When we replace a symbol with one that has fewer pins, the old pins
are released, which leaves points to them in the connection graph
dangling. This updates the pointer to use the cloned copy in the undo
stack until the connection graph is rebuilt with the new data
Fixes https://gitlab.com/kicad/code/kicad/-/issues/17851
Ignore bus member connection subgraphs. They do not have a valid net
name nor do they contain schematic items. This prevents empty nodes
from being added to the tree. They can be reintroduced in the future
if someone wants to pursue it.
Freeze the wxTreeCtrl while populating it and thaw when done to prevent
any unnecessary repainting.
Add profiling to test how long it takes to rebuild the net navigator.
The recently added populate the navigator with all nets when no net
is highlighted has exposed some potential performances issues with some
versions of wxWidgets on certain platforms. Namely wxWidgets 3.2.4 on
Linux GTK.
Fix an issue where a sheet name change would not update the highlighted
net navigator resulting in a stale human readable sheet path.
Prevent the highlighted net navigator from being rebuilt twice when
loading a schematic. SCH_EDIT_FRAME::RefreshNetNavigator() was being
called from both SCH_EDIT_FRAME::UpdateHierarchyNavigator() and
SCH_EDIT_FRAME::RecalculateConnectivity().
Add a new trace helper "KICAD_UI_PROFILE" to show trace output when
profiling user interface performance. It's used in the net navigator
profiling mentioned above.
Reuse PROF_TIMER::Show() to generate string for PROF_TIMER::to_string().
C++20 under MSVC does not like seeing enum values between two enums getting operated on.
static cast to int fixes it for what we are trying to do anyway
Refactors `SHAPE_POLY_SET::fractureSingle()` to be more efficient, while
not changing the actual algorithm:
* increase cache locality by using contiguous arrays instead of what was
effectively a linked list
* reduce latency and jitter by replacing per-edge allocator calls with
ahead-of-time std::vector reserves
* increase cache efficiency by making the vertex struct smaller
* replace O(n^2) leftmost edge search with O(n log n) std::sort
* sort the polygons instead of the edges
* cut iteration count in half in the remaining O(polygons * edges) part
Logging is useful when we find an area that cannot be triangulated.
This will be used to generated test cases.
Skipping minor untesselated areas means that the polygon will still be
considered fully tesselated (and not sent back again and again) even if
the tesselation misses an area less than the configured limit.
Currently, this is 31^2nm.
Jeff Young
RosyDev
Seth Hillbrand
Alex Shvartzkop
Wayne Stambaugh
Mike Williams
jean-pierre charras
Marek Roszko
Jon Evans
Ian McInerney
Céleste Wouters