Feature ID's in MI-JSON output

[colin-combe]

The ID's of features in MI-JSON output are auto-incremented integers.

This is fine when working with the output from converting a single PSI-MI file. However, if wishing to, for example, visualise several PSI-MI files at once the feature ID's collide.

The unique feature ID's need to be more unique - e.g. making them by concatenating the PSI-MI XML filename and an auto-incremented integer would work.