PR- [11]: Final PR by San1357 · Pull Request #259 · theochem/gbasis

San1357 · 2026-08-13T05:37:37Z

Here is the Problem statement:

#229

Summary

Combines all PR-1 to 11 implementations into a single branch viz PR-11 for review

Related PRs

Checklist

Write a good description of what the PR does.
Add tests for each unit of code added (e.g. function, class)
Update documentation
Squash commits that can be grouped together
Rebase onto master

Type of Changes

	Type
✓	🐛 Bug fix
✓	✨ New feature
✓	🔨 Refactoring
✓	📜 Docs

Closes #229

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

PR[1]: build: migrate from setuptools to scikit-build-core

PR[2] : feat: add libcint C extension bindings for various integrals

…int v6

PR[3]: feat: use PyArray_GETPTR instead of PyLong_AsVoidPtr + fix dylib rpath

PR[4]: fix: platform-aware libcint loading + moment integral libcint v6 fix

PR-5 - feat: add C shell loop for all 1-electron integrals

…pport refactor: rename macros to DEFINE_INT1E_ARRAY_FN and DEFINE_INT1E_SHELLLOOP_FN fix: fill ERI in physicist notation in C feat: add C shell-loop binding for point_charge integral feat: add C shell-loop bindings for moment integrals refactor: rename macros to DEFINE_INT1E_ARRAY_FN and DEFINE_INT1E_SHELLLOOP_FN fix: fill ERI in physicist notation in C feat: add C shell-loop binding for point_charge integral feat: add C shell-loop bindings for moment integrals feat: update momentum() to return full 3-component complex array docs: update momentum() docstring with notes test: add C shell-loop tests for 1e integrals & 2e integrals bindings in (PR-5 &PR-6) refactor: dynamic buf allocation + rename shellloop to array feat: add optimizer support feat: add optimizer support using CINTall_1e_optimizer and cint2e_sph_optimizer perf: recorrect optimizer support via token pasting — 49x speedup fix: restore optimizer line after benchmark testing

… support test: add C shell-loop tests for gradient integrals (ipkin, ipnuc, iprinv) feat: add GIAO/magnetic integral bindings (ia01p, ircxp, igkin, igovlp, ignuc) test: add GIAO/magnetic integral tests (ia01p, ircxp, igkin, igovlp, ignuc) commented: to remove unused import style: apply clang-format, black, pycodestyle, pydocstyle formatting add transform support to overlap() C-loop function add transform support to kinetic_energy() C-loop function add transform support to nuclear_attraction() C-loop function add inv_origin and transform support to rinv() C-loop function add notation and transform support to electron_repulsion() C-loop function add transform support to dipole, quadrupole, octupole C-loop functions add transform and inv_origin support to gradient_kinetic, gradient_nuclear, gradient_rinv add transform, origin, inv_origin support to all 17 C-loop wrapper functions

perf: zero only used buffer portion in eri_array instead of full buf feat: add int3c2e_cart binding, cartesian 3c2e tests, coord_type aware three_center_two_electron test: add cartesian shell-loop tests and 3c2e cart reference tests

ci: add CMake install step for scikit-build-core on GitHub Actions ci: verbose pip install, remove coverage threshold ci: fix RPATH for Linux, install libcint.so alongside extension revert: remove unnecessary libcint.so install step (static linking) ci: use no-build-isolation with explicit cmake install fix: cartesian normalization, momentum, point_charge, moment/momentum skip - Add _ovlp_minhalf cartesian normalization in __init__ - Apply normalization in all 2D methods - Apply 4D normalization in electron_repulsion - Fix point_charge: add permutation before storing val - Add momentum() method with -1j scaling - Add transform parameter to point_charge() and moment() - Skip momentum/moment tests (multi-component C fix needed) - Skip cc-pVDZ cartesian ERI (d-shell normalization TODO) fix: add CMAKE_POSITION_INDEPENDENT_CODE for Linux shared object build fix: momentum multi-component fix via DEFINE_INT1E_LOOP_FN_MULTICOMP macro fix: moment mixed orders via full multi-component dipole/quadrupole/octupole buffers fix: cc-pVDZ cartesian ERI via eri_array_cart with cint2e_cart_optimizer fix: cc-pVDZ cartesian ERI and momentum normalization fixes

