We asked a problem about an infinitely nested dictionary in Python. Where to dig to give an answer?

Question

Python

The other day I dealt with the topic "Data passing by value and by reference" and I was given a task.
Gave 3 commands.
1. a = {'k':'v'}
2. a['z'] = a
3. a['z']['k'] = 1
Do them in order.
And they gave 3 questions:
1. What are we working with in line 2?
2. Why doesn't memory grow in line 2?
3. Why will it grow to 3?
I understood the answer to the first question - we are working with a reference data type.
I see that there is a recursion. Interestingly)
But the question of memory is not at all clear to me.
What should I watch or read to get the gist and be able to respond?
Thanks a lot!

Answer 1

If the wording of the problem was given to you exactly like this, then the person who asked it probably expected an answer that in the third paragraph in memory a place for the number 1 would be allocated, and in the second paragraph in the dictionary a link to the memory block already allocated in the first paragraph would simply be stored. But this is not entirely true. First, as Yuriy already wrote , memory for collections is allocated in blocks. Second, for each element of the dictionary, a PyDictEntry structure is created, storing the hash integer and two references to the PyObject structure representing the key and value of the element. This means that even storing a reference to an already existing object will consume 94 bytes (on a 64-bit platform).

Answer 2

You can also remember that the integers 0-255 are created in advance and reused, i.e. in the 3rd line, a reference to an existing object representing the number 1 will be passed.