python 最大递归次数 RuntimeError: maximum recursion depth exceeded

转载出处：点击打开链接

帮别人看代码，偶然遇到这个问题，原来python解释器有一个默认的最大递归次数是999。

举个例子：

def recursion(n):
    if (n <= 1):
        return
    print n
    recursion(n - 1)

print "test 999"
recursion(999)   #正常运行

print "test 1000"
recursion(1000)  #报错，RuntimeError: maximum recursion depth exceeded

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• 9
• 10
• 11
• 12

可以在文件初始位置修改这个最大递归次数，如下：

import sys
sys.setrecursionlimit(10000)  # set the maximum depth as 10000

• 1
• 2
• 3

这时候递归1000次，2000次都不会报错了，但次数很大时比如5000就会报错，又查资料，结论如下：

sys.setrecursionlimit(limit) 
Set the maximum depth of the Python interpreter stack to limit. This limit prevents infinite recursion from causing an overflow of the C stack and crashing Python. 
The highest possible limit is platform-dependent. A user may need to set the limit higher when she has a program that requires deep recursion and a platform that supports a higher limit. This should be done with care, because a too-high limit can lead to a crash.

sys.setrecursionlimit() 只是修改解释器在解释时允许的最大递归次数，此外，限制最大递归次数的还和操作系统有关，经过测试：

windows下最大迭代次数约4400次，linux下最大迭代次数约为24900次（python 2.7 64位）

如下代码可以测试最大迭代次数：

import sys
import itertools

class RecursiveBlowup1:
    def __init__(self):
        self.__init__()

def test_init():
    return RecursiveBlowup1()

class RecursiveBlowup2:
    def __repr__(self):
        return repr(self)

def test_repr():
    return repr(RecursiveBlowup2())

class RecursiveBlowup4:
    def __add__(self, x):
        return x + self

def test_add():
    return RecursiveBlowup4() + RecursiveBlowup4()

class RecursiveBlowup5:
    def __getattr__(self, attr):
        return getattr(self, attr)

def test_getattr():
    return RecursiveBlowup5().attr

class RecursiveBlowup6:
    def __getitem__(self, item):
        return self[item - 2] + self[item - 1]

def test_getitem():
    return RecursiveBlowup6()[5]

def test_recurse():
    return test_recurse()

def test_cpickle(_cache={}):
    try:
        import cPickle
    except ImportError:
        print "cannot import cPickle, skipped!"
        return
    l = None
    for n in itertools.count():
        try:
            l = _cache[n]
            continue  # Already tried and it works, let's save some time
        except KeyError:
            for i in range(100):
                l = [l]
        cPickle.dumps(l, protocol=-1)
        _cache[n] = l

def check_limit(n, test_func_name):
    sys.setrecursionlimit(n)
    if test_func_name.startswith("test_"):
        print test_func_name[5:]
    else:
        print test_func_name
    test_func = globals()[test_func_name]
    try:
        test_func()
    # AttributeError can be raised because of the way e.g. PyDict_GetItem()
    # silences all exceptions and returns NULL, which is usually interpreted
    # as "missing attribute".
    except (RuntimeError, AttributeError):
        pass
    else:
        print "Yikes!"

limit = 1000
while 1:
    check_limit(limit, "test_recurse")
    # check_limit(limit, "test_add")
    # check_limit(limit, "test_repr")
    # check_limit(limit, "test_init")
    # check_limit(limit, "test_getattr")
    # check_limit(limit, "test_getitem")
    # check_limit(limit, "test_cpickle")
    print "Limit of %d is fine" % limit
    limit = limit + 100