I'm searching for semi-empirical quantum chemistry solver which would be easy to integrate into my own software. I found a few implementations which can be in principle used e.g. MOPAC, ORCA, SQM some of which are integrated e.g. into Molecular Design Toolkit. But I don't want to depend on 1GB packages when I need only a few functions. Also I wan't something where I can dig-in and modify it more easily.
I wonder if there is something more convenient with respect to following aspects:
- As much open as possible (GPL or better MIT licence)
- compact and simple to understand code-base (small number of lines, clear readability - no fortran77, minimum of external dependencies)
- C/C++ (optionally with python interface) ... I don't want to bother with frotran-C or fortran-python interface.
This is in principle continuation of my previous question Simple open-source Quantum chemistry or DFT code in C/C++. While Libint is very nice software, full Hartree-Fock is too slow for me. I really would like something cheaper, like semiempirical methods and DFTB.