kicad/pcbnew/board_items_to_polygon_shap...

703 lines
24 KiB
C++
Raw Normal View History

2014-10-23 17:53:38 +00:00
/*
* This program source code file is part of KiCad, a free EDA CAD application.
*
* Copyright (C) 2009-2018 Jean-Pierre Charras, jp.charras at wanadoo.fr
* Copyright (C) 1992-2019 KiCad Developers, see AUTHORS.txt for contributors.
2014-10-23 17:53:38 +00:00
*
* 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 <vector>
#include <bezier_curves.h>
#include <base_units.h> // for IU_PER_MM
#include <pcbnew.h>
2018-01-29 20:58:58 +00:00
#include <pcb_edit_frame.h>
#include <trigo.h>
2014-05-17 19:29:15 +00:00
#include <class_board.h>
#include <class_pad.h>
#include <class_track.h>
#include <class_drawsegment.h>
#include <class_pcb_text.h>
2012-01-29 19:29:19 +00:00
#include <class_zone.h>
#include <class_module.h>
#include <class_edge_mod.h>
#include <convert_basic_shapes_to_polygon.h>
#include <geometry/geometry_utils.h>
#include <geometry/shape_segment.h>
#include <math/util.h> // for KiROUND
// A helper struct for the callback function
// These variables are parameters used in addTextSegmToPoly.
// But addTextSegmToPoly is a call-back function,
// so we cannot send them as arguments.
struct TSEGM_2_POLY_PRMS {
int m_textWidth;
int m_error;
SHAPE_POLY_SET* m_cornerBuffer;
};
TSEGM_2_POLY_PRMS prms;
// This is a call back function, used by GRText to draw the 3D text shape:
static void addTextSegmToPoly( int x0, int y0, int xf, int yf, void* aData )
{
TSEGM_2_POLY_PRMS* prm = static_cast<TSEGM_2_POLY_PRMS*>( aData );
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
TransformOvalToPolygon( *prm->m_cornerBuffer, wxPoint( x0, y0 ), wxPoint( xf, yf ),
prm->m_textWidth, prm->m_error );
}
void BOARD::ConvertBrdLayerToPolygonalContours( PCB_LAYER_ID aLayer, SHAPE_POLY_SET& aOutlines )
2014-05-17 19:29:15 +00:00
{
// convert tracks and vias:
2019-05-31 02:30:28 +00:00
for( auto track : m_tracks )
2014-05-17 19:29:15 +00:00
{
if( !track->IsOnLayer( aLayer ) )
continue;
track->TransformShapeWithClearanceToPolygon( aOutlines, aLayer, 0 );
2014-05-17 19:29:15 +00:00
}
// convert pads
for( auto module : m_modules )
2014-05-17 19:29:15 +00:00
{
module->TransformPadsShapesWithClearanceToPolygon( aLayer, aOutlines, 0 );
2014-05-17 19:29:15 +00:00
// Micro-wave modules may have items on copper layers
module->TransformGraphicShapesWithClearanceToPolygonSet( aLayer, aOutlines, 0 );
2014-05-17 19:29:15 +00:00
}
// convert copper zones
for( ZONE_CONTAINER* zone : Zones() )
2014-05-17 19:29:15 +00:00
{
if( zone->GetLayerSet().test( aLayer ) )
zone->TransformSolidAreasShapesToPolygon( aLayer, aOutlines );
2014-05-17 19:29:15 +00:00
}
// convert graphic items on copper layers (texts)
