Add reverse inplace function for digraph

[matanco64]

I added a reverse function for a graph, it takes all the edges and reverses them,

also added tests and documentation for the function

[CLAassistant]

All committers have signed the CLA.

[coveralls]

Pull Request Test Coverage Report for Build 4924945942

• 9 of 9 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.009%) to 97.093%

---

Totals
Change from base Build 4919724815:	0.009%
Covered Lines:	14397
Relevant Lines:	14828

---

💛 - Coveralls

[mtreinish]

Overall this looks great. I think your approach here makes the most sense to do an inplace edge reversal. I left a couple of small inline comments, the thing blocking CI I expect is my comment about the docs. The other thing which is missing is a release note to document the addition of the new feature. If you don't mind could you add one, the procedure for doing this is documented here: https://github.com/Qiskit/rustworkx/blob/main/CONTRIBUTING.md#release-notes

[mtreinish]

Two small test cases that I think would be good to add is one with an empty graph. The other is for a graph where there is an edge removed in the middle (which will leave a hole in the edge indices). I don't expect there are any problems with the method's behavior with these, but they're common edge cases that are good to verify (because a lot of other methods have had issues in the past with both).

[matanco64]

Hey Sorry for the late response and update, I've added the tests

[IvanIsCoding]

Thanks for contributing this. Overall LGTM, I just have one suggestion. Also, I second Matthew's comment that we should add some extra tests

My main suggestion is naming the method reverse because NetworkX does it that way. It is not a rule in the project but it does avoid people asking what the equivalent method is. For now, we can ignore the copy keyword and just always do it in place.

[matanco64]

I've gone over all the comments

1. renamed
2. added tests
3. changed to unwraps

I also added code to the interface at digraph.pyi so it will be easier to develop in pycharm and such

[IvanIsCoding]

LGTM, you just need to run cargo fmt because you forgot but it should be good to go

[matanco64]

Thank You! ran cargo fmt, should be good now :)

[IvanIsCoding]

You’ll need to fix your docstring too, it’s giving Inline literal start-string without end-string. when building the docs

[mtreinish]

I pushed up a quick fix to update the docs syntax here to get this merged as it LGTM besides the docs failure. The underlying issue was the docstring for rustworkx. The docstrings get exported to python as rst that sphinx will need to read but the docs were constructed using rust doc so I just converted that to use python/sphinx rst.