📣 New Release: codeanalyzer-python 0.3.2

[github-actions[bot]]

This is a maintenance release for the 0.3 line, not the newest version.
The current release is 1.1.1. Upgrade to 0.3.2 only if you are pinned to 0.3.x;
everyone else should stay on the 1.x line.

What changed

numpy and pandas are no longer dependencies.

Neither package was ever imported by the analyzer, but both were declared as runtime
dependencies with tight upper bounds — numpy>=1.21.0,<1.24.0 on Python 3.9/3.10 and
<2.0.0 above it. Those bounds pinned installs to numpy releases that have no prebuilt
wheel for some targets, Red Hat UBI images in particular, so pip install codeanalyzer-python fell back to compiling numpy from source and failed.

Resolving the published package for x86_64-manylinux2014:

Python	0.3.1	0.3.2
3.9	numpy 1.23.5 + pandas 1.5.3	numpy 2.0.2 (no pandas)
3.11	numpy 1.26.4 + pandas 2.3.3	neither
3.12	numpy 1.26.4 + pandas 2.3.3	neither

On Python 3.11+ numpy leaves the dependency tree entirely — ray 2.55 does not require
it, and pandas was its only other route. On Python 3.9/3.10 ray==2.0.0 still requires
numpy, but without the <1.24 cap it now resolves to numpy 2.0.2, which ships cp39
manylinux wheels for x86_64 and aarch64. Either way, no source build.

This is a dependency-only change. No analyzer behaviour, emitted schema, or CLI surface
moves — 0.3.2 produces byte-identical output to 0.3.1.

Upgrade

pip install -U "codeanalyzer-python==0.3.2"

No migration steps. If you were working around the numpy build failure by pre-installing
a numpy build or pinning an older Python, you can drop that workaround.

Known gap

Python 3.9 and 3.10 still pull numpy in transitively through the ray==2.0.0 pin. Raising
that floor carries real compatibility risk and was deliberately left out of this release.

Links

• Release v0.3.2
• CHANGELOG
• Issue #124
• Compare v0.3.1...v0.3.2