relax SVN archive

Added a fix for the aliasing of a PDB file for when no structural data exists in the data pipe.

This is related to bug #12409 (https://gna.org/bugs/index.php?12409), so that rather than an
unrelated RelaxError, the structure is not aliased is the structural data could not be loaded in
another data pipe.

Fix for the checks in the test_load_scientific_results() system test.

Partial fix for bug #12409 (https://gna.org/bugs/index.php?12409).

Updated the add_struct() method to match that of the internal structural object.

Created the new test_load_scientific_results() system test.

This test checks the loading of a PDB file using the Scientific python PDB internal object when
reading a results file. The test_load_results() system test has been renamed to
test_load_internal_results().

Made the structure reading from a results file system test specific to the internal structure object.

Partial fix for bug #12407 (https://gna.org/bugs/index.php?12407).

This catches the "TypeError: unsupported operand type(s) for *: 'float' and 'NoneType'" error. If
the local or global tm is missing for some reason, the eliminate function returns False indicating
that the model has not been eliminated.

Fix for the test_tm_51ns() system test.

The model 'tm4' is set rather than 'm4'.

Conversion to bools in determine_model_type() and a fix for bug #12408 (https://gna.org/bugs/?12408).

All internal flags in the method have been converted to Booleans. This commit also contains a bit
of a fix for bug #12408. The bug was a false alarm and now determine_model_type() raises a more
relevant RelaxError.

Created a system test for model-free model elimination when local tm > 50 ns.

Completed the fix for bug #12407 (https://gna.org/bugs/index.php?12407).

This latest problem was caused by the use of the wrong Boolean operator in the model-free
eliminate() method. The method has also been converted to return the Booleans True or False.