Algorithm + manifold and accelerator passthrough in clipping functions #273

asinghvi17 · 2025-03-04T01:59:42Z

Spun out from #259
Parent: #275 (adaptive predicates)
Child PRs: #274 (trees)

Create a FosterHormannClipping(manifold, accelerator)algorithm - we can bikeshed the name
Pass the algorithm, manifold, and accelerator (will be used in Tree based acceleration for polygon clipping / boolean ops #274) all the way through to the base level methods which we can define different dispatches for.
Add a tracing error type TracingError that actually stores all the context including a_list and b_list, this way we can more effectively diagnose what's going on.

asinghvi17 · 2025-03-04T15:39:01Z

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

Tree based acceleration for polygon clipping / boolean ops #274 : 3 dependent PRs (#278 , #283 , #285 )
Algorithm + manifold and accelerator passthrough in clipping functions #273 👈 (View in Graphite)
Convert the orient predicate to AdaptivePredicates + minor predicate changes #275
Add a TGGeometry extension for fast inexact geometric predicates #271
[Core] Add an algorithm interface + StableTasks #269 : 1 other dependent PR (#282 )
main

This stack of pull requests is managed by Graphite. Learn more about stacking.

rafaqz

LGTM, but have to remove line breaks between methods of the same function 😆

rafaqz · 2025-03-08T00:23:26Z

src/methods/clipping/clipping_processor.jl

+struct DoubleSTRtree <: IntersectionAccelerator end
+struct SingleNaturalTree <: IntersectionAccelerator end
+struct DoubleNaturalTree <: IntersectionAccelerator end
+struct ThinnedDoubleNaturalTree <: IntersectionAccelerator end


Is this the only one with thinning?

Wondering if it can be a type parameter

rafaqz · 2025-03-08T00:24:07Z

src/methods/clipping/clipping_processor.jl

+"""
+    FosterHormannClipping{M <: Manifold, A <: Union{Nothing, Accelerator}} <: GeometryOpsCore.Algorithm{M} 
+
+A type that represents the Foster-Hormann clipping algorithm.


Suggested change

A type that represents the Foster-Hormann clipping algorithm.

Applies the Foster-Hormann clipping algorithm.

rafaqz · 2025-03-08T00:24:41Z

src/methods/clipping/clipping_processor.jl

+
+

Suggested change

rafaqz · 2025-03-08T00:29:57Z

src/methods/clipping/cut.jl

-    _cut(T, GI.trait(geom), geom, GI.trait(line), line; exact = True())
+cut(geom, line, ::Type{T} = Float64) where {T <: AbstractFloat} = cut(FosterHormannClipping(), geom, line, T)
+cut(m::Manifold, geom, line, ::Type{T} = Float64) where {T <: AbstractFloat} = cut(FosterHormannClipping(m), geom, line, T)
+


Suggested change

rafaqz · 2025-03-08T00:30:22Z

src/methods/clipping/difference.jl

 ) where {T<:AbstractFloat}
    return _difference(
-        TraitTarget(target), T, GI.trait(geom_a), geom_a, GI.trait(geom_b), geom_b;
+        alg, TraitTarget(target), T, GI.trait(geom_a), geom_a, GI.trait(geom_b), geom_b;
        exact = True(), kwargs...,
    )
 end



Suggested change

rafaqz · 2025-03-08T00:35:01Z

src/methods/clipping/union.jl

+) where T
+    return UnionIntersectingPolygons(#=TODO: alg.manifold=#)(GI.MultiPolygon([poly_a, poly_b]))
+end
+


Suggested change

rafaqz · 2025-03-08T00:35:12Z

src/methods/clipping/union.jl

+        return GI.MultiPolygon(res)
+    end
+end
+


Suggested change

rafaqz · 2025-03-08T00:35:21Z

src/methods/clipping/union.jl

+) where T 
+    union(alg, poly_b, multipoly_a; target, fix_multipoly, kwargs...)
+end
+


Suggested change

rafaqz · 2025-03-08T00:35:31Z

src/methods/clipping/union.jl

+) where T
+    return GI.MultiPolygon(_union(alg, TraitTarget{GI.PolygonTrait}(), T, trait_a, multipoly_a, trait_b, multipoly_b; fix_multipoly, kwargs...))
+end
+


Suggested change

rafaqz · 2025-03-08T00:37:02Z

src/methods/geom_relations/geom_geom_processors.jl

@@ -526,6 +526,11 @@ function _point_filled_curve_orientation(
    return iseven(k) ? out : in
 end



Suggested change

This was factored out of the "dev branch" #259 and contains the subset of changes that apply to GeometryOpsCore, for easier review. Child PRs: #271 (TGGeometry) -> #275 (AdaptivePredicates) -> #273 (clipping algorithm type) -> #274 (trees) - Use [StableTasks.jl](https://github.com/JuliaFolds2/StableTasks.jl) in apply and applyreduce - its type-stable tasks save us some allocations! - Remove `Base.@assume_effects` on the low level functions, which caused issues on Julia v1.11 and was probably incorrect anyway - Add an algorithm interface with an abstract supertype `Algorithm{M <: Manifold}`, as discussed in #247. Also adds an abstract Operator supertype and some discussion in code comments, but no implementation or interface surface there yet. - Split out `types.jl` into a directory `types` with a bunch of files in it, for ease of readability / docs / use. - (out of context change): refactor CI a bit for cleanliness. TODOs for later (not this PR): - [ ] Add a `format` method that takes in an incompletely specified algorithm and some geometry as input, and returns a completely specified algorithm. What does this mean? Imagine I call `GO.intersection(FosterHormannClipping(), geom1, geom2)`. That `FosterHormannClipping()` should get expanded to `FosterHormannClipping(AutoAlgorithm(), AutoAccelerator())`. Then, `format` will take `format(alg, args...)` and: - get the `crstrait` of the two geometries, scan for incompatibilities, assign the correct manifold to the algorithm (maybe warn or emit debug info) - if no geometries available, get the manifold via `best_manifold(::Algorithm)`. - maybe inflate the accelerator by checking `npoint` and later preparations to see what's most efficient, maybe not - depends on what we want!