…fixes fix: add setuptools_scm for Windows CI fix: add setuptools_scm for all platforms fix: install numpy before build feat: add MinGW support for Windows CI build fix: force MinGW gcc via env vars for window fix: skip test_density.py on Windows fix: test_deriv on Windows fix: ignore eval_hermite related tests on Windows fix: ignore all eval_hermite related tests on Windows fix: bump scipy to >=1.18.0 revert: restore scipy>=1.13.0 fix: eval_hermite dtype compatibility with NumPy 2.0 fix: eval_hermite dtype fix for NumPy 2.0 fix: NumPy 2.0 dtype compat on Windows fix: remove jwlawson cmake action, use pip cmake instead fix: use .real cast for eval_hermite Windows compat fix: force alphas to float64 for Windows NumPy compat fix: use np.int32 for eval_hermite Windows fix: remove win32 skip from libcint tests fix: update libcint glob to match .pyd on Windows fix: use importlib to check libcint_bindings on Windows fix: build libcint as shared lib on Windows for DLL loading fix: add DLL directory for Windows libcint loading fix: force shared lib build on Windows fix: add MinGW bin to PATH for Windows DLL loading fix: copy MinGW runtime DLLs to lib dir on Windows fix: manually copy MinGW runtime DLLs for Windows fix: pre-load libcint.dll on Windows before importing pyd fix: static link libcint + MinGW runtime on Windows fix: static link MinGW runtime for .pyd on Windows fix: add MinGW bin to PATH in conftest for Windows DLL loading fix: set MinGW PATH before pytest collection via pytest_configure fix: explicitly C-contiguous array for Windows libcint overlap

skip: libcint tests on Windows CI - heap corruption in MinGW build

msricher

Can you run the C code through clang-format again? Otherwise, check that the C code naming conventions match the Python names well enough, and please make the changes suggested in my comments. Thanks!

msricher · 2026-08-13T13:42:00Z

+        Raises
+        ------
+        ValueError
+            If ``coord_type`` is not 'spherical' or 'cartesian'.
+
+


Can you make these keyword arguments consistent? Sometimes "spherical"|"cartesian" is used, sometimes "sph"|"cart".

msricher · 2026-08-13T13:43:48Z

+    pytest.param(["He"], np.asarray([[0.0, 0.0, 0.0]]), "He", "spherical", id="He-sph"),
+    pytest.param(["C"], np.asarray([[0.0, 0.0, 0.0]]), "C", "spherical", id="C-sph"),
+    pytest.param(["H", "He"], np.asarray([[0.0, 0.0, 0.0], [1.5117, 0.0, 0.0]]), "H_He", "spherical", id="H_He-sph"),
+    pytest.param(["Be", "C"], np.asarray([[0.0, 0.0, 0.0], [1.8897, 0.0, 0.0]]), "Be_C", "spherical", id="Be_C-sph"),
+    pytest.param(["He"], np.asarray([[0.0, 0.0, 0.0]]), "He", "cartesian", id="He-cart"),
+    pytest.param(["C"], np.asarray([[0.0, 0.0, 0.0]]), "C", "cartesian", id="C-cart"),
+    pytest.param(["H", "He"], np.asarray([[0.0, 0.0, 0.0], [1.5117, 0.0, 0.0]]), "H_He", "cartesian", id="H_He-cart"),
+    pytest.param(["Be", "C"], np.asarray([[0.0, 0.0, 0.0], [1.8897, 0.0, 0.0]]), "Be_C", "cartesian", id="Be_C-cart"),
+]


If you load the reference arrays from a file, there's no need to include these placeholder [0, ..., 0] arrays in the parameter lists. You can remove them, and the corresponding arguments to the test functions where they are used.

RichRick1

Great job!

marco-2023 · 2026-08-24T16:39:11Z

@msricher, can I have a little more time to review it?

msricher · 2026-08-24T19:23:05Z

@msricher, can I have a little more time to review it?

