class or inherited list ?

[Stef Mientki]

thanks Gabriel,

Gabriel Genellina wrote:
> En Fri, 28 Mar 2008 12:15:45 -0300, Stef Mientki <stef.mientki at gmail.com>  
> escribió:
>
>   
>> Passing all kinds of data between objects,
>> I'm looking for an elegant (and simple) way to pack the data.
>> Now it looks to me that both the class and the inherited list,
>> performs equally well.
>> Till now, the most elegant way (in my view) is the list inheritance,
>> mainly because you don't need brackets and no quotes.
>> What are others opinion about this ?
>>
>> class super_list(list):
>>     pass
>>
>> def kwadraat ( value ) :
>>     return value * value
>>
>> x={}
>> x['frequency']=33
>> x['functie']=kwadraat
>> print x['functie'](2)
>>
>> y = super_list()
>> y.frequency = 33
>> y.functie = kwadraat
>> print y.functie(3)
>>     
>
> You don't use y as a list at all - you might as well inherit from object.
Good point, didn't notice that.
>   
> And in that case you get a generic attribute container - as you said, like  
> a dict but using attribute syntax (drawback: only valid names can be used  
> as keys)
>   
Perfect, I don't need invalid names.
> But I don't understand the "functie" usage - what's for? Perhaps if it  
> used y.frequency - in that case a proper method would be better.
>   
In my case the sender of the data determines the function ("functie" is 
the Dutch word for "function"),
assume this concept:
Sender:
- generates a huge array of data
Receiver:
- is a plot utility with interactive user interface
- the user can select a part of the huge dataset and wants to see the 
result of "function" on the selected part of the huge dataset
So by passing the function-name(s), I prevent that the functions(s) 
should be performed on all the data.
while they will never be used.

cheers,
Stef Mientki