kicad/libs/kimath/src/convert_basic_shapes_to_pol...

475 lines
17 KiB
C++
Raw Normal View History

/**
* @file convert_basic_shapes_to_polygon.cpp
*/
/*
* This program source code file is part of KiCad, a free EDA CAD application.
*
* Copyright (C) 2018 Jean-Pierre Charras, jp.charras at wanadoo.fr
* Copyright (C) 1992-2021 KiCad Developers, see AUTHORS.txt for contributors.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, you may find one here:
* http://www.gnu.org/licenses/old-licenses/gpl-2.0.html
* or you may search the http://www.gnu.org website for the version 2 license,
* or you may write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
*/
#include <algorithm> // for max, min
#include <math.h> // for atan2
#include <type_traits> // for swap
#include <convert_basic_shapes_to_polygon.h>
#include <geometry/geometry_utils.h>
#include <geometry/shape_line_chain.h> // for SHAPE_LINE_CHAIN
#include <geometry/shape_poly_set.h> // for SHAPE_POLY_SET, SHAPE_POLY_SE...
#include <math/util.h>
#include <math/vector2d.h> // for VECTOR2I
#include <trigo.h>
void TransformCircleToPolygon( SHAPE_LINE_CHAIN& aCornerBuffer, wxPoint aCenter, int aRadius,
int aError, ERROR_LOC aErrorLoc )
{
wxPoint corner_position;
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
int numSegs = GetArcToSegmentCount( aRadius, aError, 360.0 );
// The shape will be built with a even number of segs. Reason: the horizontal
// diameter begins and ends to points on the actual circle, or circle
// expanded by aError if aErrorLoc == ERROR_OUTSIDE.
// This is used by Arc to Polygon shape convert.
if( numSegs & 1 )
numSegs++;
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
int delta = 3600 / numSegs; // rotate angle in 0.1 degree
int radius = aRadius;
if( aErrorLoc == ERROR_OUTSIDE )
{
// The outer radius should be radius+aError
// Recalculate the actual approx error, as it can be smaller than aError
// because numSegs is clamped to a minimal value
int actual_delta_radius = CircleToEndSegmentDeltaRadius( radius, numSegs );
radius += GetCircleToPolyCorrection( actual_delta_radius );
}
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
for( int angle = 0; angle < 3600; angle += delta )
{
corner_position.x = radius;
corner_position.y = 0;
RotatePoint( &corner_position, angle );
corner_position += aCenter;
aCornerBuffer.Append( corner_position.x, corner_position.y );
}
aCornerBuffer.SetClosed( true );
}
void TransformCircleToPolygon( SHAPE_POLY_SET& aCornerBuffer, wxPoint aCenter, int aRadius,
int aError, ERROR_LOC aErrorLoc )
{
wxPoint corner_position;
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
int numSegs = GetArcToSegmentCount( aRadius, aError, 360.0 );
// The shape will be built with a even number of segs. Reason: the horizontal
// diameter begins and ends to points on the actual circle, or circle
// expanded by aError if aErrorLoc == ERROR_OUTSIDE.
// This is used by Arc to Polygon shape convert.
if( numSegs & 1 )
numSegs++;
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
int delta = 3600 / numSegs; // rotate angle in 0.1 degree
int radius = aRadius;
if( aErrorLoc == ERROR_OUTSIDE )
{
// The outer radius should be radius+aError
// Recalculate the actual approx error, as it can be smaller than aError
// because numSegs is clamped to a minimal value
int actual_delta_radius = CircleToEndSegmentDeltaRadius( radius, numSegs );
radius += GetCircleToPolyCorrection( actual_delta_radius );
}
aCornerBuffer.NewOutline();
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
for( int angle = 0; angle < 3600; angle += delta )
{
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
corner_position.x = radius;
corner_position.y = 0;
2018-01-19 20:30:43 +00:00
RotatePoint( &corner_position, angle );
corner_position += aCenter;
aCornerBuffer.Append( corner_position.x, corner_position.y );
}
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
// Finish circle
corner_position.x = radius;
corner_position.y = 0;
corner_position += aCenter;
aCornerBuffer.Append( corner_position.x, corner_position.y );
}
void TransformOvalToPolygon( SHAPE_POLY_SET& aCornerBuffer, wxPoint aStart, wxPoint aEnd,
int aWidth, int aError, ERROR_LOC aErrorLoc )
{
// To build the polygonal shape outside the actual shape, we use a bigger
// radius to build rounded ends.
