This is a fix of a bug, which was in Slic3r forever, where raw slices
were not cached, but recalculated from classified regions, where
merging the regions did not produce the original contour reliably.
Fixes [2.3.0-beta2] Odd bad slicing related to infill (?) percentage #5407
Layer newly remembers bounding boxes of slices,
the bounding boxes are used by G-code generator & newly the support
generator.
Slices are stored as ExPolygons, not ExPolygonCollection.
"Multimaterial printer switches filament at the wrong time during a print"
https://github.com/prusa3d/Slic3r/issues/607
There was a single layer between the raft top and the object first layer
missing on the wipe tower, and after this missing layer all the tool
changes were shifted by one layer, meaning two color print had the colors
switched.
discover_horizontal_shells() fron Perl to C++, where
the already calculated bridge direction was being lost.
Improved constness of the debug methods
void export_region_slices_to_svg(const char *path) const;
void export_region_fill_surfaces_to_svg(const char *path) const;
When extruding supports, the support is interleaved with interface
if possible (when extruded with the same extruder).
Otherwise the base is extruded first.
These could be calculated from the fill areas if needed.
On the other side, the non-classified (non-split) fill areas are stored
now for use in the "ensure vertical wall thickness" feature,
also the non-split fill areas are re-used when recalculating the infills.
This is safer than trying to stitch the fill region together from the
classified fragments.
Modified the "ensure vertical wall thickness" feature to use the non-split
fill areas instead of perimeter areas for the calculation
of non-supported regions. This is cheaper as the fill areas contain
roughly half the edges.
Removed dependency on Perl Math::PlanePath module.
Fixed compilation with Visual Studio and SLIC3R_DEBUG: Visual Studio older than 2015 does not support the prinf type specifier %zu. Use %Iu instead.
C++11 move semantics enabled.
