Rewrite GiveBoxMonInitialMoveset_Fast to match GiveBoxMonInitialMoveset results (but still faster)

[Sneed69]

Issue(s) that this PR fixes

The GiveBoxMonInitialMoveset_Fast function can generate pokemon that know the same move twice, if a pokemon has that move twice in its last 4 learnset slots.
It also produces different results from GiveBoxMonInitialMoveset in some niche cases and reverses the order of known moves.
Currently only used to fill the PC storage system with pokemon by the debug menu.

Discord contact info

duke5614

[mrgriffin]

Hmm... Based on the conversation yesterday, I want to say that this tweak to the going-backwards algorithm still doesn't always produce the same results as going-forwards?

Because what we need to say is "if the move is in any of the previous four learnset slots, skip it", not "if the move is already known, skip it"?

(Also the order of the moves probably comes out different, but I don't think I personally consider that to be a problem unless/until GF introduces a mechanic which cares about the specific order.)

EDIT: But obviously this is a fix to the "can have duplicates" bug, as you describe :)

[Sneed69]

Hmm... Based on the conversation yesterday, I want to say that this tweak to the going-backwards algorithm still doesn't always produce the same results as going-forwards?

Because what we need to say is "if the move is in any of the previous four learnset slots, skip it", not "if the move is already known, skip it"?

(Also the order of the moves probably comes out different, but I don't think I personally consider that to be a problem unless/until GF introduces a mechanic which cares about the specific order.)

EDIT: But obviously this is a fix to the "can have duplicates" bug, as you describe :)

Updated with fixes to these concerns.