Python 字典(Dictionary)与hash

    了解过python字典的人都知道一句话，字典的键必须不可变，所以可以用数，字符串或元组充当，列表就不行。

    那么，为什么可变类型不能当作键呢？

    自定义类型可以做键吗？

    首先，我们看看list做键值的情况

lsta = [1,2,3]
dicta = {lsta:1}

>>>TypeError: unhashable type: 'list'

    看了字典的源码dictobject.c

static int
insertdict(register PyDictObject *mp, PyObject *key, long hash, PyObject *value)

    其中有个关键参数long hash，这个值是通过key调用自己的hash方法计算出来的，每一种不同的类型都定义了自己的hash方法。

    Int类型intobject.c：

PyTypeObject PyInt_Type = {
    ...
        (hashfunc)int_hash,            /* tp_hash */
    ...
}

static long
int_hash(PyIntObject *v)
{
    /* XXX If this is changed, you also need to change the way
       Python's long, float and complex types are hashed. */
    long x = v -> ob_ival;
    if (x == -1)
        x = -2;
    return x;
}

    String类型stringobject.c：

PyTypeObject PyString_Type = {
    ...
        (hashfunc)string_hash,  /* tp_hash */
    ...
}

static long
string_hash(PyStringObject *a)
{
    register Py_ssize_t len;
    register unsigned char *p;
    register long x;

    if (a->ob_shash != -1)
        return a->ob_shash;
    len = Py_SIZE(a);
    p = (unsigned char *) a->ob_sval;
    x = *p << 7;
    while (--len >= 0)
        x = (1000003*x) ^ *p++;
    x ^= Py_SIZE(a);
    if (x == -1)
        x = -2;
    a->ob_shash = x;
    return x;
}

     Tuple类型tupleobject.c：

PyTypeObject PyTuple_Type = {
    ...
        (hashfunc)tuplehash, /* tp_hash */
    ...
}

static long
tuplehash(PyTupleObject *v)
{
    register long x, y;
    register Py_ssize_t len = Py_SIZE(v);
    register PyObject **p;
    long mult = 1000003L;
    x = 0x345678L;
    p = v->ob_item;
    while (--len >= 0) {
        y = PyObject_Hash(*p++);
        if (y == -1)
            return -1;
        x = (x ^ y) * mult;
        /* the cast might truncate len; that doesn't change hash stability */
        mult += (long)(82520L + len + len);
    }
    x += 97531L;
    if (x == -1)
        x = -2;
    return x;
}

 

     List类型listobject.c：

PyTypeObject PyList_Type = {
    ...
    (hashfunc)PyObject_HashNotImplemented,    /* tp_hash */
    ...
}

    PyObject_HashNotImplemented实在基类object中定义的object.c：

long
PyObject_HashNotImplemented(PyObject *self)
{
    PyErr_Format(PyExc_TypeError, "unhashable type: '%.200s'",
             self->ob_type->tp_name);
    return -1;
}

long
PyObject_Hash(PyObject *v)
{
    PyTypeObject *tp = v->ob_type;
    if (tp->tp_hash != NULL)
        return (*tp->tp_hash)(v);
    if (tp->tp_compare == NULL && RICHCOMPARE(tp) == NULL) {
        return _Py_HashPointer(v); /* Use address as hash value */
    }
    /* If there's a cmp but no hash defined, the object can't be hashed */
    return PyObject_HashNotImplemented(v);
}

     从上面就可以看出为什么list不能作为键，因为list类型的hash函数是PyObject_HashNotImplemented，这个函数执行时会抛出TypeError异常。

     同时，从上面tuplehash函数可以看出，tuple类型虽然可以做键值，但是如果tuple容器里的子项无法hash，那么这个tuple依然不能做键值，比如下面把list放到tuple容器中：

lista = [1,2,3]
tuplea = (lista,)
dicta = {tuplea:1}

>>>TypeError: unhashable type: 'list'

     

     对于自定义类型，如果没有重写__hash__函数，则会使用基类object的hash函数，默认返回对象的地址,如果重写了hash函数，则根据新的hash函数

#1）重写hash函数，抛出异常，无法作为键值

class my_class(object):
    def __hash__(self):
        raise TypeError, "unhashable type: my_class"

obj = my_class()
dicta = {obj:1}

>>>TypeError: unhashable type: my_class

#2）重写hash函数，不抛出异常

class my_class2(object):
    def __init__(self, value):
        self.value = value
    def __hash__(self):
        return self.value
        
obj = my_class2(1)
obj2 = my_class2(2)
obj3 = my_class2(1)
dicta = {obj:1, obj2:2, obj3:3}
print hash(obj)
print hash(obj2)
print hash(obj3)
print dicta

>>>18
   1
   2
   1
   {<__main__.my_class2 object at 0x01A94490>: 1, <__main__.my_class2 object at 0x01A944B0>: 2, <__main__.my_class2 object at 0x01A944D0>: 3}

#3）不重写hash函数

class my_class3(object):
    pass
        
obj = my_class3()
obj2 = my_class3()
dicta = {obj:1, obj2:2}
print hash(obj)
print hash(obj2)
print dicta

>>>29050192
   29050224
   {<__main__.my_class3 object at 0x01BB4550>: 1, <__main__.my_class3 object at 0x01BB4570>: 2}