[Critsium-xy]

Reminder

• I have read AGENTS.md and docs/developers_guide/agent_governance.md.
• I have linked an issue or explained why this PR does not need one.
• I have added adequate unit tests and/or case tests, or explained why not.
• I have listed the exact verification commands run and their results.
• I have described user-visible behavior changes, including INPUT parameter changes.
• I have explained core-module impact for ESolver, HSolver, ElecState, Hamilt, Operator, Psi, or other source/ changes.
• I have requested any needed governance exception below.

Linked Issue

Part of #7599 and #7604. This removes two of the includes from source_hamilt into source_lcao and source_pw, and one from source_pw into source_lcao. The remaining source_hamilt ones (exx_info_ri.h, cal_pseudo.cpp) overlap open PRs #8176 and #8222 and are left for later.

Unit Tests and/or Case Tests for my changes

• Commands run:
  • python3 tools/03_code_analysis/agent_governance_check.py --staged --format text: warnings only, listed under Governance Notes.
  • A static include-closure comparison against develop @ e5ad53046 for every file that reaches one of the changed headers. No file loses a header whose classes, aliases or macros it uses.
• Result summary: not compiled locally yet. This is a draft until CI builds it.
• Checks not run, with reason:
  • No local build or tests: the remote build machine I normally use was unavailable.
  • The relocated conversion only runs with out_mat_dh 1, nspin 2 and gamma_only 0. The two existing out_mat_dh cases (02_NAO_Gamma/scf_out_dh, 03_NAO_multik/scf_out_dh_t) use gamma_only 1 and nspin 1, so no CI case reaches it. Before marking this ready, I will run a copy of 03_NAO_multik/scf_out_dh_t with nspin 2 on this branch and on develop, and compare the written dH matrices.

What's changed?

Two commits, no intended behavior change:

1. Narrow three header includes to what the header actually uses.
   • source_hamilt/module_xc/kernels/xc_functional_op.h: source_psi/psi.h → source_base/module_device/types.h. The kernel only needs GetTypeReal (already included from source_base/macros.h) and the base_device tags.
   • source_hamilt/module_vdw/vdwd2_parameters.h: source_io/module_parameter/parameter.h → input_parameter.h. Only Input_para is used; none of its includers touch PARAM.
   • source_pw/module_pwdft/op_pw_exx.h: source_lcao/module_ri/conv_coulomb_pot_k.h → source_hamilt/module_xc/coulomb_config.h. Only Conv_Coulomb_Pot_K::Coulomb_Type is used, and it is defined there. The two files that call conv_coulomb_pot_k.h functions (ewald_vq.hpp, exx_lri.hpp) include it directly.
2. Move the dV/dR sparse conversion from module_gint to source_lcao.
   • Gint_dvlocal::cal_dvlocal() integrates ⟨φ|V|∇φ⟩ on the grid into three HContainers; that part stays.
   • Its other two members, cal_dvlocal_R_sparse and distribute_pvdpR_sparse, only converted that result into the LCAO_HS_Arrays format of the dH output. They were why module_gint included source_lcao/lcao_hs_arrays.h and source_lcao/module_ri/abfs_vector3_order.h.
   • Both are now file-local functions in source_lcao/spar_dh.cpp, next to their only caller sparse_format::cal_dH, and take the HContainers from the existing get_pvdpR*() getters.
   • The bodies are unchanged. The only edits: dead locals and the unused nspin/Grid_Driver parameters are dropped, and trace_lo is read through the public Gint::get_gint_info() instead of the protected member.
   • The wrapper ModuleGint::cal_dvlocal_R_sparse and the LCAO_HS_Arrays forward declaration in gint_interface.h are removed. veff_dh.cpp, the other user of Gint_dvlocal, only calls cal_dvlocal() and the getters and is untouched.

Governance Notes

• INPUT/docs changes: none; no INPUT parameter or output format changes.
• Core module impact: source_hamilt/module_gint, module_xc (kernel header), module_vdw (one header), source_pw/module_pwdft/op_pw_exx.h, source_lcao/spar_dh.cpp. No new files, so no CMake or Makefile changes.
• Exceptions requested:
  • Global dependency budget: net 0. The one PARAM.globalv.nlocal line in spar_dh.cpp is the existing call site, reformatted.
  • Header include warnings (three): each one replaces an include with a narrower one.

🤖 Generated with Claude Code

[ABACUS-PR-BOT]

Static review found no implementation defect in the relocation or narrowed headers. git diff --check passed; build, runtime, and CI results were not verified.

[ABACUS-PR-BOT]

Non-blocking: Please retain the planned nspin=2, multi-k out_mat_dh comparison as a regression for all three Cartesian components in both spin channels. The existing legacy CSR checks in tests/integrate/validation_tools/props_mat.sh compare only s1 files, so reusing them unchanged would leave s2 unchecked.