for( auto item : m_drawings )
2014-05-17 19:29:15 +00:00
{
if( !item->IsOnLayer( aLayer ) )
continue;
switch( item->Type() )
{
case PCB_LINE_T:
( (DRAWSEGMENT*) item )->TransformShapeWithClearanceToPolygon( aOutlines, aLayer, 0 );
2014-05-17 19:29:15 +00:00
break;
case PCB_TEXT_T:
( (TEXTE_PCB*) item )->TransformShapeWithClearanceToPolygonSet( aOutlines, 0 );
2014-05-17 19:29:15 +00:00
break;
default:
break;
}
}
}
void MODULE::TransformPadsShapesWithClearanceToPolygon( PCB_LAYER_ID aLayer,
SHAPE_POLY_SET& aCornerBuffer,
int aInflateValue, int aMaxError,
bool aSkipNPTHPadsWihNoCopper,
bool aSkipPlatedPads,
bool aSkipNonPlatedPads ) const
{
for( D_PAD* pad : m_pads )
{
if( aLayer != UNDEFINED_LAYER && !pad->IsOnLayer(aLayer) )
continue;
if( !pad->FlashLayer( aLayer ) && IsCopperLayer( aLayer ) )
continue;
// NPTH pads are not drawn on layers if the shape size and pos is the same
// as their hole:
2020-09-30 15:38:35 +00:00
if( aSkipNPTHPadsWihNoCopper && pad->GetAttribute() == PAD_ATTRIB_NPTH )
{
if( pad->GetDrillSize() == pad->GetSize() && pad->GetOffset() == wxPoint( 0, 0 ) )
{
switch( pad->GetShape() )
{
case PAD_SHAPE_CIRCLE:
if( pad->GetDrillShape() == PAD_DRILL_SHAPE_CIRCLE )
continue;
break;
case PAD_SHAPE_OVAL:
if( pad->GetDrillShape() != PAD_DRILL_SHAPE_CIRCLE )
continue;
break;
default:
break;
}
}
}
const bool isPlated = ( ( aLayer == F_Cu ) && pad->FlashLayer( F_Mask ) ) ||
( ( aLayer == B_Cu ) && pad->FlashLayer( B_Mask ) );
if( aSkipPlatedPads && isPlated )
continue;
if( aSkipNonPlatedPads && !isPlated )
continue;
wxSize clearance( aInflateValue, aInflateValue );
switch( aLayer )
{
case F_Mask:
case B_Mask:
clearance.x += pad->GetSolderMaskMargin();
clearance.y += pad->GetSolderMaskMargin();
break;
case F_Paste:
case B_Paste:
clearance += pad->GetSolderPasteMargin();
break;
default:
break;
}
// Our standard TransformShapeWithClearanceToPolygon() routines can't handle differing
// x:y clearance values (which get generated when a relative paste margin is used with
// an oblong pad). So we apply this huge hack and fake a larger pad to run the transform
// on.
// Of course being a hack it falls down when dealing with custom shape pads (where the
// size is only the size of the anchor), so for those we punt and just use clearance.x.
if( ( clearance.x < 0 || clearance.x != clearance.y )
&& pad->GetShape() != PAD_SHAPE_CUSTOM )
{
D_PAD dummy( *pad );
dummy.SetSize( pad->GetSize() + clearance + clearance );
dummy.TransformShapeWithClearanceToPolygon( aCornerBuffer, aLayer, 0, aMaxError );
}
else
{
pad->TransformShapeWithClearanceToPolygon( aCornerBuffer, aLayer, clearance.x,
aMaxError );
}
}
}
/**
* Generate shapes of graphic items (outlines) as polygons added to a buffer.
