Object Store

The purpose of this system is to store and retrieve "objects".
An object is a file; in effect, you are implementing a flat file
system, i.e., one without directories or folders.  How you store
the objects is up to you, though you must not impose any unreasonable
artificial limits on object size.  ("Unreasonable" means any limit not
dictated by the size of the disk, your space quota, the maximum value
you can store in an integer variable, etc.)  You are perfectly welcome
to store each object as a separate Linux file.

There are multiple user and groups.  The list of users and what groups
they are in are also stored as objects.  That is, your code must use
other parts of your system to retrieve this data.  The userfile is composed
of lines in this format:

	username group1 group2 ...

There are three commands you must implement: objput, objget, objlist.
The arguments are as follows:

	objput -u username objname
	objget -u username objname
	objlist -u username [-l]

All commands must have a valid username.  When an object is created, the
creator's username is stored (somewhere, somehow).  You are welcome to create
a small number of system users.  Each user has their own namespace; two
different users can create objects of the same name.

objput reads the contents of the object from stdin; objget writes the
contents of a retrieved object to stdout.  objlist lists all of the objects
belonging to the specified user; if the -l option is given (and options
can appear in any order), all metadata associated with the object is
displayed as well.  At this point, that's just the size of the object,
but that will change...

Usernames, group names, and object names can contain letters, digits, and
underscores; no other characters are legal.

Pay a lot of attention to test data, and carefully read the packaging and
submission instructions.