(analyzing the japanese kanken test 3ds game text file binary headers more)

well i managed to determine the correct byte length for each record and figured out some operations to de-shift the table so i have ordered records now, and there's a fair bit of regularity, for example you see some counting up likely indicating a question index or numbering, however im failing to find meaning in the other pointers/offets

some theories i've grappled with are that two of those pointers would be ranges (question beginning question end) but there seems to be no regularity when the left value is larger than the right value, i would assume depending on the question type different byte values are filled but i don't see a rule here, only that after a dozen questions there are sets of 5 with those middle bytes filled

also i would guess these should be 4 byte offsets but have blocks where those start with a 01 byte so idk how plausible that is, searching in the file with 2 byte offsets didn't return anything that was consistently meaningful though

as this is the kanken level 10 data i've thought about how many question types there are and from the data it looks like there are 40 antonym + 10 synonym questions but honestly i don't see a 40/10 grouping that is uniquely identifiable to me

what i assume the initial FFFFFFFF to be is a reference to an external file cuz some questions have all their data in the file while other clearly reference externally (although it might also be internally thinking about it) well i have no clue

probably about to give up, it could be that the game does weird storage logic that makes working with these variables not easily feasible which would suck