d3q Code

bug fix, thank to Truong for reporting

Added a nfar option to the d3_q2r input. When set to zero, it produces non-recentered force constants which are very similar to the ones from standard q2r, except that:
1. The cell index is R in crystal coordinates (in q2r it is R+(111))
2. The order of the Rs is different
3. Before every block of R there is the number of vectors in the block

Removed obsolete stuff

Added support for new xmltools with i/o of complex tensors

remove iotk in/out of matrices

Removed dependency on iotk

Removed dependence on iotk and a few fixes here and there

backup

Subroutine to write d3matrix using fox (reading still to be done)

minor edits to the manual