[Python-Dev] type categories

[Greg Ewing]

Guido:

> I can see how it could be done using some additional syntax similar to
> what ML uses, e.g.:
> 
>   def f(a: Cat1):
>       ...code for Cat1...
>   else f(a: Cat2):
>       ...code for Cat2...
>   else f(a: Cat3):
>       ...code for Cat3...

As long as all the implementations have to be in one place,
this is equivalent to

  def f(a):
    if belongstocategory(a, Cat1):
      ...
    elif belongstocategory(a, Cat2):
      ...
    elif belongstocategory(a, Cat3):
      ...

so you're not gaining much from the new syntax.

> It might also be possible to modify a multimethod dynamically,
> e.g. later one could write:
> 
>   def f4(a: Cat4):
>     ...code for Cat4...
> 
>   f.add(f4)

This sort of scheme makes me uneasy, because it means that any module
can change the behaviour of any call of f() in any other
module. Currently, if you know the types involved in a method call,
you can fairly easily track down in the source which piece of code
will be called. With this sort of generic function, that will no
longer be possible. It's kind of like an "import *" in reverse -- you
won't know what's coming from where, and you can get things that
you never even asked for.

Greg Ewing, Computer Science Dept, +--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg@cosc.canterbury.ac.nz	   +--------------------------------------+