Fix shared NumPy assignment_by_sum duplicate indices

[FlorianPfaff]

Summary

• Add a shared NumPy/autograd compatibility hook so assignment_by_sum accumulates repeated advanced indices instead of applying NumPy's buffered += semantics.
• Load the hook during package initialization through the existing stability compatibility path.
• Add a focused backend-portable regression for duplicate vector and tuple indices.

Bug fixed

For shared NumPy backends, assignment_by_sum(x, [1, 2], [0, 0]) previously produced [2, 0, 0] because NumPy's advanced-indexed += buffers repeated indices. PyTorch and JAX accumulation paths add both contributions and produce [3, 0, 0]. The hook uses numpy.add.at for the non-vectorized shared NumPy path to match true scatter-add semantics.

Validation

• python -m py_compile on the new compatibility hook and regression test in a local scratch copy.
• Verified the GitHub branch diff against main after resetting an earlier too-broad file replacement.
• Full repository test suite not run in this connector-only environment; CI should run the focused backend-portable regression.

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ COPYPASTE	jscpd	yes		no	no	79.24s
✅ JSON	prettier	7	0	0	0	1.13s
✅ JSON	v8r	7		0	0	4.36s
✅ MARKDOWN	markdownlint	68	0	0	0	1.97s
✅ MARKDOWN	markdown-table-formatter	68	0	0	0	0.77s
✅ PYTHON	black	1187	30	0	0	66.81s
✅ PYTHON	isort	1187	54	0	0	4.66s
✅ REPOSITORY	checkov	yes		no	no	49.62s
✅ REPOSITORY	gitleaks	yes		no	no	11.51s
✅ REPOSITORY	git_diff	yes		no	no	0.21s
✅ REPOSITORY	secretlint	yes		no	no	46.14s
✅ REPOSITORY	syft	yes		no	no	6.28s
✅ REPOSITORY	trivy-sbom	yes		no	no	6.59s
✅ REPOSITORY	trufflehog	yes		no	no	21.16s
✅ YAML	prettier	11	0	0	0	0.72s
✅ YAML	v8r	11		0	0	12.24s
✅ YAML	yamllint	11		0	0	0.59s

Notices

📣 MegaLinter 9.5.0 is out! Discover the new features and security recommendations in the release announcement. (Skip this info by defining SECURITY_SUGGESTIONS: false)

See detailed reports in MegaLinter artifacts

Your project could benefit from a custom flavor, which would allow you to run only the linters you need, and thus improve runtime performances. (Skip this info by defining FLAVOR_SUGGESTIONS: false)

• Documentation: Custom Flavors
• Command: npx mega-linter-runner@9.5.0 --custom-flavor-setup --custom-flavor-linters PYTHON_BLACK,PYTHON_ISORT,COPYPASTE_JSCPD,JSON_V8R,JSON_PRETTIER,MARKDOWN_MARKDOWNLINT,MARKDOWN_MARKDOWN_TABLE_FORMATTER,REPOSITORY_CHECKOV,REPOSITORY_GIT_DIFF,REPOSITORY_GITLEAKS,REPOSITORY_SECRETLINT,REPOSITORY_SYFT,REPOSITORY_TRIVY_SBOM,REPOSITORY_TRUFFLEHOG,YAML_PRETTIER,YAML_YAMLLINT,YAML_V8R

Show us your support by starring ⭐ the repository