i_shape provides compact, generic data structures and utilities for 2D polygon data. It includes:
- integer and floating-point path, contour, shape, and shapes aliases;
- flat contour and shape buffers for allocation-efficient geometry pipelines;
- area, winding, convexity, simplification, despiking, and deduplication helpers;
- conversion between floating-point and integer coordinates through
i_float; - a common
ShapeResourceinterface for accepting contours, shapes, and flat buffers.
The crate is no_std, uses alloc, and supports i16, i32, and i64 integer coordinates.
[dependencies]
i_shape = "4.0"Enable serde when serialized geometry or flat buffers are required:
[dependencies]
i_shape = { version = "4.0", features = ["serde"] }use i_shape::int::area::Area;
use i_shape::int::path::ContourExtension;
use i_shape::int_path;
let contour = int_path![[0_i32, 0], [10, 0], [10, 10], [0, 10]];
assert_eq!(contour.area_two(), 200_i64);
assert!(!contour.is_clockwise_ordered());
assert!(contour.contains_point(i_shape::int::IntPoint::new(5, 5)));Integer operations use the wide type associated with the coordinate type. See the
i_float coordinate-range documentation
for the arithmetic preconditions that apply to geometric predicates.
Flat buffers store points contiguously and describe contours and shapes with index ranges. This is useful when passing geometry between algorithms without rebuilding nested Vecs.
use i_shape::flat::buffer::FlatContoursBuffer;
use i_shape::int_path;
let contour = int_path![[0_i32, 0], [4, 0], [4, 4], [0, 4]];
let mut buffer = FlatContoursBuffer::default();
buffer.add_contour(&contour);
assert_eq!(buffer.ranges, [0..4]);
assert_eq!(buffer.to_contours(), vec![contour]);Floating-point equivalents are available as FloatFlatContoursBuffer<P> and FloatFlatShapesBuffer<P>, where P can be any FloatPointCompatible type such as [f32; 2] or [f64; 2].
Licensed under the MIT License. See the license file.