关于python:如何读取包含默认参数值的函数签名?

How can I read a function's signature including default argument values?

给定一个函数对象,我该如何获得它的签名? 例如,对于:

1
2
def myMethod(firt, second, third='something'):
    pass

我想得到"myMethod(firt, second, third='something')"


1
2
3
4
5
6
7
import inspect

def foo(a, b, x='blah'):
    pass

print(inspect.getargspec(foo))
# ArgSpec(args=['a', 'b', 'x'], varargs=None, keywords=None, defaults=('blah',))

但请注意,自Python 3.0以来,不推荐使用inspect.getargspec()

Python 3.0--3.4推荐inspect.getfullargspec()

Python 3.5+建议使用inspect.signature()


可以说,找到函数签名的最简单方法是help(function)

1
2
3
4
5
>>> def function(arg1, arg2="foo", *args, **kwargs): pass
>>> help(function)
Help on function function in module __main__:

function(arg1, arg2='foo', *args, **kwargs)

此外,在Python 3中,一个名为signatureinspect模块中添加了一个方法,该方法用于表示可调用对象的签名及其返回注释:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
>>> from inspect import signature
>>> def foo(a, *, b:int, **kwargs):
...     pass

>>> sig = signature(foo)

>>> str(sig)
'(a, *, b:int, **kwargs)'

>>> str(sig.parameters['b'])
'b:int'

>>> sig.parameters['b'].annotation
<class 'int'>


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
#! /usr/bin/env python

import inspect
from collections import namedtuple

DefaultArgSpec = namedtuple('DefaultArgSpec', 'has_default default_value')

def _get_default_arg(args, defaults, arg_index):
   """ Method that determines if an argument has default value or not,
    and if yes what is the default value for the argument

    :param args: array of arguments, eg: ['first_arg', 'second_arg', 'third_arg']
    :param defaults: array of default values, eg: (42, 'something')
    :param arg_index: index of the argument in the argument array for which,
    this function checks if a default value exists or not. And if default value
    exists it would return the default value. Example argument: 1
    :return: Tuple of whether there is a default or not, and if yes the default
    value, eg: for index 2 i.e. for"second_arg" this function returns (True, 42)
   """

    if not defaults:
        return DefaultArgSpec(False, None)

    args_with_no_defaults = len(args) - len(defaults)

    if arg_index < args_with_no_defaults:
        return DefaultArgSpec(False, None)
    else:
        value = defaults[arg_index - args_with_no_defaults]
        if (type(value) is str):
            value = '"%s"' % value
        return DefaultArgSpec(True, value)

def get_method_sig(method):
   """ Given a function, it returns a string that pretty much looks how the
    function signature would be written in python.

    :param method: a python method
    :return: A string similar describing the pythong method signature.
    eg:"my_method(first_argArg, second_arg=42, third_arg='something')"
   """


    # The return value of ArgSpec is a bit weird, as the list of arguments and
    # list of defaults are returned in separate array.
    # eg: ArgSpec(args=['first_arg', 'second_arg', 'third_arg'],
    # varargs=None, keywords=None, defaults=(42, 'something'))
    argspec = inspect.getargspec(method)
    arg_index=0
    args = []

    # Use the args and defaults array returned by argspec and find out
    # which arguments has default
    for arg in argspec.args:
        default_arg = _get_default_arg(argspec.args, argspec.defaults, arg_index)
        if default_arg.has_default:
            args.append("%s=%s" % (arg, default_arg.default_value))
        else:
            args.append(arg)
        arg_index += 1
    return"%s(%s)" % (method.__name__,",".join(args))


if __name__ == '__main__':
    def my_method(first_arg, second_arg=42, third_arg='something'):
        pass

    print get_method_sig(my_method)
    # my_method(first_argArg, second_arg=42, third_arg="something")


尝试在对象上调用help来查找它。

1
2
3
4
5
6
>>> foo = [1, 2, 3]
>>> help(foo.append)
Help on built-in function append:

append(...)
    L.append(object) -- append object to end

也许派对有点晚了,但是如果你还想保持参数的顺序和它们的默认值,那么你可以使用抽象语法树模块(ast)。

这是一个概念证明(注意对参数进行排序并将它们与默认值匹配的代码肯定可以改进/更清晰):

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
import ast

for class_ in [c for c in module.body if isinstance(c, ast.ClassDef)]:
    for method in [m for m in class_.body if isinstance(m, ast.FunctionDef)]:
        args = []
        if method.args.args:
            [args.append([a.col_offset, a.id]) for a in method.args.args]
        if method.args.defaults:
            [args.append([a.col_offset, '=' + a.id]) for a in method.args.defaults]
        sorted_args = sorted(args)
        for i, p in enumerate(sorted_args):
            if p[1].startswith('='):
                sorted_args[i-1][1] += p[1]
        sorted_args = [k[1] for k in sorted_args if not k[1].startswith('=')]

        if method.args.vararg:
            sorted_args.append('*' + method.args.vararg)
        if method.args.kwarg:
            sorted_args.append('**' + method.args.kwarg)

        signature = '(' + ', '.join(sorted_args) + ')'

        print method.name + signature


如果你要做的只是打印函数,那么使用pydoc。

1
2
3
4
5
6
7
8
import pydoc    

def foo(arg1, arg2, *args, **kwargs):                                                                    
    '''Some foo fn'''                                                                                    
    pass                                                                                                

>>> print pydoc.render_doc(foo).splitlines()[2]
foo(arg1, arg2, *args, **kwargs)

如果您正在尝试实际分析函数签名,请使用检查模块的argspec。在将用户的钩子脚本函数验证到一般框架中时,我必须这样做。


示例代码:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
import inspect
from collections import OrderedDict


def get_signature(fn):
    params = inspect.signature(fn).parameters
    args = []
    kwargs = OrderedDict()
    for p in params.values():
        if p.default is p.empty:
            args.append(p.name)
        else:
            kwargs[p.name] = p.default
    return args, kwargs


def test_sig():
    def fn(a, b, c, d=3, e="abc"):
        pass

    assert get_signature(fn) == (
        ["a","b","c"], OrderedDict([("d", 3), ("e","abc")])
    )

在命令行(IPython)中使用%pdef,它将仅打印签名。

例如%pdef np.loadtxt

1
 np.loadtxt(fname, dtype=<class 'float'>, comments='#', delimiter=None, converters=None, skiprows=0, usecols=None, unpack=False, ndmin=0, encoding='bytes')