Recurring identity collisions on python objects

Question

I understand that the built-in python id() returns an identifier unique to the life of the object. I understand that objects with non-overlapping lifetimes can have the same identifier. However, I am trying to understand this rather confusing behavior:

 >>> id(matplotlib.image.BboxImage.set_cmap) 4424372944 >>> id(numpy.ma.core.MaskedArray.sum) 4424372944 

And in fact, in multiple instances of the interpreter, the behavior is repeatable:

 Mac:~$ python2.7 -c "import matplotlib.image; import numpy; print id(matplotlib.image.BboxImage.set_cmap), id(numpy.ma.core.MaskedArray.sum)" 4343186208 4343186208 Mac:~$ python2.7 -c "import matplotlib.image; import numpy; print id(matplotlib.image.BboxImage.set_cmap), id(numpy.ma.core.MaskedArray.sum)" 4521153312 4521153312 Mac:~$ python2.7 -c "import matplotlib.image; import numpy; print id(matplotlib.image.BboxImage.set_cmap), id(numpy.ma.core.MaskedArray.sum)" 4358591264 4358591264 Mac:~$ python2.7 -c "import matplotlib.image; import numpy; print id(matplotlib.image.BboxImage.set_cmap), id(numpy.ma.core.MaskedArray.sum)" 4389970720 4389970720 

It seems that matplotlib.image.BboxImage.set_cmap and numpy.ma.core.MaskedArray.sum always get the same identifier from each other, even with different installations of the python interpreter.

Now I understand that this is consistent with the docs for id() , as these two objects are created dynamically upon access and, therefore, will have non-overlapping lives. But why do these two unrelated objects always end with the same identifier?

(This question is different from Object methods of the same class have the same identifier? Because here I ask why this is repeated in multiple instances of the interpreter, and not just a false id collision with objects with non-overlapping lifetimes.)

Answer 1

This will be true for any instance method that is a type of both objects that you are testing. Each time you access the descriptor, a new instancemethod object is created. Since you do not save the reference to the newly created method, it has the right to garbage collection as soon as the id returned, and, apparently, it is collected before the next id call, so the same memory location can be reused.

This behavior can be seen in a different way:

 >>> x = matplotlib.image.BboxImage.set_cmap >>> y = matplotlib.image.BboxImage.set_cmap >>> id(x) == id(y) False 

x and y are references to two different objects, because the set_cmap binding creates a new instancemethod object each time.