[p5.js 2.0+ Bug Report]: p5.Vector.rem() silently ignores negative divisors, a 1.x regression

[rajanpanth]

Most appropriate sub-area of p5.js?

Math

p5.js version

2.x main (4b096e2)

Actual vs expected behavior

The reference for rem() says all components are set to their values modulo the divisor. The 2.x implementation guards each component with > 0, so negative divisors are silently ignored:

new p5.Vector(3, 4, 5).rem(-2);        // actual [3, 4, 5], 1.x gave [1, 0, 1]
new p5.Vector(3, 4, 5).rem([-2, 3, -4]); // actual [3, 1, 5], only y reduced
// plain JS: 3 % -2 === 1

1.x applied % for any nonzero divisor (calculateRemainder2D/3D used if (xComponent !== 0), v1.11.3 line 440), so this is a 2.x regression. The zero-skip (division by zero) is the only case that should be skipped.

Steps to reproduce

Outputs above are from executed runs against current main.

Note

I have a one-line-class fix ready (change both > 0 checks to !== 0) with a "negative divisors" test suite. Verified: with > 0 restored both new tests fail with expected 3 to deeply equal 1; with the fix the vector suite passes 211. Filing for approval per the contributing guide; will open the PR once approved.

[ksen0]

Thanks @rajanpanth please go ahead! Consider Math.abs(x) > 0 rather than x !== 0 unless there is some reason to use !== specifically here. We did also see your other files issues, and in general we recommend new contributors especially to work on 1 PR at a time to get used to the codebase. Sounds like you already went through the contributor guide, please have a look at the AI usage policy as well