Allow range(stop) with default start for override grammar

[victorjoos]

Motivation

To make the range function more like the default python range(), and easier to type, I added a check if only start was added to set stop to start and start to zero.

Have you read the Contributing Guidelines on pull requests?

Yes

Test Plan

tested by adding a parametrize argument in test_basic_sweeper.py.

Related Issues and PRs

N/A

[Jasha10]

Thanks for the PR!

It would be good to have a few more tests -- see the tests/test_overrides_parser.py file, where most of the range tests currently reside.
In particular, I think it would be good to pin down behavior for the following:

• behavior when some inputs are zero or negative, e.g. range(-10, step=1), range(10, step=-1), range(0), ...
  for some of these cases, it is not immediately obvious to me what the behavior should be.
• behavior when called on float inputs

[victorjoos]

Aaah, Thanks for the tips !

I was a bit fast in implementing this, didn't notice I had to also modify the parser. Let me modify the right files.

Not sure about the float inputs, but for the other cases I would say the best is to compare to the default python range implementation:

• range(-10) == []
• range(-2, step=-1) == [0, -1]
• range(0) == []
• range(10, step=-1) == []

[victorjoos]

Actually didn't need to change the parser. I added some tests and a new example in the docs.

[jieru-hu]

one nit, otherwise lgtm. Thanks!

[jieru-hu]

@omry could you take a look as well?

[omry]

Of course. but as I am in the middle of releasing 1.1 this will wait.

[omry]

Thanks. Looking pretty good.
See comments.

[victorjoos]

Thank you, sorry for the delay. I implemented your suggestions.

[omry]

Looking good, thanks.

[omry]

We have some existing CI failures we are looking at. Will likely need you to rebase once those are fixed to get CI green again.

[jieru-hu]

@victorjoos could you rebase master and force push to this branch? Many thanks!