#7012 Fix: Can't turn grid off

[andrewvarga]

Fixes #7012

Could be reproduced by having a project which has an import:

import "cube.kcl" as cube
cube
  |> translate(
       %,
       x = 5.14,
       y = 2.14,
       z = 1.14,
     )

Make sure you have the file being imported "cube.kcl" in this case that can contain anything:

sketch001 = startSketchOn(XZ)
profile001 = startProfile(sketch001, at = [-2, 2])
  |> yLine(length = -4)
  |> xLine(length = 4)
  |> yLine(length = 4)
  |> line(endAbsolute = [profileStartX(%), profileStartY(%)])
  |> close()
extrude001 = extrude(profile001, length = 4)

Then:

• toggle "Show scale grid" either via command bar settings or user settingg: this will work
• toggle it again: this will not work
• change anything in the kcl code to cause re-execution (eg. change a number) -> grid will update properly

I was trying to add a test, it was a bit tricky for me at first without using a screenshot based test or without using mocks..

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Preview	Comments	Updated (UTC)
modeling-app	Ready	Preview	Comment	Nov 28, 2025 3:58pm

[andrewvarga]

Previously we never called write_old_ast in this CheckImportsOnly arm of the match after reapply_settings so the value was never updated in the cache, so next time the grid changed it was compared against an outdated cached value.

This is similar code to further below in this function which also calls write_old_ast after reapply_settings.

[codspeed-hq]

CodSpeed Performance Report

Merging #9101 will not alter performance

_{Comparing andrewvarga/7012/fix-cant-turn-grid-off (15f97b0) with main (b7b0f68)¹}

Summary

✅ 143 untouched
🆕 2 new
⏩ 92 skipped²

Benchmarks breakdown

	Mode	Benchmark	BASE	HEAD	Change
🆕	Simulation	parse_crash-box	N/A	33.1 ms	N/A
🆕	Simulation	parse_hex-nut-with-chamfer	N/A	13.6 ms	N/A

Footnotes

1. No successful run was found on main (8506634) during the generation of this report, so b7b0f68 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

2. 92 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