Of course, we are finished now, so I'm just inviting the core gbasis maintainers to review before merging it myself.,

Aayush Gupta and others added 25 commits June 5, 2026 00:42

build: migrate from setuptools to scikit-build-core

0fb4fce

build: restore comments and fix numpy version consistency

8a37c73

build: add CMakeLists.txt with FetchContent and platform detection

b7c0938

style: fix comments

4b3ac64

build: fix misleading comment about setuptools in pyproject.toml

467ae73

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

build: address Copilot review comments

334ecc4

Merge pull request #1 from San1357/pr-1/scikit-build-migration

f5841ad

PR[1]: build: migrate from setuptools to scikit-build-core

feat: add libcint C extension bindings for various integrals

816a7ff

added documentation

66a9036

Merge branch 'dev' into pr-2/c-extension-bindings

0f94b86

Merge pull request #2 from San1357/pr-2/c-extension-bindings

78b8456

PR[2] : feat: add libcint C extension bindings for various integrals

feat: switch to NumPy C-API (PyArray_GETPTR) + fix dylib runtime loading

14c0912

fix: platform-aware libcint loading + moment integral naming for libc…

1d71026

…int v6

fix: rename libcint dylib at install time

50e63fa

remove SKBUILD_SOABI block

2ef1438

fix: platform-aware loading + moment integral libcint v6 fix

6a6d32d

Merge pull request #3 from San1357/pr-3/numpy-c-api-bindings

0d9e60d

PR[3]: feat: use PyArray_GETPTR instead of PyLong_AsVoidPtr + fix dylib rpath

Merge branch 'dev' into pr-4/libcint-py-fix

74a71a9

Merge pull request #4 from San1357/pr-4/libcint-py-fix

6e5feb5

PR[4]: fix: platform-aware libcint loading + moment integral libcint v6 fix

feat: add C shell loop for overlap only

ee40693

feat: macro-based C shell loop bindings for all 1-electron integrals

cf9b7f5

moved the import at top

c00d380

fix: rename shellloop methods and improve docstrings per review

fd88efc

build: update CMakeLists.txt per mentor review

09e96d2

Merge pull request theochem#253 from San1357/pr-5/shell-loop-c-new

d927c06

PR-5 - feat: add C shell loop for all 1-electron integrals

San1357 changed the title ~~Pr 11/final pr~~ PR- [11]: Final PR Aug 13, 2026

Aayush Gupta added 4 commits August 13, 2026 11:18

Aayush Gupta added 2 commits August 13, 2026 11:19

test: skip libcint tests on Windows CI pending MinGW heap corruption fix

6360595

skip: libcint tests on Windows CI - heap corruption in MinGW build

San1357 force-pushed the pr-11/final-PR branch from fed05d7 to 6360595 Compare August 13, 2026 05:50

docs: improve docstrings and comments

9ef5886

San1357 requested a review from msricher August 13, 2026 12:18

msricher requested changes Aug 13, 2026

View reviewed changes

Aayush Gupta added 4 commits August 13, 2026 20:01

fix: clang-format, coord_type quotes, 3c2e test cleanup

8ba1aaa

Jupyter Tutorial notebook

2d1fe90

Final tutorial notebook

96fa7e5

docs: add installation section in tutorial notebook

91470d2

msricher requested review from FarnazH, RichRick1 and marco-2023 August 18, 2026 13:09

RichRick1 approved these changes Aug 18, 2026

View reviewed changes

msricher approved these changes Aug 24, 2026

View reviewed changes

msricher marked this pull request as ready for review August 26, 2026 15:10

Navigation Menu

Uh oh!

PR- [11]: Final PR - #259

PR- [11]: Final PR #259
San1357 wants to merge 36 commits into
theochem:masterfrom
San1357:pr-11/final-PR

San1357 commented Aug 13, 2026 •

edited

Loading

Uh oh!

msricher left a comment

Uh oh!

msricher Aug 13, 2026

Uh oh!

San1357 Aug 13, 2026

Uh oh!

msricher Aug 13, 2026

Uh oh!

San1357 Aug 13, 2026

Uh oh!

RichRick1 left a comment

Uh oh!

marco-2023 commented Aug 24, 2026

Uh oh!

msricher commented Aug 24, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Uh oh!

Conversation

San1357 commented Aug 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Here is the Problem statement:

Summary

Related PRs

Checklist

Type of Changes

Uh oh!

msricher left a comment

Choose a reason for hiding this comment

Uh oh!

msricher Aug 13, 2026

Choose a reason for hiding this comment

Uh oh!

San1357 Aug 13, 2026

Choose a reason for hiding this comment

Uh oh!

msricher Aug 13, 2026

Choose a reason for hiding this comment

Uh oh!

San1357 Aug 13, 2026

Choose a reason for hiding this comment

Uh oh!

RichRick1 left a comment

Choose a reason for hiding this comment

Uh oh!

marco-2023 commented Aug 24, 2026

Uh oh!

msricher commented Aug 24, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

San1357 commented Aug 13, 2026 •

edited

Loading