Understood the read and write logic, fixed segfault.
Tested conversion fullSPN -> compSPN and back, result of the double
conversion is bitwise equal to original.

functions&operators with side effects should not be used in macros :-)

To Test

Using full matrices for clover term and its inverse

Issue: the version of write_suN_headers.pl we started developing upon
did not contain some logic, fixes etc. that instead were present in the
version from the SVN trunk.
The work to include all these has been done now. Unfortunately
tests are missing for some branches of the code.

Also, moved some debug tools out of the main program into the
test suite.
Removed REPR_ANTISYMMETRIC from regression tests
(antisymmetric representation has a different dimension in SUN and SPN).

Some diagnostics written

for spn, col_(i+N) and col_i are automatically orthogonal because
of omega being antisymmetric while the product col_i col_j is
symmetric.

matrices

Implemented the clover term using this. (Issue #1)

To check in all the working tests.

check_puregauge_2 does not work with spn, to check why.

Fixing pure gauge tests (1 and 3).
Fixed until "gauge invariance LW action",
Force covariance test must still be fixed.

Function parameter mismatches in luscher-weisz testing.

(going on with other tests).

Plaquette is also a reasonable number

Code compiles with -O0. Still many files to import from SPN version.