// However, the width of the segment is too big.
// so, later, we will clamp the polygonal shape with the bounding box
// of the segment.
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
int radius = aWidth / 2;
int numSegs = GetArcToSegmentCount( radius, aError, 360.0 );
int delta = 3600 / numSegs; // rotate angle in 0.1 degree
int correction = GetCircleToPolyCorrection( aError );
if( aErrorLoc == ERROR_OUTSIDE )
radius += correction;
// end point is the coordinate relative to aStart
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
wxPoint endp = aEnd - aStart;
wxPoint startp = aStart;
wxPoint corner;
SHAPE_POLY_SET polyshape;
polyshape.NewOutline();
// normalize the position in order to have endp.x >= 0
// it makes calculations more easy to understand
if( endp.x < 0 )
{
endp = aStart - aEnd;
startp = aEnd;
}
// delta_angle is in radian
double delta_angle = atan2( (double)endp.y, (double)endp.x );
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
int seg_len = KiROUND( EuclideanNorm( endp ) );
// Compute the outlines of the segment, and creates a polygon
// Note: the polygonal shape is built from the equivalent horizontal
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
// segment starting at {0,0}, and ending at {seg_len,0}
// add right rounded end:
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
for( int angle = 0; angle < 1800; angle += delta )
{
corner = wxPoint( 0, radius );
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
RotatePoint( &corner, angle );
corner.x += seg_len;
polyshape.Append( corner.x, corner.y );
}
// Finish arc:
corner = wxPoint( seg_len, -radius );
polyshape.Append( corner.x, corner.y );
// add left rounded end:
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
for( int angle = 0; angle < 1800; angle += delta )
{
corner = wxPoint( 0, -radius );
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
RotatePoint( &corner, angle );
polyshape.Append( corner.x, corner.y );
}
// Finish arc:
corner = wxPoint( 0, radius );
polyshape.Append( corner.x, corner.y );
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
// Now trim the edges of the polygonal shape which will be slightly outside the
// track width.
SHAPE_POLY_SET bbox;
bbox.NewOutline();
// Build the bbox (a horizontal rectangle).
int halfwidth = aWidth / 2; // Use the exact segment width for the bbox height
corner.x = -radius - 2; // use a bbox width slightly bigger to avoid
// creating useless corner at segment ends
corner.y = halfwidth;
bbox.Append( corner.x, corner.y );
corner.y = -halfwidth;
bbox.Append( corner.x, corner.y );
corner.x = radius + seg_len + 2;
bbox.Append( corner.x, corner.y );
corner.y = halfwidth;
bbox.Append( corner.x, corner.y );
// Now, clamp the shape
polyshape.BooleanIntersection( bbox, SHAPE_POLY_SET::PM_STRICTLY_SIMPLE );
// Note the final polygon is a simple, convex polygon with no hole
// due to the shape of initial polygons
// Rotate and move the polygon to its right location
polyshape.Rotate( delta_angle, VECTOR2I( 0, 0 ) );
polyshape.Move( startp );
aCornerBuffer.Append( polyshape);
}
// Return a polygon representing a round rect centered at {0,0}
void TransformRoundRectToPolygon( SHAPE_POLY_SET& aCornerBuffer, const wxSize& aSize,
int aCornerRadius, int aError, ERROR_LOC aErrorLoc )
{
SHAPE_POLY_SET outline;
wxPoint centers[4];
wxSize size( aSize / 2 );
size.x -= aCornerRadius;
size.y -= aCornerRadius;
// Ensure size is > 0, to avoid generating unusable shapes which can crash kicad.
