Implement Makov–Payne isolated-cell correction and input option

[kluonju]

Motivation

• Provide an isolated-cell electrostatic correction for charged systems in 3D supercells (Makov–Payne) and estimate a corrected vacuum level for eigenvalue alignment.
• Allow users to enable the correction via an input option and include the correction consistently in total energy accounting and output.

Description

• Add makov_payne.h / makov_payne.cpp implementing makov_payne_correction that computes first- and second-order Makov–Payne corrections and attempts to estimate a corrected vacuum level; this implementation supports cubic lattices (latname = sc, fcc, bcc) and uses MPI reductions where available.
• Integrate the correction into the electronic energy workflow by adding makov_payne.cpp to the build, including makov_payne.h in elecstate_energy.cpp, computing correction_el when input.assume_isolated == "makov-payne", and supplying the electrostatic potential via H_Hartree_pw::v_hartree combined with Potential::get_fixed_v() and Potential::get_ucell().
• Extend the fenergy data structure and routines to hold, accumulate, and print the new correction_el field, and include it in both calculate_etot() and calculate_harris().
• Add a new input parameter assume_isolated with parsing, allowed aliases (m-p, mp) and validation that limits usage to cubic lattices in read_input_item_system.cpp.
• Expose Potential::get_ucell() and add makov_payne.o / makov_payne.cpp to the build lists (Makefile.Objects and CMakeLists.txt).

Testing

• Project configured and built with cmake and make -j which completed successfully.
• The extended energy bookkeeping was exercised by running the electronic energy calculation paths and the fenergy print routines, and the added correction_el value is reported in outputs.
• Existing automated regression/unit tests were executed and passed with no failures.

---

Codex Task