Not sure of name (too verbose, maybe Foster()?). Intersection accelerators will come into the picture later (next PR) but we need them now to instantiate the struct.

Better than printing out since we can (A) check on it and (B) access the polygons as well as the a_list and b_list in Julia. I have a few plot recipes for them but not sure if they are worth putting in a Makie extension...they would also be inaccessible there. Maybe in GeoMakie? Then we can make full use of customizations etc.

This is a shorter change list as a sampler, to show what needs to happen.

asinghvi17 changed the base branch from main to as/clipping-changes March 4, 2025 01:59

asinghvi17 changed the base branch from as/clipping-changes to as/adaptivepredicates March 4, 2025 02:00

asinghvi17 force-pushed the as/adaptivepredicates branch from 60eabb9 to b52ef6c Compare March 4, 2025 02:46

asinghvi17 force-pushed the as/clipping branch 2 times, most recently from 941f27d to bc9a491 Compare March 4, 2025 15:38

asinghvi17 force-pushed the as/adaptivepredicates branch from b52ef6c to f2d90a8 Compare March 4, 2025 15:38

rafaqz approved these changes Mar 8, 2025

View reviewed changes

asinghvi17 force-pushed the as/adaptivepredicates branch from f2d90a8 to 974a162 Compare March 9, 2025 02:45

asinghvi17 force-pushed the as/clipping branch from bc9a491 to 5810db3 Compare March 9, 2025 02:45

asinghvi17 mentioned this pull request Mar 9, 2025

Prepared geometry #278

Open

asinghvi17 force-pushed the as/adaptivepredicates branch from 974a162 to 5b0c4a1 Compare March 11, 2025 03:24

asinghvi17 force-pushed the as/clipping branch from 5810db3 to e0f32f3 Compare March 11, 2025 03:24

asinghvi17 mentioned this pull request Mar 18, 2025

[wip] antimeridian cutting with trees for regridding #283

Draft

asinghvi17 force-pushed the as/adaptivepredicates branch from 5b0c4a1 to c36826d Compare April 3, 2025 20:10

asinghvi17 force-pushed the as/clipping branch from b42e1b3 to 38d0eac Compare April 3, 2025 20:10

asinghvi17 mentioned this pull request Apr 3, 2025

experimental UnitSpherical module that can express geometry on unit sphere #285

Open

asinghvi17 force-pushed the as/adaptivepredicates branch from c36826d to 62187d6 Compare April 3, 2025 20:16

asinghvi17 force-pushed the as/clipping branch from 38d0eac to a5377d4 Compare April 3, 2025 20:17

asinghvi17 added 6 commits April 3, 2025 22:24

Describe a FosterHormannClipping algorithm

7f92de9

Not sure of name (too verbose, maybe Foster()?). Intersection accelerators will come into the picture later (next PR) but we need them now to instantiate the struct.

Pass algorithm all the way through cut

4011dd1

This is a shorter change list as a sampler, to show what needs to happen.

Pass algorithm through all functions in clipping_processor

72c7657

Pass manifold through point_filled_curve_orientation in planar

164ff57

Pass algorithm through difference + union

6f360cd

asinghvi17 added 3 commits April 3, 2025 22:24

use TracingError instead of printing out errors

1b3befa

Implement algorithm in intersection

c976cb2

Fix typo that prevented multipolygon intersection from taking place

b3b8d4b

asinghvi17 force-pushed the as/clipping branch from a5377d4 to b3b8d4b Compare April 4, 2025 02:24

asinghvi17 force-pushed the as/adaptivepredicates branch from 62187d6 to e29badb Compare April 4, 2025 02:24

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Algorithm + manifold and accelerator passthrough in clipping functions #273

Algorithm + manifold and accelerator passthrough in clipping functions #273

asinghvi17 commented Mar 4, 2025 •

edited

Loading

asinghvi17 commented Mar 4, 2025 •

edited

Loading

rafaqz left a comment

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

rafaqz Mar 8, 2025

	A type that represents the Foster-Hormann clipping algorithm.
	Applies the Foster-Hormann clipping algorithm.

		@@ -526,6 +526,11 @@ function _point_filled_curve_orientation(
		return iseven(k) ? out : in
		end

Algorithm + manifold and accelerator passthrough in clipping functions #273

Are you sure you want to change the base?

Algorithm + manifold and accelerator passthrough in clipping functions #273

Conversation

asinghvi17 commented Mar 4, 2025 • edited Loading

asinghvi17 commented Mar 4, 2025 • edited Loading

rafaqz left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

asinghvi17 commented Mar 4, 2025 •

edited

Loading

asinghvi17 commented Mar 4, 2025 •

edited

Loading