For loop overriding outer variables instead of creating new ones

Question

I stumbled upon another basic concept I've missed in Python:

Having this basic for (foreach) loop:

x = 15
for x in range(10):
    continue
print(x)

The value for x I expected was 15, but instead I got 9.

The same code snippet in C returns x's original value – 15:

#include <stdio.h>

int main(void) 
{
  int x = 15;
  for (int x = 0; x < 10; x++)
  {
    continue;
  }
  printf("%d", x);
  return 0;
}

I can't figure out how the variable scope works here.

Since x is declared outside the for loop scope, shouldn't a new local variable be created during the lifetime of the loop?

Why is x being overridden in the Python version then?

I don't know about c, but in python, any variable will always take the value it's assigned — Nathan, Nov 17 '18 at 23:29
try adding a print statement `print(x)` before continue to see what happens. — Nathan, Nov 17 '18 at 23:29
@Nathan adding the print statement would print each value of x, from 0 to 9, but I can't understand why does it override x instead of creating a new local variable, like for example C does. — Kamen Hristov, Nov 17 '18 at 23:32
A `for` loop doen't create a new scope in Python. For scoping rules, see for example https://stackoverflow.com/questions/291978/short-description-of-the-scoping-rules — Thierry Lathuille, Nov 17 '18 at 23:33
@ThierryLathuille Thanks! Great post, I should've found it earlier! — Kamen Hristov, Nov 17 '18 at 23:40

Matthieu Brucher · Answer 1 · 2018-11-17T23:31:26.957

3

This is not the same. In C, you explicitly create a new variable, whereas in Python, you reuse the name in the for scope, ending up overriding the previous value.

So the C equivalent really is:

#include <stdio.h>

int main(void) 
{
  int x = 15;
  for (x = 0; x < 10; ++x)
  {
    continue;
  }
  --x; // To accommodate the different behavior of the range loop
  printf("%d", x);
  return 0;
}

Don't forget that in Python, variables are just entries in a dictionary, dynamically created, whereas in C, they are independent, static items.

edited Nov 17 '18 at 23:31

answered Nov 17 '18 at 23:30

Matthieu Brucher

21,634
7
38
62

`x < 9` actually – Marco Bonelli Nov 17 '18 at 23:31
it is x<10, but there should not be a final increment. In a way, it's closer to the C++ container loop than any C for loop. – Matthieu Brucher Nov 17 '18 at 23:33
"it is `x < 10` but there should not be a final increment"... yeah, so it is `x < 9`... lol – Marco Bonelli Nov 17 '18 at 23:34
2

No, because there are 10 iterations in the loop. If we were doing something inside the loop instead of doing continue, we would see 10 passes, not 9. – Matthieu Brucher Nov 17 '18 at 23:35
Yeah ant if there was a `return 0` inside the loop then no iterations would have happened, you can imagine what you want but what OP wrote was an empty loop. In fact, GCC even optimizes that program to `printf("%d", 9);` – Marco Bonelli Nov 17 '18 at 23:39
Anyway yah I'm just nitpicking, it makes sense to explain it either way – Marco Bonelli Nov 17 '18 at 23:41
Yes indeed, it's an important difference. – Matthieu Brucher Nov 17 '18 at 23:53

For loop overriding outer variables instead of creating new ones

1 Answers1

Linked