Improve search cache R/W speeds #4
Description
As it stands, the current search cache implementation is still way too slow. This is probably due to a mix of the code being inefficient in some places, and fread/fwrite being unsuitable candidates for this type of work. (However, I failed to find any other suitable working file IO APIs built into nxdk, meaning it's pretty much restricted to these for the time being)
This issue will be updated as soon as new replacements and implementations are found.