introspection and functions

[yagyala]

Hi. I would like to be able to tell, at run time, how many parameters
a function requires. Ideally I would like to be able to tell which are
optional as well. I've tried looking at the functions attributes, but
haven't found one that helps in this. How can I do this?

Thanks

[Wildemar Wildenburger]

yagyala wrote:

Hi. I would like to be able to tell, at run time, how many parameters
a function requires. Ideally I would like to be able to tell which are
optional as well. I've tried looking at the functions attributes, but
haven't found one that helps in this. How can I do this?

I've never used it before, but there is the "inspect" module.
<url:http://docs.python.org/lib/module-inspect.html>.
That any good?

/W

[James Stroud]

yagyala wrote:

Hi. I would like to be able to tell, at run time, how many parameters
a function requires. Ideally I would like to be able to tell which are
optional as well. I've tried looking at the functions attributes, but
haven't found one that helps in this. How can I do this?

Thanks

pydef doit(a, b, c, x=14):
.... pass
....
pydoit.func_code.co_argcount
4
pydoit.func_code.co_varnames
('a', 'b', 'c', 'x')
pydoit.func_defaults
(14,)

James

[Bruno Desthuilliers]

yagyala a écrit :

Hi. I would like to be able to tell, at run time, how many parameters
a function requires. Ideally I would like to be able to tell which are
optional as well. I've tried looking at the functions attributes, but
haven't found one that helps in this. How can I do this?

>>def myfunc(arg1, arg2, arg3='toto', *args, **kw): pass

....

>>import inspect
inspect.getargspec(myfunc)

(['arg1', 'arg2', 'arg3'], 'args', 'kw', ('toto',))
HTH

[Scott David Daniels]

yagyala wrote:

Hi. I would like to be able to tell, at run time, how many parameters
a function requires. Ideally I would like to be able to tell which are
optional as well. I've tried looking at the functions attributes, but
haven't found one that helps in this. How can I do this?

Thanks

This really only will work for those functions that are simply
constructed. You are better off not trying to do this, or code
like the following will confound your code:

from functools import partial

def somefunction(a=23, b=14, c='hi'):
print 'Finally'

f = partial(somefunction, b=13)
g = partial(f, a=19)
h = partial(g, c=123)
print whatargs(h)

[Ayaz Ahmed Khan]

"James Stroud" typed:

pydef doit(a, b, c, x=14):
... pass
...
pydoit.func_code.co_argcount
4
pydoit.func_code.co_varnames
('a', 'b', 'c', 'x')
pydoit.func_defaults
(14,)

Neat.

--
Ayaz Ahmed Khan

I have not yet begun to byte!

[=?ISO-8859-1?Q?Ricardo_Ar=E1oz?=]

Ayaz Ahmed Khan wrote:

"James Stroud" typed:

>pydef doit(a, b, c, x=14):
... pass
...
pydoit.func_code.co_argcount
4
pydoit.func_code.co_varnames
('a', 'b', 'c', 'x')
pydoit.func_defaults
(14,)

Neat.

How do you know the 14 corresponds to x ?

[Wildemar Wildenburger]

Ricardo Aráoz wrote:

Ayaz Ahmed Khan wrote:

>"James Stroud" typed:

>>pydef doit(a, b, c, x=14):
... pass
...
pydoit.func_code.co_argcount
4
pydoit.func_code.co_varnames
('a', 'b', 'c', 'x')
pydoit.func_defaults
(14,)

Neat.

How do you know the 14 corresponds to x ?

Well, there is one optional argument (len(doit.func_defaults)==1) and
those *must* come after mendatory arguments. More generally: If a
function has n optional args, they're the n last ones. There is never
any ambiguity.

/W