Improve your disc file system so that file size is only limited
by disc size.

Do not assume that discs will always be as small as they have to
be under our virtual machine. Your implementation must be able
to work for any disc size for which a block number will fit into
a 32 bit int (4,294,967,296 blocks so total 2 TB). Clearly you
won't be able to run tests at such sizes, so reasonable errors
at scale will not be penalised.

You must use a dynamically changeable multi-level index format
as explored in class (class 6 mostly but there was some spread).
You decide on what the meta-data should be. Minimally it must
include exact file data size in bytes, and times and dates of
last modification and original creation, but think of a few
potetntially useful extras to add.

If the file's data is small enough to fit inthe header block, it
must do so (i.e. level 0 must be implemented).

Level 0 should be a trivial change to your original single block
system. If you are sensible you will make that first and test it
well, then move on to a strictly 0 or 1 level system. With a one
level index structure working properly, moving on to multiple
levels will not be very difficult. Doing everything at once will
be. A file's structure must always have the minimum number of
levels required for its size.

You do not have to manage any kind of free block list at this stage.
It must be possible to delete files, but their blocks may remain
permanently inaccessible. That comes next.