feat(diskann): expose Python update APIs

[ScissorJack-ever]

Description

Expose the standalone updatable DiskANN index through Python bindings.

Python callers can build, load, search, insert, and remove vectors using external IDs only. External-ID deletion resolves the corresponding internal slot and reuses the existing internal deletion implementation.

Related Issue

N/A

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Performance improvement
• Code refactoring
• Documentation update
• CI/CD changes

Changes Made

• Add the alayalite.diskann Python module with build, load, search, insert, batch insert, remove, batch remove, contains, and flush APIs.
• Maintain a live external-ID-to-internal-slot mapping for updatable DiskANN indexes.
• Resolve external IDs before deletion and reuse the shared internal single and batch deletion implementations.
• Validate reserved, duplicate, and unknown external IDs before mutating the index.
• Rebuild the external-ID mapping when loading an updatable index and keep it synchronized across insertion, deletion, slot reuse, flush, and reload.
• Convert missing external IDs to Python KeyError.
• Add C++ and Python coverage for external-ID updates, validation atomicity, slot reuse, and persistence.

Testing

• Unit tests added/updated
• Integration tests added/updated
• Manual testing performed

/tmp/alayalite-python-update-build/tests/diskann/test_diskann_index

ALAYA_DISKANN_UPDATE_IO=blocking \
  /tmp/alayalite-python-update-build/tests/diskann/test_diskann_update_e2e

PYTHONPATH=/tmp/alayalite-python-update-build/python:python/src \
  .venv/bin/python -c \
  "import sys, _alayalitepy; sys.modules['alayalite._alayalitepy'] = _alayalitepy; import pytest; raise SystemExit(pytest.main(['-q', 'python/tests/diskann/test_update.py']))"

Results:

• DiskANN C++ unit tests: 16 passed
• DiskANN update E2E tests: 23 passed
• Python DiskANN update tests: 4 passed
• Commit-time lint, formatting, license, and static checks passed

Checklist

• My code follows the project's coding style
• I have run make lint and fixed any issues
• I have added tests that prove my fix/feature works
• All new and existing tests pass (make test)
• I have updated documentation if needed
• My commits follow the conventional commits (https://www.conventionalcommits.org/) format

Additional Notes

The Python update API intentionally exposes external IDs only. Internal slot IDs remain an implementation detail of the C++ index.

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!