为什么两个相同长度的字符串之间的两个不同的相等测试需要不同的时间

时间:2012-09-16 14:35:19

标签: python time string-comparison

根据Python reference manual

  

使用字符的数字等价物(内置函数ord()的结果)按字典顺序比较字符串。在这种行为中,Unicode和8位字符串是完全可互操作的。

含义散列不用于此目的。

现在,让我们假设在相等测试期间内部Python首先检查两个字符串的长度并继续进行词典比较,如果两者的长度相同(我认为它对所有其他比较也是如此)。

那么,如果是这样,为什么以下两个不同的比较会消耗不同的时间?

>>> str1 = "foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge1"
>>> 
>>> str2 = "foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge2"
>>> 
>>> def compare():
     t1 = time.time()
     for x in xrange(100000000):
         str1 == str2
     print time.time() - t1

>>> 
>>> compare()
13.001019001
>>> 
>>> str2 = "foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge1"
>>> 
>>> compare()
7.41645097733

在比较中str1str2的长度相同。在第一个中,两者的区别仅在于 last 字符,而在第二个中则不同。

PS:使用长字符串和大型迭代来区分差异。

4 个答案:

答案 0 :(得分:1)

在Python中有一些字符串文字池,就像在许多语言中一样,在第二种情况下,两个字符串实际上都是同一个对象,因此它们通过引用进行比较,而不是通过实际值进行比较。

基于参考语言的典型字符串比较函数:

if (ref(a) == ref(b)) return true;
if (len(a) != len(b)) return false;
return compare_actual_data(a, b);

答案 1 :(得分:1)

正如其他答案所推测的那样,比较is indeed short-circuited if the strings are identical,这是具有相同内容的字符串文字的情况。

答案 2 :(得分:1)

在第二种情况下,这两个字符串实际上是同一个对象,具有相同的id。

str1 = "foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge1"
str2 = "foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge2"

def compare():
    t1 = time.time()
    for x in xrange(100000000):
        str1 == str2
    print time.time() - t1


print "%d, %d" % (id(str1), id(str2))
compare()

str2 = "foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge_foo_bar_cruft_kludge1"

print "%d, %d" % (id(str1), id(str2))
compare()

在第二种情况下,id是相同的,因此python解释器不必逐字节地比较整个字符串。

答案 3 :(得分:0)

如果两个字符串是'interned',则可以通过指针比较来比较它们。也许在这种情况下,字节码生成器注意到字符串是常量并实现它们的内容?

你可以通过显示较小的时间在很大程度上独立(或不是)字符串的长度来证明(或反驳)。

事实上,我得到了这些结果:

$ ./test.py 
17.4877259731
8.61179995537
$ ./test_halflengthstrings.py 
12.6504788399
8.58732700348