HK: fix grubhunt required grubs count

[qwint]

What is this fixing or adding?

somehow this mixup got into the final grubhunt PR

the original code that somehow got mixed up in the final grubhunt PR (in a per player loop, which is why we need the loop in the new code)

if grub.location and grub.location.player in relevant_groups:
    # not counting our grubs stuck in item links because we also will count the group's copy
    pass
else:
    world.grub_count += 1

How was this tested?

gen'd with a breakpoint to confirm world.grub_count with 4 grubhunt: all players using both replacement grubs and null, and confirmed that the number gets honored when not using all still

If this makes graphical changes, please attach screenshots.

[Exempt-Medic]

Given how complicated this seems to be, can you add unit tests for it?

[Exempt-Medic]

I asked in DMs, but I should put this here as well:
Why is the expected grub count 69 when there are 92 Grubs in the pool and playthrough also expects players to get 92 Grubs?

[Exempt-Medic]

Apparently only one of the players is linking the replacement Grub. I hate itemlinks

[Mysteryem]

Yes. There are always 1 fewer replacement items, per linked item, than the number of items being linked together.

If there were 6 players in an item link, linking an item between them would become 1 linked item + 5 replacements.

[Exempt-Medic]

Oh, what I said is also apparently completely wrong. The link_relacement is for the group, the replacement item is per player

[Exempt-Medic]

Tested some configurations of players and saw that the proper values were found. The unit tests pass and the code makes sense after I, regretfully, learned a lot more about itemlinks

[black-sliver]

I trust medic