This topic was archived here:

https://web.archive.org/web/20230000000000fw_/https://www.zenhax.com/viewtopic.php?t=11410

Velosity, posted Sun May 12, 2019 5:41 pm (47736)

---

Samples: https://mega.nz/#!dsMDiSzC!lrpG95ivr6ui ... UsX_-ZlhDI

Coding Language is Python afaik. Would be great if anyone could help unpacking them.

tachiorz, posted Mon May 13, 2019 4:47 pm (47745)

---

https://gist.github.com/Tachiorz/2dec93 ... b905095103
I'm guessing archive is read recursively from the end of list to the beginning. For example archives in quest.tpk have different count of elements, but last one always is an array object with length of 45 elements that is populated with values, arrays and dicts before it.
I would really appreciate help with type 7 element that is object description I guess.

tachiorz, posted Tue May 14, 2019 11:18 am (47753)

---

byte type; // 7
short index_of_classname;
int count_or_size; // depends on classname?
// rest is if Dict - count of pairs of uints that are indexes to key/values
// if Array - count of uints that are indexes to values
// if Canvas_Static - size bytes that is blob with image
// etc
if count_or_size == -1 { // this case I still don't get
byte unknown1
byte unknown2
int count_or_size
}

aluigi, posted Sat Jun 01, 2019 9:35 pm (48266)

---

http://aluigi.org/bms/thing_archive_tarc.bms

But I think that some files may be compressed, for example there are PNG files but they have some bytes on front and aren't correctly visualized.
Give me your feedback.