CS入门学习笔记14-MIT 6.00.1x- Lecture 12 Object Oriented Programming (Class Generator

Lectrue 12

1. Inheritance

inheritance 指的是不仅要建立class，还要在一个class中建立许多的subclass，让它们都可以共享主class中的standard。

1.1 Using inheritance

Let’s build an application that organizes info about people!
– Person: name, birthday
- Get last name
- Sort by last name
- Get age

import datetimeclass Person(object):def __init__(self, name):'''create a person called name'''self.name = nameself.birthday = Noneself.lastName = name.split(' ')[-1]def getLastName(self):"""return self's last name"""return self.lastNamedef __lt__(self, other):'''return True if self's ame is lexicographically less thanother's name, and False otherwise'''if self.lastName == other.lastName:return self.name < other.namereturn self.lastName < other.lastNamedef setBirthday(self,month, day, year):'''sets selfs birthday to birthDate'''self.birthday =datetime.date(year, month, day)def getAge(self):'''returns self's current age in days'''if self.birthday == None:raise ValueErrorreturn (datetime.date.today() - self.birthday).daysdef __str__(self):"""return self's name"""return self.name

1.2 how sort() work in python
用了一个叫做 __lt__的内置method
1.3 进一步丰富此例子

class MITPerson(Person): # This is a subclass of Person -- inheritancenextIdNum = 0 # Next ID number to assign,\# this is a class attribute, not an attribute for instancesdef __init__(self, name):Person.__init__(self, name) #initialize Person attributes# new MITPerson attribute: a unique ID numberself.idNum = MITPerson.nextIdNum # idNum is an instance attribute# 此处idNum需要调用class attribute nextIdNum的值MITPerson.nextIdNum += 1def getIdNum(self):return self.idNum# sorting MIT people uses their ID number, not name!def __lt__(self, other):return self.idNum < other.idNump1 = MITPerson('Eric')
p2 = MITPerson('John')
p3 = MITPerson('John')
p4 = Person('John')>p1 < p2 # 判断哪个在前哪个在后
>>> True
>p3 < p2
>>> False
>p4 < p1
>>> False
>p1 < p4
>>>报错 AttributeError
"""
此处报错的原因在于
This __lt__ method "shadows" the Person method, meaning that if we compare an MITPerson object,since its environment inherits from the MITPerson class environment, Python will see this version of __lt__ not the Person version. 但是反过来，在MITPerson中定义的__lt__method，只有局部的视野，并不能看到Person中定义的__lt__,所以调用会报错。下图更清晰地解释了：
"""

1.4 继承的两个好处（from 廖雪峰教程）

获得父类的全部功能
多态
当子类和父类都存在相同的run()方法时，我们说，子类的run()覆盖了父类的run()，在代码运行的时候，总是会调用子类的run()。（此处run仅为一个例子）
多态的好处就是，当我们需要传入Dog、Cat、Tortoise……时，我们只需要接收Animal类型就可以了，因为Dog、Cat、Tortoise……都是Animal类型，然后，按照Animal类型进行操作即可。由于Animal类型有run()方法，因此，传入的任意类型，只要是Animal类或者子类，就会自动调用实际类型的run()方法，这就是多态的意思：

对于一个变量，我们只需要知道它是Animal类型，无需确切地知道它的子类型，就可以放心地调用run()方法，而具体调用的run()方法是作用在Animal、Dog、Cat还是Tortoise对象上，由运行时该对象的确切类型决定，这就是多态真正的威力：调用方只管调用，不管细节，而当我们新增一种Animal的子类时，只要确保run()方法编写正确，不用管原来的代码是如何调用的。这就是著名的“开闭”原则：

