Tracking object allocation in python -

is there method can override allow me use print statements / pdb / etc. keep track of every time instance of class allocated? while unpickling objects seeming never have either __setstate__ or __init__ called on them. tried overriding __new__ , printing out id of every object make in __new__, still encountering objects ids never printed.

edit: here code use altering (instrumenting) __new__ of class , of super-classes except object itself:

class allocator:     def __init__(self, my_class):        self.my_class = my_class        self.old_new = my_class.__new__      def new(self, * args, ** kargs):         rval = self.old_new(*args, ** kargs)         #rval = super(self.my_class,cls).__new__(cls)         print 'made '+str(self.my_class)+' id '+str(id(rval))         return rval  def replace_allocator(cls):     if cls == object:         return      setattr(cls,'__new__',allocator(cls).new)     print cls.__base__      try:         parent in cls.__base__:             replace_allocator(parent)    except:         replace_allocator(cls.__base__) 

i call replace_allocator on classes' parent class imported in main script. class has custom __new__ begin with, prints out id.

(this more of comment answer.)

quoting guido's unifying types , classes in python 2.2:

there situations new instance created without calling __init__ (for example when instance loaded pickle). there no way create new instance without calling __new__ (although in cases can away calling base class's __new__).

if using new-style classes (descendants of object), __new__() should called. don't think obscure cases "you can away calling base class's __new__" in happen accidently, though don't know these cases are.

and add example:

in [1]: class a(object):    ...:     def __new__(cls):        ...:         print "a"    ...:         return object.__new__(cls)    ...:       in [2]: a() out[2]: <__main__.a object @ 0xa3a95cc>  in [4]: object.__new__(a) out[4]: <__main__.a object @ 0xa3a974c>