Addresses Issue #236 by enabling access to Cartesian one-body property derivatives.
Acts similarly to the DISABLE_ONEBODY_PROPERTY_DERIVS by just skipping formation of the files if SphericalMultipoleOper && deriv_level > 0 is detected.
Obviously, the configure.ac file has no effect for when Libint is switched to CMake, but I've changed the build_libint.cc with Lori's branch new-cmake-2023-take2-b with Psi4 before and this works with the change as well.
In response to @loriab Issue #236, I am able to match Psi4's McMurchie-Davidson dipole derivatives through a separate implementation in the GitHub repo Quax (https://github.com/EricaCMitchell/Quax/blob/dipoles/quax/integrals/libint_interface.cc). So the dipole derivative ints Psi4 is seeking may just be a small issue on the Psi4 side or fixed by improvements to Libint since the issue was raised.
Addresses Issue #236 by enabling access to Cartesian one-body property derivatives.
Acts similarly to the
DISABLE_ONEBODY_PROPERTY_DERIVS
by just skipping formation of the files if SphericalMultipoleOper && deriv_level > 0 is detected.Obviously, the configure.ac file has no effect for when Libint is switched to CMake, but I've changed the build_libint.cc with Lori's branch new-cmake-2023-take2-b with Psi4 before and this works with the change as well.
In response to @loriab Issue #236, I am able to match Psi4's McMurchie-Davidson dipole derivatives through a separate implementation in the GitHub repo Quax (https://github.com/EricaCMitchell/Quax/blob/dipoles/quax/integrals/libint_interface.cc). So the dipole derivative ints Psi4 is seeking may just be a small issue on the Psi4 side or fixed by improvements to Libint since the issue was raised.