Python's int function performance

Question

Does Python's built-in function int still try to convert the submitted value even if the value is already an integer?

More concisely: is there any performance difference between int('42') and int(42) caused by conversion algorithm?

Answer 1

This is handled in function long_long in Objects/longobject.c, as explained in more detail by thefourtheye:

static PyObject *
long_long(PyObject *v)
{
    if (PyLong_CheckExact(v))
        Py_INCREF(v);
    else
        v = _PyLong_Copy((PyLongObject *)v);
    return v;
}

So, when the argument is already an int, the reference count is incremented and the same object returned.

You can assume similar behavior for immutable types in general,. For example, tuple(mytuple) returns a new reference to mytuple, while, by contrast, list(mylist) creates a copy of mylist.

Answer 2

If you pass an int object to int(), you get the same object back (CPython 3.3.2):

>>> a = 1000 * 1000 # large enough to avoid interning
>>> b = int(a)
>>> a is b
True

I don't know what you mean by "algorithmic performance difference", but it doesn't create a new object.