"Gre7g Luterman" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED] |I suppose I was lulled into complacency by how Python makes so many things | look like classes, but I'm starting to realize that they're not, are they? | | I'm writing a C program which handles Python objects in different ways based | on their type. I do a PyInstance_Check(PyObj) to determine if the PyObj is | an instance, but it is returning 0 on a lot of stuff that I thought would be | an instance. So I did the following simple test on three things that look | like instances: | | Python 2.5.1 (r251:54863, Apr 18 2007, 08:51:08) [MSC v.1310 32 bit (Intel)] | on win32 | Type "help", "copyright", "credits" or "license" for more information. | >>> class a: pass | ... | >>> type(a()) | <type 'instance'> | >>> type(Exception()) | <type 'exceptions.Exception'> | >>> class b(dict): pass | ... | >>> type(b()) | <class '__main__.b'> | | I was relieved that a() returns an instance, but I was surprised that | Exceptions aren't really instances at all. And what's the deal with derving | a class from a standard type like a dictionary? I thought for sure, that | would be an instance, but this shows it is a class?!? | | Can anyone explain the last one and/or give me a simple test I can do in C | to determine whether a Python object is "instance-like"?
Your problem is mixing two different meanings of 'instance', one obsolete. Everything is an instance of its type/class. However, for old-style classes, all instances of user defined classes are instances of type 'instance' instead of their class. This is usually not very helpful. New style classes fix this. >>> class a(object): pass >>> type(a()) <class '__main__.a'> In 2.5, Exception changed from old-style to new. Terry Jan Reedy -- http://mail.python.org/mailman/listinfo/python-list
