Negative dimensions will mirror rect() again

[martinleopold]

Resolves #7353

Changes:

Fix a regression with negative values for width and/or height on rect(), ellipse(), and arc()

Screenshots of the change:

N/A

PR Checklist

• npm run lint passes
• Inline documentation is included / updated
• Unit tests are included / updated

[martinleopold]

This resolves the immediate issue by adjusting the position of the shape when using negative dimensions.
I'm wondering if we need to look into what happens with textured shapes. In this case the texture should actually be flipped i.e. properly mirrored, I guess?

[perminder-17]

Thank you so much for your hard work on this! I truly appreciate the effort you've put in. There is one potential issue I wanted to mention regarding texture image handling. While you’ve done a great job adjusting the positioning in CORNERS, CORNER, etc., it seems that the image itself isn’t being flipped when necessary—only the position is being shifted. Adding a flip where needed could enhance the solution. Please let us know if there’s any way we can assist in exploring or addressing this aspect.

Expected:

Your's:

[martinleopold]

@perminder-17 Thanks for pointing that out, will look into it.

[martinleopold]

@perminder-17 I think my latest commit should restore the original behavior. This means that rect() will mirror, while ellipse() and arc() will NOT.

[limzykenneth]

@martinleopold Looks good. Thanks!