size.x = std::max( 1, size.x );
size.y = std::max( 1, size.y );
centers[0] = wxPoint( -size.x, size.y );
centers[1] = wxPoint( size.x, size.y );
centers[2] = wxPoint( size.x, -size.y );
centers[3] = wxPoint( -size.x, -size.y );
Clean up arc/circle polygonization. 1) For a while now we've been using a calculated seg count from a given maxError, and a correction factor to push the radius out so that all the error is outside the arc/circle. However, the second calculation (which pre-dates the first) is pretty much just the inverse of the first (and yields nothing more than maxError back). This is particularly sub-optimal given the cost of trig functions. 2) There are a lot of old optimizations to reduce segcounts in certain situations, someting that our error-based calculation compensates for anyway. (Smaller radii need fewer segments to meet the maxError condition.) But perhaps more importantly we now surface maxError in the UI and we don't really want to call it "Max deviation except when it's not". 3) We were also clamping the segCount twice: once in the calculation routine and once in most of it's callers. Furthermore, the caller clamping was inconsistent (both in being done and in the clamping value). We now clamp only in the calculation routine. 4) There's no reason to use the correction factors in the 3Dviewer; it's just a visualization and whether the polygonization error is inside or outside the shape isn't really material. 5) The arc-correction-disabling stuff (used for solder mask layer) was somewhat fragile in that it depended on the caller to turn it back on afterwards. It's now only exposed as a RAII object which automatically cleans up when it goes out of scope. 6) There were also bugs in a couple of the polygonization routines where we'd accumulate round-off error in adding up the segments and end up with an overly long last segment (which of course would voilate the error max). This was the cause of the linked bug and also some issues with vias that we had fudged in the past with extra clearance. Fixes https://gitlab.com/kicad/code/kicad/issues/5567
2020-09-10 23:05:20 +00:00
int numSegs = GetArcToSegmentCount( aCornerRadius, aError, 360.0 );
// Choppy corners on rounded-corner rectangles look awful so enforce a minimum of
// 4 segments per corner.
if( numSegs < 16 )
numSegs = 16;
int delta = 3600 / numSegs; // rotate angle in 0.1 degree
int radius = aCornerRadius;
if( aErrorLoc == ERROR_OUTSIDE )
radius += GetCircleToPolyCorrection( aError );
auto genArc =
[&]( const wxPoint& aCenter, int aStart, int aEnd )
2021-04-08 22:50:40 +00:00
{
for( int angle = aStart + delta; angle < aEnd; angle += delta )
{
wxPoint pt( -radius, 0 );
RotatePoint( &pt, angle );
pt += aCenter;
outline.Append( pt.x, pt.y );
}
};
outline.NewOutline();
outline.Append( centers[0] + wxPoint( -radius, 0 ) );
genArc( centers[0], 0, 900 );
outline.Append( centers[0] + wxPoint( 0, radius ) );
outline.Append( centers[1] + wxPoint( 0, radius ) );
genArc( centers[1], 900, 1800 );
outline.Append( centers[1] + wxPoint( radius, 0 ) );
outline.Append( centers[2] + wxPoint( radius, 0 ) );
genArc( centers[2], 1800, 2700 );
outline.Append( centers[2] + wxPoint( 0, -radius ) );
outline.Append( centers[3] + wxPoint( 0, -radius ) );
genArc( centers[3], 2700, 3600 );
outline.Append( centers[3] + wxPoint( -radius, 0 ) );
outline.Outline( 0 ).SetClosed( true );
// The created outlines are bigger than the actual outlines, due to the fact
// the corner radius is bigger than the initial value when building a shape outside the
// actual shape.
// However the bounding box shape does not need to be bigger: only rounded corners must
// be modified.
