Tkinter question: master?

[chajadan]

Thanks :-D

--chajadan

At 11:08 AM 3/8/02 +0000, Eric Brunel wrote:
><posted & mailed>
>
>chajadan wrote:
>[snip]
> > class Application(Frame):
> >
> > def __init__(self, master):
> > Frame.__init__(self, master)
>[snip]
> > in "def __init__(self, master):", self refers to this class, i.e.
> > Application (though how those are inherently linked [by name or arguement
> > position--assuming position] I have no clue. master will be set equal to
> > whatever the next arguement is, which seems to be Frame...
>
>Nope: "self" does *not* refer to the class, neither do "master" refer to
>Frame. "self" refers to the instance being built. There's no link at all
>between the arguments in the __init__ method and what you specified in the
>class declaration (Application(Frame)), which describes the inherited
>classes.
>
>Consider this far simpler example:
>
>class C1:
>   def __init__(self, a, b):
>     self.x = a + b
>
>class C2(C1):
>   def __init__(self, a, b, c):
>     C1.__init__(self, a, b)
>     self.y = c
>
>C1 does not inherit from anything, so you define it like above: "class C1"
>with nothing behind. The method __init__ is defined with two parameters,
>that will be passed as arguments when the object is built. So if you have
>to build an instance of C1, you'll have to do that:
>
>o1 = C1(2, 4)
>
>(Try to pass any other number of arguments, and you'll see it fails...)
>
>Now C2 inherits from C1, so its definition is written "class C2(C1)". But
>its __init__ method takes 3 parameters, so when you create an instance of
>C2, you'll write:
>
>o2 = C2(1, 3, 5)
>
>I see what you find confusing: in the class definition C2(C1) means
>"inheritance", but in object creation, C2(1, 3, 5) means "call to the
>__init__ method". The notations are similar, but the meaning are different.
>Which is indeed even more confusing when you inherit from one class and the
>__init__ method has one argument, like in your example.
>
> > the (Frame) in the class definition must just be for
> > inheritance purposes and have ~NOTHING~ to do with arguements
>
>You're right on that.
>
>HTH
>  - eric -
>
>--
>http://mail.python.org/mailman/listinfo/python-list