Reverse argument order
bokr at accessone.com
Mon Oct 22 22:17:07 CEST 2001
On Sat, 20 Oct 2001 23:16:54 GMT, slinkp23 at yahoo.com (Paul Winkler) wrote:
>On Sat, 20 Oct 2001 21:53:52 GMT, Bengt Richter <bokr at accessone.com> wrote:
>>Brute force, but clear:
>> >>> def wait(first=None,second=None,third=None):
>> ... if third: h,m,s = first,second,third
>> ... elif second: h,m,s=0,first,second
>> ... elif first: h,m,s=0,0,first
>> ... else: h,m,s=0,0,0
>> ... print "h=%d, m=%d, s=%d" % (h,m,s) # or whatever you want to do
>> >>> wait()
>> h=0, m=0, s=0
>> >>> wait(1)
>> h=0, m=0, s=1
>> >>> wait(1,2)
>> h=0, m=1, s=2
>> >>> wait(1,2,3)
>> h=1, m=2, s=3
>This is broken!
>h=1, m=2, s=3
>h=0, m=0, s=1
>Clearly the user intended to get h=1, m=0, s=0.
>Fix: Change it so your tests look like "if third is not None:"
>... etc. Then it works:
>h=0, m=1, s=0
>h=1, m=0, s=0
>h=1, m=0, s=5
>But it's still not great, because the function def implies that this
>function can take keyword args when in fact it can't:
>Traceback (most recent call last):
> File "<stdin>", line 1, in ?
> File "<stdin>", line 6, in wait
>TypeError: an integer is required
>That could confuse a user who actually bothered to learn python.
>I think Kirill's solution (among others) is better; it does the right
Me too, as I noted when I saw it.
How do you retract a[n embarrassing] post?
>thing with inappropriate keyword args:
>TypeError: wait() got an unexpected keyword argument 'foo'
Summary: A clear case of PIHCAL (post in haste, cringe at leisure) ;-/
More information about the Python-list