that needs decorator.
  # Here is some codes after the original function.
  

• Decorators在Python frameworks(其他人用Python寫的有用工具)中很常見。

  例如：Flask就是一個Python framework，可以用來建立網頁伺服器(web server)。

  若想建立一個網頁伺服器，傳統上要設定TCP和HTTP protocols(網際網路協議)。但在Flask中，可以直接使用Flask提供的decorator，快速建立一個網頁伺服器。

  from flask import Flask
  
  app = Flask(__name__)
  
  @app.route("/hello")
  def hello():
      return "Hello World!"
  
  if __name__ == "__main__":
      app.run(debug=True)
  

Generator

Recall the generator syntax that we learned from chapter 3. (The one that is very similar to comprehension.)

• yield

  # for loop
  def cube(n):
      result = []
      for x in range(n):
          result.append(x ** 3)
      return result
  
  for i in cube(10):
      print(i)
  
  # disadvantage:
  # result的長度越長，占用的記憶體位置越多。
  

  # generator
  def cube(n):
      for x in range(n):
          yield x ** 3
  
  print(cube(10)) # <generator object cube at ...>
  
  for element in cube(10):
      print(element)
  

• yield from

  def sub_generator(x):
  		for i in range(x):
  				yield i ** 2
  
  def gen(y):
  		yield from sub_generator(y)
  
  for num in gen(15):
  		print(num)
  

  使用情境：

  def get_all_users():                # 可以取得所有使用者的id，但若想單獨取老師或學生
      for id in get_teachers_id():    # 的部分，需要重新定義下列兩個函數。
          yield id
      for id in get_students_id():
          yield id
  
  def teacher_id_generator():
      for i in get_teachers_id():
          yield id
  
  def student_id_generator():
      for i in get_students_id():
          yield id
  

  def get_all_users():                   # 主要用於協助重新架構程式碼，使程式碼更易於
      yield from teacher_id_generator()  # 維護，並減少重複的程式碼。
      yield from student_id_generator()
  
  def teacher_id_generator():
      for i in get_teachers_id():
          yield id
  
  def student_id_generator():
      for i in get_students_id():
          yield id
  

Iteration, Iterable, and Iterator