* @aCornerBuffer = the buffer to store polygons
* @aInflateValue = a value to inflate shapes
* @aError = the maximum error to allow when approximating curves with segments
* @aIncludeText = indicates footprint text items (reference, value, etc.) should be included
* in the outline
*/
void MODULE::TransformGraphicShapesWithClearanceToPolygonSet( PCB_LAYER_ID aLayer,
SHAPE_POLY_SET& aCornerBuffer,
int aInflateValue,
int aError,
bool aIncludeText,
bool aIncludeEdges ) const
{
std::vector<TEXTE_MODULE *> texts; // List of TEXTE_MODULE to convert
for( auto item : GraphicalItems() )
{
if( item->Type() == PCB_MODULE_TEXT_T && aIncludeText )
{
TEXTE_MODULE* text = static_cast<TEXTE_MODULE*>( item );
if( aLayer != UNDEFINED_LAYER && text->GetLayer() == aLayer && text->IsVisible() )
texts.push_back( text );
}
if( item->Type() == PCB_MODULE_EDGE_T && aIncludeEdges )
{
EDGE_MODULE* outline = (EDGE_MODULE*) item;
if( aLayer != UNDEFINED_LAYER && outline->GetLayer() == aLayer )
outline->TransformShapeWithClearanceToPolygon( aCornerBuffer, aLayer, 0, aError );
}
}
if( aIncludeText )
{
if( Reference().GetLayer() == aLayer && Reference().IsVisible() )
texts.push_back( &Reference() );
if( Value().GetLayer() == aLayer && Value().IsVisible() )
texts.push_back( &Value() );
}
prms.m_cornerBuffer = &aCornerBuffer;
for( TEXTE_MODULE* textmod : texts )
{
bool forceBold = true;
int penWidth = 0; // force max width for bold text
2020-04-14 12:25:00 +00:00
prms.m_textWidth = textmod->GetEffectiveTextPenWidth() + ( 2 * aInflateValue );
prms.m_error = aError;
wxSize size = textmod->GetTextSize();
if( textmod->IsMirrored() )
size.x = -size.x;
GRText( NULL, textmod->GetTextPos(), BLACK, textmod->GetShownText(),
textmod->GetDrawRotation(), size, textmod->GetHorizJustify(),
textmod->GetVertJustify(), penWidth, textmod->IsItalic(), forceBold,
addTextSegmToPoly, &prms );
}
}
void ZONE_CONTAINER::TransformSolidAreasShapesToPolygon( PCB_LAYER_ID aLayer,
SHAPE_POLY_SET& aCornerBuffer,
int aError ) const
{
if( !m_FilledPolysList.count( aLayer ) || m_FilledPolysList.at( aLayer ).IsEmpty() )
return;
// Just add filled areas if filled polygons outlines have no thickness
if( !GetFilledPolysUseThickness() || GetMinThickness() == 0 )
{
const SHAPE_POLY_SET& polys = m_FilledPolysList.at( aLayer );
aCornerBuffer.Append( polys );
return;
}
// Filled areas have polygons with outline thickness.
// we must create the polygons and add inflated polys
SHAPE_POLY_SET polys = m_FilledPolysList.at( aLayer );
auto board = GetBoard();
int maxError = ARC_HIGH_DEF;
if( board )
maxError = board->GetDesignSettings().m_MaxError;
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( GetMinThickness(), maxError, 360.0 );
polys.InflateWithLinkedHoles( GetMinThickness()/2, numSegs, SHAPE_POLY_SET::PM_FAST );
aCornerBuffer.Append( polys );
}
void EDA_TEXT::TransformBoundingBoxWithClearanceToPolygon( SHAPE_POLY_SET* aCornerBuffer,
int aClearanceValue ) const
{
if( GetText().Length() == 0 )
return;
wxPoint corners[4]; // Buffer of polygon corners
2020-04-14 12:25:00 +00:00
EDA_RECT rect = GetTextBox();
rect.Inflate( aClearanceValue + Millimeter2iu( DEFAULT_TEXT_WIDTH ) );
corners[0].x = rect.GetOrigin().x;
corners[0].y = rect.GetOrigin().y;
corners[1].y = corners[0].y;
corners[1].x = rect.GetRight();
corners[2].x = corners[1].x;
corners[2].y = rect.GetBottom();
corners[3].y = corners[2].y;
corners[3].x = corners[0].x;
aCornerBuffer->NewOutline();
for( wxPoint& corner : corners )
{
// Rotate polygon
RotatePoint( &corner.x, &corner.y, GetTextPos().x, GetTextPos().y, GetTextAngle() );
aCornerBuffer->Append( corner.x, corner.y );
}
}
2012-01-29 19:29:19 +00:00
/**
* Function TransformShapeWithClearanceToPolygonSet
* Convert the text shape to a set of polygons (one per segment).
