the right way of handling **args for __init__ and other methods

[Bjorn Pettersen]

Oivvio Polite wrote:
> 
> Hi all.
> I guess this is more of a design question than a howto question.
> 
> It has to do with implementing arbitrary-argument set methods.
> I want the user to be able to use the same interface with __init__ and read.
> This is a stripped down version:
> class argeater:
>     def __init__(self, somearg, **args):
>         print "init ",args
>         self.data = {}
>         self.read(somearg, argdict=args)

you can use apply(self.read, (), args) and avoid the special case in
read, or...

>     def read(self, somearg, argdict=None,**args):
>         print "read ",args
>         print "argdict ",argdict
> 
>         #putting everything from argdict into args
>         #feels a bit awkward.
>         if argdict:
>             assert type(argdict) == types.DictionaryType
>             for key in argdict.keys():
>                 if not args.has_key(key):
>                     args[key] = argdict[key]
> 
>         #now do a lot of stuff to self.data
>         #based on contents of args

Since args looks to be always empty when argdict != None, you can say

  args.update(argdict)

which will overwrite args, or if you want to be able to pass both a
dictionary and keyword args to read:

  tmp = {}
  tmp.update(args)
  tmp.update(argdict)

-- bjorn