Shadowman 2econd Coming [PS2]

[SpaceDog]

Hello. Help to unpack the game archive Shadowman 2econd Coming. Resources are stored in the file PS2DATA.bin. Packed with RNC2. The source files can be divided into several parts.
Link to resources
https://drive.google.com/folderview?id= ... lZISkhXTGM
Link to the disk image https://drive.google.com/folderview?id= ... zBuZ0NWcEk

[aluigi]

http://aluigi.org/bms/shadowman_2econd.bms

[SpaceDog]

Thank you! You are the best!

[SpaceDog]

I think that the unpacking is not quite correct, because in PS2DATA.bin the second and third file to be joined. I think that if the RNC before the header is 4 bytes different from 0 it is the beginning of a new file but if set to 0 then the continuation of the file. But I could be wrong.

[aluigi]

I understand, chunked files.
I will take another look at it.

[aluigi]

I don't think there are chunked files used in the archive, in fact there are even huge files of over 20 Mb.
For example your first image is taken from 000020e2.pme that terminates at 0x8000 but the file of the second image doesn't exist, I have no idea where you got it.
The files that contain "rs\icicle\iceberg3.tex" are all perfect files of the F2DA format.
In short there are no chunked files in my opinion and I have no idea why 000020e2.pme terminates before its (supposed) real end.

*update* I made other tests and I confirm that everything is correct in the script, including the 000020e2.pme file.

[SpaceDog]

In the file PS2DATA 20173 files with the header RNC2. I think that your script takes an offset from the dir file. And he correctly decompresses and reads the file from the archive. But if the file is broken into several parts that is extracted only the first file the rest are not retrieved . The PS2DATA file located at offset 0x5e88 (PME). and it continued to offset 0х7544.

[SpaceDog]

also, you can check for files with the extension .f2d. Header (46324441), the next 4 bytes of file size - 4.

[aluigi]

That explains the 85% of coverage that I missed.
I will check it again.

[aluigi]

Script 0.2, now with 99% of coverage

Archives like this are really bad because it's difficult to notice these problems during the analysis, moreover if only the compressed files use this particular method and that 32bit field doesn't have a clear meaning (I have simply defined it as REMAINING to check if it's different than zero).

[SpaceDog]

Thank you so much. It seems to be true.)