* @aCornerBuffer = SHAPE_POLY_SET to store the polygon corners
* @aClearanceValue = the clearance around the text
* @aError = the maximum error to allow when approximating curves
*/
void TEXTE_PCB::TransformShapeWithClearanceToPolygonSet( SHAPE_POLY_SET& aCornerBuffer,
int aClearanceValue, int aError ) const
{
wxSize size = GetTextSize();
2012-01-29 19:29:19 +00:00
if( IsMirrored() )
size.x = -size.x;
2012-01-29 19:29:19 +00:00
bool forceBold = true;
2020-04-14 12:25:00 +00:00
int penWidth = GetEffectiveTextPenWidth();
prms.m_cornerBuffer = &aCornerBuffer;
2020-04-14 12:25:00 +00:00
prms.m_textWidth = GetEffectiveTextPenWidth() + ( 2 * aClearanceValue );
prms.m_error = aError;
COLOR4D color = COLOR4D::BLACK; // not actually used, but needed by GRText
2012-01-29 19:29:19 +00:00
if( IsMultilineAllowed() )
2012-01-29 19:29:19 +00:00
{
wxArrayString strings_list;
wxStringSplit( GetShownText(), strings_list, '\n' );
std::vector<wxPoint> positions;
positions.reserve( strings_list.Count() );
2020-04-14 12:25:00 +00:00
GetLinePositions( positions, strings_list.Count() );
2012-01-29 19:29:19 +00:00
for( unsigned ii = 0; ii < strings_list.Count(); ii++ )
2012-01-29 19:29:19 +00:00
{
wxString txt = strings_list.Item( ii );
GRText( NULL, positions[ii], color, txt, GetTextAngle(), size, GetHorizJustify(),
GetVertJustify(), penWidth, IsItalic(), forceBold, addTextSegmToPoly, &prms );
2012-01-29 19:29:19 +00:00
}
}
else
{
GRText( NULL, GetTextPos(), color, GetShownText(), GetTextAngle(), size, GetHorizJustify(),
GetVertJustify(), penWidth, IsItalic(), forceBold, addTextSegmToPoly, &prms );
2012-01-29 19:29:19 +00:00
}
}
void DRAWSEGMENT::TransformShapeWithClearanceToPolygon( SHAPE_POLY_SET& aCornerBuffer,
PCB_LAYER_ID aLayer,
int aClearanceValue, int aError,
bool ignoreLineWidth ) const
{
int width = ignoreLineWidth ? 0 : m_Width;
width += 2 * aClearanceValue;
// Creating a reliable clearance shape for circles and arcs is not so easy, due to
// the error created by segment approximation.
// for a circle this is not so hard: create a polygon from a circle slightly bigger:
// thickness = width + s_error_max, and radius = initial radius + s_error_max/2
// giving a shape with a suitable internal radius and external radius
// For an arc this is more tricky: TODO
switch( m_Shape )
{
case S_CIRCLE:
if( width == 0 )
TransformCircleToPolygon( aCornerBuffer, GetCenter(), GetRadius(), aError );
else
TransformRingToPolygon( aCornerBuffer, GetCenter(), GetRadius(), aError, width );
break;
case S_RECT:
{
std::vector<wxPoint> pts = GetRectCorners();
if( width == 0 )
{
aCornerBuffer.NewOutline();
for( const wxPoint& pt : pts )
aCornerBuffer.Append( pt );
}
if( width > 0 )
{
// Add in segments
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
TransformOvalToPolygon( aCornerBuffer, pts[0], pts[1], width, aError );
TransformOvalToPolygon( aCornerBuffer, pts[1], pts[2], width, aError );
TransformOvalToPolygon( aCornerBuffer, pts[2], pts[3], width, aError );
TransformOvalToPolygon( aCornerBuffer, pts[3], pts[0], width, aError );
}
}
break;
case S_ARC:
TransformArcToPolygon( aCornerBuffer, GetCenter(), GetArcStart(), m_Angle, aError, width );
break;
case S_SEGMENT:
TransformOvalToPolygon( aCornerBuffer, m_Start, m_End, width, aError );
break;
case S_POLYGON:
if( IsPolyShapeValid() )
{
// The polygon is expected to be a simple polygon
// not self intersecting, no hole.
