Don't inherit Shape from NamedTuple

[awygle]

Fixes issue #421 .

[codecov]

Codecov Report

Merging #422 into master will decrease coverage by 0.12%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #422      +/-   ##
==========================================
- Coverage   81.88%   81.75%   -0.13%     
==========================================
  Files          35       35              
  Lines        6094     6079      -15     
  Branches     1236     1238       +2     
==========================================
- Hits         4990     4970      -20     
- Misses        927      931       +4     
- Partials      177      178       +1     

Impacted Files	Coverage Δ
nmigen/hdl/ast.py	88.76% <100.00%> (+0.12%)	⬆️
nmigen/lib/io.py	73.68% <0.00%> (-8.67%)	⬇️
nmigen/back/pysim.py	90.40% <0.00%> (-0.51%)	⬇️
nmigen/build/run.py	31.25% <0.00%> (ø)

---

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 175c8a5...1500165. Read the comment docs.

[awygle]

Because I'm bad at python and didn't know iter() existed, and also I've been writing Rust a lot and this is how you have to do it there.

…

On Mon, Jul 6, 2020, 2:34 AM whitequark ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In nmigen/hdl/ast.py <#422 (comment)>: > @@ -55,8 +71,15 @@ class Shape(typing.NamedTuple): signed : bool If ``False``, the value is unsigned. If ``True``, the value is signed two's complement. """ - width: int = 1 - signed: bool = False + def __init__(self, width=1, signed=False): + if not isinstance(width, int) or width < 0: + raise TypeError("Width must be a non-negative integer, not {!r}" + .format(width)) + self.width = width + self.signed = signed + + def __iter__(self): + return ShapeIter(self.width, self.signed) Why not return iter((self.width, self.signed))? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#422 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB35VVT5TIKM7BOMG36KKNTR2GLBDANCNFSM4ORHGAXQ> .

[whitequark]

Thanks!