Overhaul the current filesystem to use content hashes

The goals of the new filesystem are:

* allow files to be stored within Mahara (as we do now).
* use a consistent and flexible approach for all file handling throughout Mahara
* track where files originally came from
* avoid redundant storage, when the same file is used twice
* fully support Unicode file names, irrespective of the capabilities of the underlying file system

Roadmap:
* Use sha-256 to calculate the content hash of files (this is because sha-1 collisions are feasible now)
* Changes to the database structure to enable use of the new file structure
* A migration/upgrade script to facilitate moving from the old file structure to the new one
* Backwards compatibility with the old file structure for plugins that are yet to be updated (potentially)