Hatchery: Faster sorting algorithm

[Xwaler]

Currently each loop of the Hatchery Automation sorts the whole pokémon list only to use at most the first 4 to fill the hatchery queue, this results in a significant overhead and some visible stuttering with large parties. On my 1100% save the automation used 99ms every second.

This PR proposes to use partial sorting to only compute the sorted top 4 pokémons, avoiding spending a lot of time sorting the 1300+ other.

Only caveat of this is that the pokerus spreading priority currently requires the whole list to be sorted to pick relevant uninfected/contagious pokémons depending on the current hatchery state. For instance the first contagious pokémon with the needed types to spread might be at the very end of the sorted list for all I know.
Therefore, when this priority is enabled (and some pokémons are uninfected), the previous complete sort is done instead to preserve functionnality at the cost of performance. This seems like a reasonable compromise considering it's very quick to be done spreading the pokerus.

The results show that with the attribute sort set to Breeding Efficiency (the most expensive), the loop goes from 99ms to 19ms. Stutters are also gone.

Before

After

[Farigh]

Hi,

Sorry for the late review, I couldn't find time sooner 😞

That's an awesome improvement, I did not think about this optimalisation when I tried to optimise a bit the scripts to run on my Py4. Good job !
If you figure-out a way to include the pokérus filter at some point, don't hesitate to submit a new PR 😉

Side note : I fixed a few style issues, and added a bit more comments, please try to fit the project's style and documentation for futur work.

Thanks again 🚀

[Farigh]

I also renamed the partialSort Utils method, since it did not do a partial sort per say, which would still return a list with all the elements (and have only the first part being sorted).
I chose the getSortedSubRange name, as it seems more appropriate.

Additionally, I optimised a bit the bisect method by returning directly if the exact match was found.

Finally, I fixed a risk of overflow. It could not happen in the hatchery case (since the user would definitely have at least 4 pokémons at this point), but since Utils methods are meant to be reused in any other case of the project, I prefer to be exhaustive.

[Xwaler]

Thank you for the improvements !

As for the style issues, my apologies I completly forgot to refactor to the (former) partialSort function, kind of written it to benchmark and then left it there haha. I'll do my best to better fit the project's guidelines from now on 😉