MODULE* module = GetParentModule(); // NULL for items not in footprints
double orientation = module ? module->GetOrientation() : 0.0;
wxPoint offset;
if( module )
offset = module->GetPosition();
// Build the polygon with the actual position and orientation:
std::vector< wxPoint> poly;
poly = BuildPolyPointsList();
for( wxPoint& point : poly )
{
RotatePoint( &point, orientation );
point += offset;
}
if( IsPolygonFilled() || width == 0 )
{
aCornerBuffer.NewOutline();
for( wxPoint& point : poly )
aCornerBuffer.Append( point.x, point.y );
}
if( width > 0 )
{
wxPoint pt1( poly[ poly.size() - 1] );
for( wxPoint pt2 : poly )
{
if( pt2 != pt1 )
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
TransformOvalToPolygon( aCornerBuffer, pt1, pt2, width, aError );
pt1 = pt2;
}
}
}
break;
case S_CURVE: // Bezier curve
{
std::vector<wxPoint> ctrlPoints = { m_Start, m_BezierC1, m_BezierC2, m_End };
BEZIER_POLY converter( ctrlPoints );
std::vector< wxPoint> poly;
converter.GetPoly( poly, m_Width );
if( width != 0 )
{
for( unsigned ii = 1; ii < poly.size(); ii++ )
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
TransformOvalToPolygon( aCornerBuffer, poly[ii-1], poly[ii], width, aError );
}
}
break;
default:
wxFAIL_MSG( "DRAWSEGMENT::TransformShapeWithClearanceToPolygon no implementation for "
+ STROKE_T_asString( m_Shape ) );
break;
}
}
void TRACK::TransformShapeWithClearanceToPolygon( SHAPE_POLY_SET& aCornerBuffer,
PCB_LAYER_ID aLayer,
int aClearanceValue, int aError,
bool ignoreLineWidth ) const
{
wxASSERT_MSG( !ignoreLineWidth, "IgnoreLineWidth has no meaning for tracks." );
switch( Type() )
{
case PCB_VIA_T:
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 = ( m_Width / 2 ) + aClearanceValue;
TransformCircleToPolygon( aCornerBuffer, m_Start, radius, aError );
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
}
break;
case PCB_ARC_T:
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
{
const ARC* arc = static_cast<const ARC*>( this );
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 width = m_Width + ( 2 * aClearanceValue );
VECTOR2D center( arc->GetCenter() );
double angle = arc->GetAngle();
TransformArcToPolygon( aCornerBuffer, (wxPoint) center, GetStart(), angle, aError, width );
}
break;
default:
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 width = m_Width + ( 2 * aClearanceValue );
TransformOvalToPolygon( aCornerBuffer, m_Start, m_End, width, aError );
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
}
break;
}
}
void D_PAD::TransformShapeWithClearanceToPolygon( SHAPE_POLY_SET& aCornerBuffer,
PCB_LAYER_ID aLayer,
int aClearanceValue, int aError,
bool ignoreLineWidth ) const
{
wxASSERT_MSG( !ignoreLineWidth, "IgnoreLineWidth has no meaning for pads." );
// minimal segment count to approximate a circle to create the polygonal pad shape
// This minimal value is mainly for very small pads, like SM0402.
// Most of time pads are using the segment count given by aError value.
const int pad_min_seg_per_circle_count = 16;
double angle = m_orient;
int dx = m_size.x / 2;
int dy = m_size.y / 2;
wxPoint padShapePos = ShapePos(); // Note: for pad having a shape offset,
// the pad position is NOT the shape position
switch( GetShape() )
{
case PAD_SHAPE_CIRCLE:
case PAD_SHAPE_OVAL:
if( dx == dy )
{
TransformCircleToPolygon( aCornerBuffer, padShapePos, dx + aClearanceValue, aError );
}
else
{
int half_width = std::min( dx, dy );
wxPoint delta( dx - half_width, dy - half_width );
RotatePoint( &delta, angle );
TransformOvalToPolygon( aCornerBuffer, padShapePos - delta, padShapePos + delta,
( half_width + aClearanceValue ) * 2, aError );
}
break;
case PAD_SHAPE_TRAPEZOID:
case PAD_SHAPE_RECT:
{
int ddx = GetShape() == PAD_SHAPE_TRAPEZOID ? m_deltaSize.x / 2 : 0;
int ddy = GetShape() == PAD_SHAPE_TRAPEZOID ? m_deltaSize.y / 2 : 0;
wxPoint corners[4];
corners[0] = wxPoint( -dx - ddy, dy + ddx );
corners[1] = wxPoint( dx + ddy, dy - ddx );
corners[2] = wxPoint( dx - ddy, -dy + ddx );
corners[3] = wxPoint( -dx + ddy, -dy - ddx );
SHAPE_POLY_SET outline;
outline.NewOutline();
for( wxPoint& corner : corners )
{
RotatePoint( &corner, angle );
corner += padShapePos;
outline.Append( corner.x, corner.y );
}
if( aClearanceValue )
{
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 = std::max( GetArcToSegmentCount( aClearanceValue, aError, 360.0 ),
pad_min_seg_per_circle_count );
int clearance = aClearanceValue + GetCircleToPolyCorrection( aError );
outline.Inflate( clearance, numSegs );
// TODO: clamp the inflated polygon, because it is slightly too big:
// it was inflated by a value slightly too big to keep rounded corners
// ouside the pad area.
