sPuReMD: standardize QM/MM interface. Add test codes for QM/MM interface for AVE dataset in Python and Fortran.

sPuReMD: merge changes from branch 'new_setup'. Rework interface for geometry initialization without file I/O. Other code clean-up.

sPuReMD: rework Fortran F90 interfaces for Amber specifics and declare this interface for such usage.

sPuReMD: fix Coulomb energy calculations for QM/MM mode. Indicate QEq model not supported with QM/MM mode. Add Fortran interface for QM/MM mode.

sPuReMD: add QMMM library interface for external codes (e.g., Amber). Rework file I/O parsing to jive with multiple consecutive simulation code changes. Improve file parser error checking (numeric types). Other code clean-up.

sPuReMD: fixes for multiple consecutive simulation codepath. Reorganization to this codepath (some data structure now carry a flag to designation their allocation state). Fix bug in force field parser (atom type not truncated correctly). Charge sparse_matrix structs to be non-pointers to jive with new allocation state flag. Other code clean-up.

sPuReMD: finalize sPuReMD multiple consecutive simulation code (for integration with other codebases).

sPuReMD: add functionality to run multiple MD simulations (for usage via integration with external codebases).

sPuReMD: output PDB files according to write_freq value in control file in addition to outputting to the trajectory file. Small tweak to PDB output format to be in compliance with specification.

sPuReMD: add control file parameter to compute atomic system pressure for non-NPT ensembles. Fix issues with isotropic NPT ensemble (incorrect pressure and temperature scalers). Change atomic position remapping behavior to include upper boundaries of the simulation box (shared memory code only). Other code clean-up.

sPuReMD: corrections to external pressure calculations for NPT-type ensemble simulations (use correct periodic images between reneighboring steps). Other code clean-up.

sPuReMD: fix issue with torsion & conjugation forces being incorrect due to rel_box arithmetic being wrong (incorrect directions for position vectors). Redesign atomic distance and position update functions to support both periodic boundary conditions and non-periodic boundary conditions.  Behavior change: let random_velocity parameter in control file take precedent over specifying a non-zero initial temperature value. Other code clean-up.

sPuReMD: small updates to Python wrapper (data structure changes). Fix issue with system energies not being computed for step 0 with the wrapper.

sPuReMD: fix issue with certain preconditioners not being initialized for runs using restart files. Fix issue with part of simulation box struct not being initialized for runs using restart files (box->min). Other code clean-up from debugging issues with SMALL_BOX_SUPPORT.

sPuReMD: add safety check for BGF input files (CRYSTX). Remove compiler warnings when enabling DEBUG_FOCUS.

sPuReMD: refactoring to better align with PuReMD code. PuReMD: backport sPuReMD changes from validation efforts against Fortran ReaxFF code.

sPuReMD: fix issue with hbonds list indices not being reset correctly. Small algebraic rearrangements of valence angle calculations to match Fortran ReaxFF code (no numeric changes). Fix issue with rvec_isZero being incorrect. Fix issue with debugging arrays for TEST_FORCES not having indicies initialized. Other clean-up work.

sPuReMD: re-enable SMALL_BOX_SUPPORT. Fix issue with far nbrs list being incorrectly generated for small boxes (missing nbr ID). Fix issue with ACKS2 matrix being incorrect for small boxes due to multiple non-zeros being added for each perioidic image of an atom (accumulate instead). Add missing ACKS2-specific energy and forces term. Clean-up far nbr list estimation and generation code for small boxes. Do not add far list struct entries unless the atom is below the specified cutoff.

sPuReMD: refactoring to align code with PuReMD. Disable SMALL_BOX_SUPPORT. Refactor to remove union-based polymorphic behavior for reax_list. Other general code clean-ups.

sPuReMD: backport changes for reax_list, struct declaration/definition separation, interaction_function.

sPuReMD: make CoM logic consistent with PuReMD. Other build system updates for debugging and newer versions of MPICH.

sPuReMD: add checks on string and file handling. Rewrite PDB writing to be line-by-line. Other refactoring.

sPuReMD: remove debug prints. Default to no good initial guess for out of range values for spline extrapolation.

sPuReMD: move NaN checks on system energy to after logging. This allows better diagnostics for, e.g., the potential logging.

sPuReMD: separate total energy computation and fix Python driver bug with structure definition for ctypes. Allow disabling file logging during simulation.

sPuReMD: fix GMRES bug introduced from previous refactoring. Explicitly link OpenMP for library. Refactoring to remove global variables. Remove unnecessary OpenMP barriers after single regions.

sPuReMD: split into library and driver for, e.g., optimization integration. Fix bug with interaction lists being allocated for hydrogen bonds when no hydorgen atoms present. Fix minimum allocation size of 3-body lists (uncovered with small lithium geometries). Refactor naming of conflicting data structures. Update .gitignore.

PG-PuReMD: fix error due to function prototype change. PuReMD: fix issues with GCC new inline behavior, fix reax_list bugs. sPuReMD: change exit code to zero.

sPuReMD: add safe memory allocation functions. Fix issue with certain parameters for shielded van der Waals interactions not being correctly initialized. Other general refactoring.

sPuReMD: further OpenMP work on force computations and interaction lists. Minor data structure scope changes. Fix numereous non-critical memory leaks. Add tear-down code. Other general refactoring.