// So clamp the too big shape by the actual bounding box
if( aErrorLoc == ERROR_OUTSIDE )
{
SHAPE_POLY_SET bbox;
bbox.NewOutline();
wxSize bbox_size = aSize/2;
bbox.Append( wxPoint( -bbox_size.x, -bbox_size.y ) );
bbox.Append( wxPoint( bbox_size.x, -bbox_size.y ) );
bbox.Append( wxPoint( bbox_size.x, bbox_size.y ) );
bbox.Append( wxPoint( -bbox_size.x, bbox_size.y ) );
bbox.Outline( 0 ).SetClosed( true );
outline.BooleanIntersection( bbox, SHAPE_POLY_SET::PM_FAST );
// The result is a convex polygon, no need to simplify or fracture.
}
// Add the outline:
aCornerBuffer.Append( outline );
}
void TransformRoundChamferedRectToPolygon( SHAPE_POLY_SET& aCornerBuffer, const wxPoint& aPosition,
const wxSize& aSize, double aRotation,
int aCornerRadius, double aChamferRatio,
int aChamferCorners, int aError, ERROR_LOC aErrorLoc )
{
SHAPE_POLY_SET outline;
TransformRoundRectToPolygon( outline, aSize, aCornerRadius, aError, aErrorLoc );
if( aChamferCorners )
{
// Now we have the round rect outline, in position 0,0 orientation 0.0.
// Chamfer the corner(s).
int chamfer_value = aChamferRatio * std::min( aSize.x, aSize.y );
SHAPE_POLY_SET chamfered_corner; // corner shape for the current corner to chamfer
int corner_id[4] =
{
RECT_CHAMFER_TOP_LEFT, RECT_CHAMFER_TOP_RIGHT,
RECT_CHAMFER_BOTTOM_LEFT, RECT_CHAMFER_BOTTOM_RIGHT
};
// Depending on the corner position, signX[] and signY[] give the sign of chamfer
// coordinates relative to the corner position
// The first corner is the top left corner, then top right, bottom left and bottom right
int signX[4] = {1, -1, 1,-1 };
int signY[4] = {1, 1, -1,-1 };
for( int ii = 0; ii < 4; ii++ )
{
if( (corner_id[ii] & aChamferCorners) == 0 )
continue;
VECTOR2I corner_pos( -signX[ii]*aSize.x/2, -signY[ii]*aSize.y/2 );
if( aCornerRadius )
{
// We recreate a rectangular area covering the full rounded corner
// (max size = aSize/2) to rebuild the corner before chamfering, to be sure
// the rounded corner shape does not overlap the chamfered corner shape:
chamfered_corner.RemoveAllContours();
chamfered_corner.NewOutline();
chamfered_corner.Append( 0, 0 );
chamfered_corner.Append( 0, signY[ii] * aSize.y / 2 );
chamfered_corner.Append( signX[ii] * aSize.x / 2, signY[ii] * aSize.y / 2 );
chamfered_corner.Append( signX[ii] * aSize.x / 2, 0 );
chamfered_corner.Move( corner_pos );
outline.BooleanAdd( chamfered_corner, SHAPE_POLY_SET::PM_STRICTLY_SIMPLE );
}
// Now chamfer this corner
chamfered_corner.RemoveAllContours();
chamfered_corner.NewOutline();
chamfered_corner.Append( 0, 0 );
chamfered_corner.Append( 0, signY[ii] * chamfer_value );
chamfered_corner.Append( signX[ii] * chamfer_value, 0 );
chamfered_corner.Move( corner_pos );
outline.BooleanSubtract( chamfered_corner, SHAPE_POLY_SET::PM_STRICTLY_SIMPLE );
}
}
// Rotate and move the outline:
if( aRotation != 0.0 )
outline.Rotate( DECIDEG2RAD( -aRotation ), VECTOR2I( 0, 0 ) );
outline.Move( VECTOR2I( aPosition ) );
// Add the outline:
aCornerBuffer.Append( outline );
}
void TransformArcToPolygon( SHAPE_POLY_SET& aCornerBuffer, wxPoint aStart, wxPoint aMid,
wxPoint aEnd, int aWidth, int aError, ERROR_LOC aErrorLoc )
{
SHAPE_ARC arc( aStart, aMid, aEnd, aWidth );
SHAPE_LINE_CHAIN arcSpine = arc.ConvertToPolyline( aError );
int radial_offset = ( aWidth + 1 ) / 2;
SHAPE_POLY_SET polyshape;
std::vector<VECTOR2I> outside_pts;
/// We start by making rounded ends on the arc
TransformCircleToPolygon( polyshape,
wxPoint( arcSpine.GetPoint( 0 ).x, arcSpine.GetPoint( 0 ).y ), radial_offset, aError,
aErrorLoc );
TransformCircleToPolygon( polyshape,
wxPoint( arcSpine.GetPoint( -1 ).x, arcSpine.GetPoint( -1 ).y ), radial_offset, aError,
aErrorLoc );
// The circle polygon is built with a even number of segments, so the
// horizontal diameter has 2 corners on the biggest diameter
// Rotate these 2 corners to match the start and ens points of inner and outer
// end points of the arc appoximation outlines, build below.
