Codecov Report

Merging #698 into master will decrease coverage by 0.02%.
The diff coverage is 47.05%.

@@            Coverage Diff            @@
##           master    #698      +/-   ##
=========================================
- Coverage   48.73%   48.7%   -0.03%     
=========================================
  Files          81      81              
  Lines       15509   15561      +52     
  Branches     3797    3806       +9     
=========================================
+ Hits         7558    7579      +21     
- Misses       6153    6181      +28     
- Partials     1798    1801       +3

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5e74d53...2c8657c. Read the comment docs.

How does this work with staticmethod in comparison to CPython?

In CPython, super() doesn't work in a staticmethod anyway, does it?

I don't think this works. If you get class from the first argument then uses of super further up the inheritance chain don't get the right thing and you get an infinite loop.

class A(object):
    def f(self):
        pass

class B(A):
    def f(self):
        super().f()

class C(B):
    def f(self):
        super().f()

C().f()

Running `target/debug/rustpython test.py`

thread 'main' has overflowed its stack
fatal runtime error: stack overflow

The cell var isn't too hard to add. It's easier to do than describe, so #703.

Merged #703, hopefully not too soon, so you can use it here as well.

Thanks @cthulahoops that helped :). I was not sure how to fetch the __class__ so I added a get to Scope. I am not sure if that is the best solution...

Rather than introducing Scope.get, can't you just use load_name from the ScopeProtocol?

If you're worried about the search going into builtins, I'd add a load_cell to the scope protocol to search non-local scope only (with the same shape as load_name).

I Added load_cell.

This looks great! It's excellent to finally have super.

Got some sort of vm mutability issue now (from the frame change?), but other than that I think we're good to go.