对扩展开放：允许新增Animal子类；

对修改封闭：不需要修改依赖Animal类型的run_twice()等函数。

1.5 作业题中启示

关于父类与子类的调用关系：

# -*- coding: utf-8 -*-
class Spell(object):def __init__(self, incantation, name):self.name = nameself.incantation = incantationdef __str__(self):return self.name + ' ' + self.incantation + '\n' + self.getDescription()def getDescription(self):return 'No description'def execute(self):print self.incantation    class Accio(Spell):def __init__(self):Spell.__init__(self, 'Accio', 'Summoning Charm')class Confundo(Spell):def __init__(self):Spell.__init__(self, 'Confundo', 'Confundus Charm')def getDescription(self):return 'Causes the victim to become confused and befuddled.'def studySpell(spell):print spellspell = Accio()
# note 这里输出Accio 是因为Spell 的 __init__函数里面的参数写反了！
# note spell 的 name = Summoning Charm incantation = Accio
spell.execute()
# note print函数会去调用 __str__ 将__str__返回的结果输出,
# note __str__ 函数返回的是  Summoning Charm Accio \nNo description  (\n 是换行符号)
studySpell(spell)
# note Confundo 是继承自 Spell的，但是它重载override 了父类的函数， getDescription()
# note 但是由于他没有重载父类的 __str__ 函数，所以print函数依然会调用 Spell里面的__str__ 函数
# note __str__函数里面调用了getDescription函数，因为Confundus自己有getDescription函数，这里会去调用自己的这个函数
# note 返回的就是 Confundus Charm Confundo \n Causes the victim to become confused and befuddled.
studySpell(Confundo())"""
Output:Accio
Summoning Charm Accio
No description
Confundus Charm Confundo
Causes the victim to become confused and befuddled.
"""

2.1 进一步丰富前文例子，形成Class Hierarchy结构

# -*- coding: utf-8 -*-
import datetimeclass Person(object):def __init__(self, name):'''create a person called name'''self.name = nameself.birthday = Noneself.lastName = name.split(' ')[-1]def getLastName(self):"""return self's last name"""return self.lastNamedef __lt__(self, other):'''return True if self's ame is lexicographically less thanother's name, and False otherwise'''if self.lastName == other.lastName:return self.name < other.namereturn self.lastName < other.lastNamedef setBirthday(self,month, day, year):'''sets selfs birthday to birthDate'''self.birthday =datetime.date(year, month, day)def getAge(self):'''returns self's current age in days'''if self.birthday == None:raise ValueErrorreturn (datetime.date.today() - self.birthday).daysdef __str__(self):"""return self's name"""return self.nameclass MITPerson(Person): # This is a subclass of Person -- inheritancenextIdNum = 0 # Next ID number to assign,\# this is a class attribute, not an attribute for instancesdef __init__(self, name):Person.__init__(self, name) #initialize Person attributes# new MITPerson attribute: a unique ID numberself.idNum = MITPerson.nextIdNum # idNum is an instance attribute# 此处idNum需要调用class attribute nextIdNum的值MITPerson.nextIdNum += 1def getIdNum(self):return self.idNum# sorting MIT people uses their ID number, not name!def __lt__(self, other):return self.idNum < other.idNumclass UG(MITPerson):def __init__(self, name, classYear):MITPerson.__init__(self, name)self.year = classYeardef getClass(self):return self.yearclass Grad(MITPerson):# 这里虽然没有定义任何东西，但因为可以直接继承MITPerson中的method和attribute# 此处建立grad的原因是为了区别grad和undergraduate两种学生passdef isStudent(obj):return isinstance(obj, UG) or isinstance(obj, Grad)

如果此时需要添加一种transfer student类别，再次进行isStudent校验时就会出现错误，必须要再修改isStudent函数中的判断才行。——> 这时其实最好的办法是改变整个hierarchy，在mitstudent下再建立一个subclass：student，让三种学生都成为student的subclass。如下图所示：

2.2 Be careful when overriding methods in a subclass!

Sbusititution principle
Important behaviors of superclass should be supported by all subclasses

