fix: lerp colors in current color mode

[pupuking723]

Resolves #8883

Changes:

• Makes Color.lerp() interpolate in the explicit mode passed by lerpColor(), instead of interpolating in the source colors' original color space and only formatting the result in the current mode.
• Keeps the existing fallback behavior when no mode is passed.
• Updates the existing HSL/HSB lerpColor() expectations to match current-mode interpolation. I did not add a new regression test, following the maintainer request in [p5.js 2.0+ Bug Report]: lerpColor() not affected by colorMode() #8883 to start with just the fix.

Screenshots of the change:
Not applicable.

PR Checklist

• npm run lint passes
• [Inline reference] is included / updated
• [Unit tests] are included / updated

Verification:

• npm test -- --run test/unit/color/creating_reading.js test/unit/color/p5.Color.js
• npm run lint

AI assistance disclosure: I used Codex assistively while preparing this change. I reviewed the issue discussion, source changes, test updates, and verification output myself, and I take responsibility for the PR.

[p5-bot]

Continuous Release

CDN link

• https://raw.esm.sh/pr/p5@e5b5f2e/lib/p5.min.js

Published Packages

• https://pkg.pr.new/p5@e5b5f2e

Commit hash: 10cf56d

Previous deployments

---

This is an automated message.

[pupuking723]

Thanks for taking a look at this.

I checked the current failing CI job. The only failing test I found is in test/unit/webgl/p5.Framebuffer.js for saveCanvas, where the assertion expected document.createElement("a") but the recorded calls were canvas. That appears unrelated to the lerpColor() / color-mode change in this PR.

For the color-related change, I verified the affected tests locally:

• npm test -- --run test/unit/color/creating_reading.js test/unit/color/p5.Color.js
• npx eslint src/color/p5.Color.js test/unit/color/creating_reading.js

One clarification on the test diff: I did not add a new regression test, but the existing HSL/HSB lerpColor() expectations encoded the old RGB interpolation behavior. After making the interpolation use the current color mode, those existing expectations needed to be updated to match the corrected behavior.

If you would prefer a source-only PR first, I can adjust the test changes based on your preference.

[ksen0]

Why have the RGB test cases changed?

[davepagurek]

It looks like it's just HSL tests that have modifications here, and the one RGB test in between these if you expand some of the code diffs is unmodified. So it might be ok!

[pupuking723]

Thanks for checking this.

The RGB test between these cases was not changed. The changed expectations are for the HSL/HSB lerpColor() cases, because those existing assertions were encoding the old behavior where interpolation still happened through RGB space.

After the fix, those cases now interpolate in the active color mode, so the expected channel values changed. I agree the diff can look confusing because the unchanged RGB case sits between the modified cases.