// The final shape is much better.
double arc_angle_end_deg = arc.GetStartAngle();
2021-06-26 17:44:18 +00:00
if( arc_angle_end_deg != 0 && arc_angle_end_deg != 180.0 )
polyshape.Outline(0).Rotate( arc_angle_end_deg * M_PI/180.0, arcSpine.GetPoint( 0 ) );
arc_angle_end_deg = arc.GetEndAngle();
2021-06-26 17:44:18 +00:00
if( arc_angle_end_deg != 0 && arc_angle_end_deg != 180.0 )
polyshape.Outline(1).Rotate( arc_angle_end_deg * M_PI/180.0, arcSpine.GetPoint( -1 ) );
if( aErrorLoc == ERROR_OUTSIDE )
radial_offset += aError;
else
radial_offset -= aError/2;
if( radial_offset < 0 )
radial_offset = 0;
polyshape.NewOutline();
VECTOR2I center = arc.GetCenter();
int radius = ( arc.GetP0() - center ).EuclideanNorm();
int last_index = arcSpine.GetPointCount() -1;
for( std::size_t ii = 0; ii <= last_index; ++ii )
{
VECTOR2I offset = arcSpine.GetPoint( ii ) - center;
int curr_rd = radius;
// This correction gives a better position of intermediate points of the sides of arc.
if( ii > 0 && ii < last_index )
curr_rd += aError/2;
polyshape.Append( offset.Resize( curr_rd - radial_offset ) + center );
outside_pts.emplace_back( offset.Resize( curr_rd + radial_offset ) + center );
}
for( auto it = outside_pts.rbegin(); it != outside_pts.rend(); ++it )
polyshape.Append( *it );
// Can be removed, but usefull to display the outline:
polyshape.Simplify( SHAPE_POLY_SET::PM_FAST );
aCornerBuffer.Append( polyshape );
}
void TransformRingToPolygon( SHAPE_POLY_SET& aCornerBuffer, wxPoint aCentre, int aRadius,
int aWidth, int aError, ERROR_LOC aErrorLoc )
{
int inner_radius = aRadius - ( aWidth / 2 );
int outer_radius = inner_radius + aWidth;
if( inner_radius <= 0 )
{ //In this case, the ring is just a circle (no hole inside)
TransformCircleToPolygon( aCornerBuffer, aCentre, aRadius + ( aWidth / 2 ), aError,
aErrorLoc );
return;
}
SHAPE_POLY_SET buffer;
TransformCircleToPolygon( buffer, aCentre, outer_radius, aError, aErrorLoc );
// Build the hole:
buffer.NewHole();
TransformCircleToPolygon( buffer.Hole( 0, 0 ), aCentre, inner_radius, aError, aErrorLoc );
2018-01-19 20:30:43 +00:00
buffer.Fracture( SHAPE_POLY_SET::PM_FAST );
aCornerBuffer.Append( buffer );
}