Geometry Nodes: Port the trim curve node to the new data-block

The trim functionality is implemented in the geometry module, and
generalized a bit to be potentially useful for bisecting in the future.
The implementation is based on a helper type called `IndexRangeCyclic`
which allows iteration over all control points between two points on a
curve.

Catmull Rom curves are now supported-- trimmed without resampling first.
However, maintaining the exact shape is not possible. NURBS splines are
still converted to polylines using the evaluated curve concept.

Performance is equivalent or faster then a 3.1 build with regards to
node timings. Compared to 3.3 and 3.2, it's easy to observe test cases
where the node is at least 3 or 4 times faster.

Differential Revision: https://developer.blender.org/D14481

3 files changed, 55 insertions, 0 deletions

diff --git a/source/blender/blenlib/BLI_array_utils.hh b/source/blender/blenlib/BLI_array_utils.hh
new file mode 100644
index 00000000000..dd65147a926
--- /dev/null
+++ b/source/blender/blenlib/BLI_array_utils.hh
@@ -0,0 +1,35 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+
+#pragma once
+
+#include "BLI_generic_span.hh"
+#include "BLI_generic_virtual_array.hh"
+#include "BLI_index_mask.hh"
+#include "BLI_task.hh"
+
+namespace blender::array_utils {
+
+/**
+ * Fill the destination span by copying masked values from the src array. Threaded based on
+ * grainsize.
+ */
+void copy(const GVArray &src, IndexMask selection, GMutableSpan dst, int64_t grain_size = 4096);
+
+/**
+ * Fill the destination span by copying values from the src array. Threaded based on
+ * grainsize.
+ */
+template<typename T>
+inline void copy(const Span<T> src,
+                 const IndexMask selection,
+                 MutableSpan<T> dst,
+                 const int64_t grain_size = 4096)
+{
+  threading::parallel_for(selection.index_range(), grain_size, [&](const IndexRange range) {
+    for (const int64_t index : selection.slice(range)) {
+      dst[index] = src[index];
+    }
+  });
+}
+
+}  // namespace blender::array_utils
diff --git a/source/blender/blenlib/CMakeLists.txt b/source/blender/blenlib/CMakeLists.txt
index 4a635e34205..470ffebcad4 100644
--- a/source/blender/blenlib/CMakeLists.txt
+++ b/source/blender/blenlib/CMakeLists.txt
@@ -47,6 +47,7 @@ set(SRC
   intern/array_store.c
   intern/array_store_utils.c
   intern/array_utils.c
+  intern/array_utils.cc
   intern/astar.c
   intern/bitmap.c
   intern/bitmap_draw_2d.c
@@ -164,6 +165,7 @@ set(SRC
   BLI_array_store.h
   BLI_array_store_utils.h
   BLI_array_utils.h
+  BLI_array_utils.hh
   BLI_asan.h
   BLI_assert.h
   BLI_astar.h
diff --git a/source/blender/blenlib/intern/array_utils.cc b/source/blender/blenlib/intern/array_utils.cc
new file mode 100644
index 00000000000..d4266295944
--- /dev/null
+++ b/source/blender/blenlib/intern/array_utils.cc
@@ -0,0 +1,18 @@
+#include "BLI_array_utils.hh"
+#include "BLI_task.hh"
+
+namespace blender::array_utils {
+
+void copy(const GVArray &src,
+          const IndexMask selection,
+          GMutableSpan dst,
+          const int64_t grain_size)
+{
+  BLI_assert(src.type() == dst.type());
+  BLI_assert(src.size() == dst.size());
+  threading::parallel_for(selection.index_range(), grain_size, [&](const IndexRange range) {
+    src.materialize_to_uninitialized(selection.slice(range), dst.data());
+  });
+}
+
+}  // namespace blender::array_utils