Python中hash的问题

ref：http://heipark.iteye.com/blog/1743819

在下面这个例子里：

class Item(object):  
  
    def __init__(self, foo, bar):  
        self.foo = foo  
        self.bar = bar  
      
    def __repr__(self):  
        return "Item(%s, %s)" % (self.foo, self.bar)  
      
print set([Item('', ''), Item('', '')])  
  
# 输出： set([Item(1, 2), Item(1, 2)])

逻辑上讲，set中的两个对象是貌似相同的，那么set中应该只有一个对象

实际上不是这样

set是根据两个元素的hash value判断这两个对象是不是相同的。元素的hash value是通过hash方法得到的（内部__hash__() magic method）。

根据文档：

All of Python’s immutable built-in objects are hashable; mutable containers (such as lists or dictionaries) are not. Objects which are instances of user-defined classes are hashable by default. They all compare unequal (except with themselves), and their hash value is derived from their<span>id()</span>.

可知道只有非可变对象才可hash，并且instances of user-defined classes的hash value是根据他们的id得到的。这个id（ref：https://docs.python.org/3/library/functions.html#id），可以理解为对象在内存中的地址，所以例子里的输出就不奇怪了

关于__hash__()的自定义实现，文档（ref：https://docs.python.org/3/reference/datamodel.html#object.__hash__）是这么说的：

it is advised to mix together the hash values of the components of the object that also play a part in comparison of objects by packing them into a tuple and hashing the tuple.

并且举了一个例子：

def __hash__(self):
    return hash((self.name, self.nick, self.color))

这里再引入一个概念：hashable，文档是这么写得：

An object ishashableif it has a hash value which never changes during its lifetime (it needs a<span>__hash__()</span>method), and can be compared to other objects (it needs an<span>__eq__()</span>method). Hashable objects which compare equal must have the same hash value.

简单来说，hashable的对象必须实现__hash__ and __equal__两个方法

我们之前说过了hash方法怎么实现，但是仅仅实现hash方法，是不能让刚开始的例子中输出正确的结果的。原因如下（ref：https://docs.python.org/3/reference/datamodel.html#object.__hash__）：

If a class does not define an<span>__eq__()</span>method it should not define a<span>__hash__()</span>operation either; if it defines<span>__eq__()</span>but not<span>__hash__()</span>, its instances will not be usable as items in hashable collections. If a class defines mutable objects and implements an<span>__eq__()</span>method, it should not implement<span>__hash__()</span>, since the implementation of hashable collections requires that a key’s hash value is immutable (if the object’s hash value changes, it will be in the wrong hash bucket).

如果定义了eq，没有定义hash，那么显然，由于hash value不同，刚开始的例子中输出结果是错误的

定义hash的同时要定义eq

之所以要定义eq，是为了处理set中有两个对象的hash value相同，这时候要怎样处理。

eq不是每次把元素放进set里都要调用的。如果某个元素和set中的已有元素的hash value都不同，那就没有调用eq的必要了。如果即使两个元素的hash value不同，也要调用eq的话，就失去了hash的意义

所以在刚开始的例子里加上：

...
    def __eq__(self, other):  
        if isinstance(other, Item):  
            return ((self.foo == other.foo) and (self.bar == other.bar))  
        else:  
            return False  
      
    def __hash__(self):  
        return hash(self.foo + " " + self.bar) 
    ...

就ok了