Add tol to cut and intersect

[adam-urbanczyk]

Adds fuzzy cut and intersect

[codecov]

Codecov Report

Merging #1332 (1920135) into master (3c576f4) will increase coverage by 0.00%.
The diff coverage is 100.00%.

❗ Current head 1920135 differs from pull request most recent head e3fbbe5. Consider uploading reports for the commit e3fbbe5 to get more accurate results

@@           Coverage Diff           @@
##           master    #1332   +/-   ##
=======================================
  Coverage   94.29%   94.30%           
=======================================
  Files          26       26           
  Lines        5575     5583    +8     
  Branches      950      954    +4     
=======================================
+ Hits         5257     5265    +8     
  Misses        187      187           
  Partials      131      131           

Impacted Files	Coverage Δ
cadquery/cq.py	92.23% <100.00%> (ø)
cadquery/occ_impl/shapes.py	92.85% <100.00%> (+0.04%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[jmwright]

One thing that I think is expected behavior but seems like it could confuse users, is that setting the tolerance too high will lead to no resulting geometry, with no error. The following code shows nothing in CQ-editor, but does not throw an error either.

import cadquery as cq

box1 = cq.Workplane().box(10, 10, 10)
box2 = cq.Workplane().center(5, 5).box(10, 10, 10)

res = box2.intersect(box1, tol=3.0)

show_object(res)

That is a contrived example, but it is not hard to imagine that it will happen with lower tolerances with complex geometry. There have been a few models in the KiCAD library that had fuse glitches. In those cases the issue was solved by shifting parts and/or breaking sharp edges between geometries, but I think there is potential there for the same kinds of issues with cut and intersect. I don't have a solution except maybe to note that setting the tolerance too high may result in incorrect or missing geometry.

[jmwright]

I made one comment, but the code looks good to me and works as I expected when experimenting with it in CQ-editor. Thanks @adam-urbanczyk !

[lorenzncode]

LGTM. I've done some manual testing and did not find any problem.