Fix Prim's algorithm implementation and update tests

[Alokzh]

Fixes: #28

Description:

This PR fixes the implementation of Prim's algorithm to align with the API conventions.

• Removed explicit return in the run method.
• Added reconstructMST method to retrieve the edges.
• Updated tests to use reconstructMST instead of relying on the return value of run.

[Ducasse]

Thanks for the PR. What is reconstructMST?

[Alokzh]

Hi @Ducasse, I think I made a mistake in naming the method here, In prim's algo MST edges are already computed and stored, so there’s no need to "reconstruct" them.

Do you think I should name it as something like getMSTEdges ( Here MST stands for Minimum Spanning Tree ) or getTreeEdges.
Please let me know , which is better or if you have some other suggestions.

[Ducasse]

I would call them minimumSpanningTreeEdges (get feels like a Java hack).

Now I do not get the name of the variable because does it returns a tree or edges?

prim run.
tree := prim reconstructMST collect: [ :e | e asTuple ].

[Alokzh]

Thnx @Ducasse, I have renamed the variables.

Please have a look whenever you get time.

[jordanmontt]

Nice refactor :)
the name minimumSpanningTreeEdges it is way better.

Thanks for your PR @Alokih

[jordanmontt]

BTW we have a failing CI (it is not your fault). I believe it can be a good exercise for you if you can make a PR to have all the CIs green. Of course, only if it interest you. What do you think @Alokih ?

[Alokzh]

For sure @jordanmontt I am working on it. Thnx

[Alokzh]

@jordanmontt I tested the workflow run on my fork repo & found 2 main issues :

1. There was a mismatch in naming of AINetworkFlowEdge.class.st. Actual file was named AINetworkFLowEdge.class.st in repo ( Note the uppercase L in FLow ).
   So I fixed it & again ran workflow then I got another issue
   Here is the workflow run link in forked repo link

2. CI was failing for Pharo64-9.0 specifically on the Windows runner, the issue was related to VM I think ( sqMakeMemoryNotExecutableFromTo in sqWin32SpurAlloc.c ) You could also check from above workflow link I provided.
   So what I did was removed Pharo64-9.0 from jobs & ran again
   Here is the link
   Please let me know what you think, if Pharo64-9.0 is not crictical we could remove it either only for windows or all OS.
   If it is required can you please give me some suggestions here, I am stuck at this point.

Also I have fixed this warning

[jordanmontt]

Hello @Alokih

1. Yes, you are right about the uppercase L. I don't know how it got there... Somebody must have edited the file manually.
2. Indeed, there is a VM failure for the Pharo 9 on WINDOWS. It is working just fine on my mac. What I just did it is just to remove Pharo 9, if we have users that need it in Pharo 9 we will investigate further.

I fixed those 2 things and know we have a green CI. Thanks for your investigation!!

[Alokzh]

Glad I could help ! @jordanmontt Happy to see CI green now 🎉