}
2016-04-06 18:15:49 +00:00
aCornerBuffer.Append( outline );
}
break;
case PAD_SHAPE_CHAMFERED_RECT:
2016-04-06 18:15:49 +00:00
case PAD_SHAPE_ROUNDRECT:
{
int radius = GetRoundRectCornerRadius() + aClearanceValue;
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 clearance = aClearanceValue + GetCircleToPolyCorrection( aError );
wxSize shapesize( m_size );
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
radius = radius + GetCircleToPolyCorrection( aError );
shapesize.x += clearance * 2;
shapesize.y += clearance * 2;
bool doChamfer = GetShape() == PAD_SHAPE_CHAMFERED_RECT;
SHAPE_POLY_SET outline;
TransformRoundChamferedRectToPolygon( outline, padShapePos, shapesize, angle, radius,
doChamfer ? GetChamferRectRatio() : 0.0,
doChamfer ? GetChamferPositions() : 0, aError );
aCornerBuffer.Append( outline );
}
break;
case PAD_SHAPE_CUSTOM:
{
SHAPE_POLY_SET outline;
MergePrimitivesAsPolygon( &outline, aLayer );
outline.Rotate( -DECIDEG2RAD( m_orient ) );
outline.Move( VECTOR2I( m_pos ) );
if( aClearanceValue )
{
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 = std::max( GetArcToSegmentCount( aClearanceValue, aError, 360.0 ),
pad_min_seg_per_circle_count );
int clearance = aClearanceValue + GetCircleToPolyCorrection( aError );
outline.Inflate( clearance, numSegs );
2020-09-22 21:05:09 +00:00
outline.Simplify( SHAPE_POLY_SET::PM_FAST );
outline.Fracture( SHAPE_POLY_SET::PM_FAST );
}
aCornerBuffer.Append( outline );
}
break;
default:
wxFAIL_MSG( "D_PAD::TransformShapeWithClearanceToPolygon no implementation for "
+ PAD_SHAPE_T_asString( GetShape() ) );
break;
}
}
2016-04-06 18:15:49 +00:00
bool D_PAD::TransformHoleWithClearanceToPolygon( SHAPE_POLY_SET& aCornerBuffer, int aInflateValue,
int aError ) const
{
wxSize drillsize = GetDrillSize();
2014-05-17 19:29:15 +00:00
if( !drillsize.x || !drillsize.y )
return false;
const SHAPE_SEGMENT* seg = GetEffectiveHoleShape();
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
TransformOvalToPolygon( aCornerBuffer, (wxPoint) seg->GetSeg().A, (wxPoint) seg->GetSeg().B,
seg->GetWidth() + aInflateValue * 2, aError );
return true;
}
void ZONE_CONTAINER::TransformShapeWithClearanceToPolygon( SHAPE_POLY_SET& aCornerBuffer,
PCB_LAYER_ID aLayer, int aClearance,
int aError, bool ignoreLineWidth ) const
{
wxASSERT_MSG( !ignoreLineWidth, "IgnoreLineWidth has no meaning for zones." );
if( !m_FilledPolysList.count( aLayer ) )
return;
aCornerBuffer = m_FilledPolysList.at( aLayer );
int numSegs = GetArcToSegmentCount( aClearance, aError, 360.0 );
aCornerBuffer.Inflate( aClearance, numSegs );
aCornerBuffer.Simplify( SHAPE_POLY_SET::PM_STRICTLY_SIMPLE );
}