I'm writing a program that extracts and adds files to the xbox 360's STFS files. The STFS structure is a mini file system, it has hashtables, a file table, etc.
Extracting the files is simple enough. I have the starting block of the file and the amount of blocks in the file, so I just need to find the block offsets, read the block lengths, and send that out as the file. What happens, though, when I need to replace or remove a file? I've read that on Windows and computers in general, files aren't actually deleted, they're just removed from the file table and are overwritten when something else needs the space. When I'm writing a file, then, how do I find an unused sequence of blocks large enough to hold it? The blocks are 0x1000 bytes in length and fill remaining space with empty bytes, so everything evens out nicely, but I can't think of an efficient way to find an unused range of blocks that will fit the file I want to add.
My current plan is to rewrite everything on removing or adding a file so that I don't have large amounts of unused space that I'm unable to figure out how to overwrite. Is there a good introduction to file systems like NTFS or FAT32 that I could read that won't take days to understand and will contain the necessary information to write a basic file manager?
reference to structure: http://free60.org/STFS
edit: on Second though, I would create a list of ranges for each file in the table. That is, the start offset and end offset based on size. When looking for an open range to insert a file, I would start at 0 and check if the end start or end of each range is inside the range needed by the file to be inserted. If either the start or end is inside the range, I would move on to the end of the other file's end offset. This is better than my initial idea, but still seems inefficient. I would have to make multiple comparisons for every file in the file table.