2.3 在hierarchy中如何调用method以及attribute的取值

class A(object):def __init__(self):self.a = 1def x(self):print "A.x"def y(self):print "A.y"def z(self):print "A.z"class B(A):def __init__(self):A.__init__(self)self.a = 2self.b = 3def y(self):print "B.y"def z(self):print "B.z"class C(object):def __init__(self):self.a = 4self.c = 5def y(self):print "C.y"def z(self):print "C.z"class D(C, B):def __init__(self):C.__init__(self)B.__init__(self)self.d = 6def z(self):print "D.z"> obj = D()
# 找attribute的顺序：按照代码执行的顺序，从最下面开始，一层一层往上找
# 先找D，无a；再找C，a = 4；再找B，B找A，a = 1；再找B自己的，a = 2
> print obj.a
>>> 2
> print obj.b
>>> 3
> print obj.c
>>> 5
> print obj.d
>>> 6# 找method的顺序：从最下面开始找，找到的第一个便直接执行，不再继续向上寻找
> obj.x()
>>> A.X
> obj.y()
>>> C.y
> obj.z()
>>> D.z

2.4 继续补全上述代码，并形成完整example- A Gradebook

# -*- coding: utf-8 -*-
# A Gradebookimport datetimeclass Person(object):def __init__(self, name):"""create a person called name"""self.name = nameself.birthday = Noneself.lastName = name.split(' ')[-1]def getLastName(self):"""return self's last name"""return self.lastNamedef setBirthday(self,month,day,year):"""sets self's birthday to birthDate"""self.birthday = datetime.date(year,month,day)def getAge(self):"""returns self's current age in days"""if self.birthday == None:raise ValueErrorreturn (datetime.date.today() - self.birthday).daysdef __lt__(self, other):"""return True if self's ame is lexicographicallyless than other's name, and False otherwise"""if self.lastName == other.lastName:return self.name < other.namereturn self.lastName < other.lastNamedef __str__(self):"""return self's name"""return self.name# me = Person("William Eric Grimson")
# print me
# me.getLastName()
# me.setBirthday(1,2,1927)
# me.getAge()
# her = Person("Cher")
# her.getLastName()
# plist = [me, her]
# for p in plist: print p
# plist.sort()
# for p in plist: print pclass MITPerson(Person):nextIdNum = 0 # next ID number to assigndef __init__(self, name):Person.__init__(self, name) # initialize Person attributes# new MITPerson attribute: a unique ID numberself.idNum = MITPerson.nextIdNumMITPerson.nextIdNum += 1def getIdNum(self):return self.idNum# sorting MIT people uses their ID number, not name!def __lt__(self, other):return self.idNum < other.idNum# p1 = MITPerson('Eric')
# p2 = MITPerson('John')
# p3 = MITPerson('John')
# p4 = Person('John')# print p1# p1.getIdNum()
# p2.getIdNum()
# p1 < p2
# p3 < p2
# p4 < p1# p1 < p4class UG(MITPerson):def __init__(self, name, classYear):MITPerson.__init__(self, name)self.year = classYeardef getClass(self):return self.yearclass Grad(MITPerson):passdef isStudent(obj):return isinstance(obj,UG) or isinstance(obj,Grad)#s1 = UG('Fred', 2016)
#s2 = Grad('Angela')
#isStudent(s1)
#isStudent(s2)class TransferStudent(MITPerson):pass# go back and define
# class Student(MITPerson)
# change inheritance for UG, Grad and TransferStudent
# change def isStudent(obj):
#            return isinstance(obj, Student)class Grades(object):"""A mapping from students to a list of grades"""def __init__(self):"""Create empty grade book"""self.students = [] # list of Student objectsself.grades = {} # maps idNum -> list of grades# grades这个dictionary的key是学生学号，value是一个list of gradesself.isSorted = True # true if self.students is sorteddef addStudent(self, student):"""Assumes: student is of type StudentAdd student to the grade book"""if student in self.students:raise ValueError('Duplicate student')self.students.append(student)self.grades[student.getIdNum()] = []self.isSorted = Falsedef addGrade(self, student, grade):"""Assumes: grade is a floatAdd grade to the list of grades for student"""try:self.grades[student.getIdNum()].append(grade)except KeyError:raise ValueError('Student not in grade book')def getGrades(self, student):"""Return a list of grades for student"""try: # return copy of student's gradesreturn self.grades[student.getIdNum()][:]#just make a copy of the grade listexcept KeyError:raise ValueError('Student not in grade book')def allStudents(self):"""Return a list of the students in the grade book"""if not self.isSorted:self.students.sort()self.isSorted = Truereturn self.students[:]# return copy of list of studentsdef gradeReport(course):"""Assumes: course if of type grades"""report = []for s in course.allStudents():tot = 0.0numGrades = 0for g in course.getGrades(s):tot += gnumGrades += 1try:average = tot/numGradesreport.append(str(s) + '\'s mean grade is '+ str(average))except ZeroDivisionError:report.append(str(s) + ' has no grades')return '\n'.join(report)ug1 = UG('Jane Doe', 2014)
ug2 = UG('John Doe', 2015)
ug3 = UG('David Henry', 2003)
g1 = Grad('John Henry')
g2 = Grad('George Steinbrenner')six00 = Grades()
six00.addStudent(g1)
six00.addStudent(ug2)
six00.addStudent(ug1)
six00.addStudent(g2)for s in six00.allStudents():six00.addGrade(s, 75)
"""
此处若将代码改写为
for s in six00.students:print s
是能得出一样的结果的，
但是倾向于不这么做，因为this violates the data hiding aspect of an object,
and exposes internal representation.
-- If I were to change how I want to represent a grade book, I should only needto change the methods within that object, not external procedures that use it"""
six00.addGrade(g1, 100)
six00.addGrade(g2, 25)six00.addStudent(ug3)# print gradeReport(six00)

3. generator

1. definition

Any procedure or method with a yield statement is called a generator

def genTest():yield 1yield 2> genTest()
>>> <generator object genTest at 0x201b 878>

Generators have a next() method which starts/resumes execution of the procedure. Inside of generator（在call generator的next()时可能发生的两种情况）:
– yield suspends execution and returns a value
– Returning from a generator raises a StopIteration exception

next() 的作用是先从程序的开头开始执行，直到reach a “yield” statement，这是会推迟执行，并且返回一个value。
当我们准备好执行接下来的代码时，就再call这个generator的next() function。如果在yield后没有其他语句执行，则会返回一个StopIteration

def genTest():yield 1yield 2> foo = genTest()
> foo.next()
>>> 1 # 返回了第一个yield后面的value
> foo.next()
>>> 2 # 继续执行了语句，并返回了第二个yield后面的value
> foo.next()
>>>
StopIterationTraceback (most recent call last)
<ipython-input-43-93eccb67fafc> in <module>()
----> 1 foo.next()StopIteration:

2. 用途

2.1 using inside a looping structure
We can use a generator inside a looping structure, as it will continue until it gets a StopIteration exception:

> for n in genTest():print n
>>> 1
>>> 2
>

# a Fancier example - 计算斐波那契数列
def genFib():fibn_1 = 1 #fib(n-1)fibn_2 = 0 #fib(n-2)while True:#fib(n) = fib(n-1) + fib(n-2)next = fibn_1 + fibn_2yield nextfibn_2 = fibn_1fibn_1 = next# 每次call genFib_instance.next()都会出现数列中下一个数
'''
for n in genFib():print n
# 使用此方法可不断打印出此数列中的数，直到键盘中断为止
'''

** 2.2 Why Generators**

** 相比之下，第一种方法要一次性列出所有人，如果学生很多就会非常低效，而第二种方法每次只给出一个学生的名字，更加的efficient。

2.3 generator例子

'''
Write a generator, genPrimes, that returns the sequence of prime numbers on successive calls to its next() method: 2, 3, 5, 7, 11, ...
如下是两种方法
'''def genPrimes():prime_list = []next = 1while True:next += 1for p in prime_list:if next % p == 0:breakelse:prime_list.append(next)yield next'''
for else 用法：
Loop statements may have an else clause;
it is executed when the loop terminates through exhaustion of the iterable (with for)
or when the condition becomes false (with while),
but not when the loop is terminated by a break statement.
'''def genPrimes():primes = [2]yield primes[0]guess = 3while True:if all(guess%x != 0 for x in primes):primes.append(guess)        if guess == primes[-1]:yield primes[-1]guess += 2
'''
!!!！注意！在使用generator的时候，要先用函数定义一个instance，再调用next(),
而不是直接在generator后面+next。如下方法才是正确的！primeGenerator = genPrimes()
primeGenerator.next()
'''