From: JunsuChoi Date: Mon, 6 Jul 2020 07:35:10 +0000 (+0900) Subject: SvgLoader: Support arc_to draw X-Git-Tag: accepted/tizen/unified/20200806.062539~60^2 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=refs%2Fchanges%2F64%2F237864%2F4;p=platform%2Fcore%2Fgraphics%2Ftizenvg.git SvgLoader: Support arc_to draw Change-Id: I950c8e850605f990d6a0aa59a067ced571ffdb51 --- diff --git a/src/loaders/svg_loader/tvgSvgPath.cpp b/src/loaders/svg_loader/tvgSvgPath.cpp index c3d3cac..ba01ae7 100644 --- a/src/loaders/svg_loader/tvgSvgPath.cpp +++ b/src/loaders/svg_loader/tvgSvgPath.cpp @@ -14,7 +14,7 @@ static char* _skipComma(const char* content) } -static inline bool _parseNumber(char** content, float* number) +static bool _parseNumber(char** content, float* number) { char* end = NULL; *number = strtof(*content, &end); @@ -26,7 +26,7 @@ static inline bool _parseNumber(char** content, float* number) } -static inline bool _parseLong(char** content, int* number) +static bool _parseLong(char** content, int* number) { char* end = NULL; *number = strtol(*content, &end, 10) ? 1 : 0; @@ -36,6 +36,183 @@ static inline bool _parseLong(char** content, int* number) return true; } +void _pathAppendArcTo(vector* cmds, vector* pts, float* arr, Point* cur, Point* curCtl, float x, float y, float rx, float ry, float angle, bool largeArc, bool sweep) +{ + float cxp, cyp, cx, cy; + float sx, sy; + float cosPhi, sinPhi; + float dx2, dy2; + float x1p, y1p; + float x1p2, y1p2; + float rx2, ry2; + float lambda; + float c; + float at; + float theta1, deltaTheta; + float nat; + float delta, bcp; + float cosPhiRx, cosPhiRy; + float sinPhiRx, sinPhiRy; + float cosTheta1, sinTheta1; + int segments, i; + + //Some helpful stuff is available here: + //http://www.w3.org/TR/SVG/implnote.html#ArcImplementationNotes + sx = cur->x; + sy = cur->y; + + //If start and end points are identical, then no arc is drawn + if ((fabs(x - sx) < (1.0f / 256.0f)) && (fabs(y - sy) < (1.0f / 256.0f))) return; + + //Correction of out-of-range radii, see F6.6.1 (step 2) + rx = fabs(rx); + ry = fabs(ry); + if ((rx < 0.5f) || (ry < 0.5f)) { + Point p = {x, y}; + cmds->push_back(PathCommand::LineTo); + pts->push_back(p); + *cur = p; + return; + } + + angle = angle * M_PI / 180.0f; + cosPhi = cosf(angle); + sinPhi = sinf(angle); + dx2 = (sx - x) / 2.0f; + dy2 = (sy - y) / 2.0f; + x1p = cosPhi * dx2 + sinPhi * dy2; + y1p = cosPhi * dy2 - sinPhi * dx2; + x1p2 = x1p * x1p; + y1p2 = y1p * y1p; + rx2 = rx * rx; + ry2 = ry * ry; + lambda = (x1p2 / rx2) + (y1p2 / ry2); + + //Correction of out-of-range radii, see F6.6.2 (step 4) + if (lambda > 1.0f) { + //See F6.6.3 + float lambdaRoot = sqrt(lambda); + + rx *= lambdaRoot; + ry *= lambdaRoot; + //Update rx2 and ry2 + rx2 = rx * rx; + ry2 = ry * ry; + } + + c = (rx2 * ry2) - (rx2 * y1p2) - (ry2 * x1p2); + + //Check if there is no possible solution + //(i.e. we can't do a square root of a negative value) + if (c < 0.0f) { + //Scale uniformly until we have a single solution + //(see F6.2) i.e. when c == 0.0 + float scale = sqrt(1.0f - c / (rx2 * ry2)); + rx *= scale; + ry *= scale; + //Update rx2 and ry2 + rx2 = rx * rx; + ry2 = ry * ry; + + //Step 2 (F6.5.2) - simplified since c == 0.0 + cxp = 0.0f; + cyp = 0.0f; + //Step 3 (F6.5.3 first part) - simplified since cxp and cyp == 0.0 + cx = 0.0f; + cy = 0.0f; + } else { + //Complete c calculation + c = sqrt(c / ((rx2 * y1p2) + (ry2 * x1p2))); + //Inverse sign if Fa == Fs + if (largeArc == sweep) c = -c; + + //Step 2 (F6.5.2) + cxp = c * (rx * y1p / ry); + cyp = c * (-ry * x1p / rx); + + //Step 3 (F6.5.3 first part) + cx = cosPhi * cxp - sinPhi * cyp; + cy = sinPhi * cxp + cosPhi * cyp; + } + + //Step 3 (F6.5.3 second part) we now have the center point of the ellipse + cx += (sx + x) / 2.0f; + cy += (sy + y) / 2.0f; + + //Sstep 4 (F6.5.4) + //We dont' use arccos (as per w3c doc), see + //http://www.euclideanspace.com/maths/algebra/vectors/angleBetween/index.htm + //Note: atan2 (0.0, 1.0) == 0.0 + at = atan2(((y1p - cyp) / ry), ((x1p - cxp) / rx)); + theta1 = (at < 0.0f) ? 2.0f * M_PI + at : at; + + nat = atan2(((-y1p - cyp) / ry), ((-x1p - cxp) / rx)); + deltaTheta = (nat < at) ? 2.0f * M_PI - at + nat : nat - at; + + if (sweep) { + //Ensure delta theta < 0 or else add 360 degrees + if (deltaTheta < 0.0f) deltaTheta += 2.0f * M_PI; + } else { + //Ensure delta theta > 0 or else substract 360 degrees + if (deltaTheta > 0.0f) deltaTheta -= 2.0f * M_PI; + } + + //Add several cubic bezier to approximate the arc + //(smaller than 90 degrees) + //We add one extra segment because we want something + //Smaller than 90deg (i.e. not 90 itself) + segments = (int)(fabs(deltaTheta / M_PI_2)) + 1.0f; + delta = deltaTheta / segments; + + //http://www.stillhq.com/ctpfaq/2001/comp.text.pdf-faq-2001-04.txt (section 2.13) + bcp = 4.0f / 3.0f * (1.0f - cos(delta / 2.0f)) / sin(delta / 2.0f); + + cosPhiRx = cosPhi * rx; + cosPhiRy = cosPhi * ry; + sinPhiRx = sinPhi * rx; + sinPhiRy = sinPhi * ry; + + cosTheta1 = cos(theta1); + sinTheta1 = sin(theta1); + + for (i = 0; i < segments; ++i) { + //End angle (for this segment) = current + delta + float c1x, c1y, ex, ey, c2x, c2y; + float theta2 = theta1 + delta; + float cosTheta2 = cos(theta2); + float sinTheta2 = sin(theta2); + static Point p[3]; + + //First control point (based on start point sx,sy) + c1x = sx - bcp * (cosPhiRx * sinTheta1 + sinPhiRy * cosTheta1); + c1y = sy + bcp * (cosPhiRy * cosTheta1 - sinPhiRx * sinTheta1); + + //End point (for this segment) + ex = cx + (cosPhiRx * cosTheta2 - sinPhiRy * sinTheta2); + ey = cy + (sinPhiRx * cosTheta2 + cosPhiRy * sinTheta2); + + //Second control point (based on end point ex,ey) + c2x = ex + bcp * (cosPhiRx * sinTheta2 + sinPhiRy * cosTheta2); + c2y = ey + bcp * (sinPhiRx * sinTheta2 - cosPhiRy * cosTheta2); + cmds->push_back(PathCommand::CubicTo); + p[0] = {c1x, c1y}; + p[1] = {c2x, c2y}; + p[2] = {ex, ey}; + pts->push_back(p[0]); + pts->push_back(p[1]); + pts->push_back(p[2]); + *curCtl = p[1]; + *cur = p[2]; + + //Next start point is the current end point (same for angle) + sx = ex; + sy = ey; + theta1 = theta2; + //Avoid recomputations + cosTheta1 = cosTheta2; + sinTheta1 = sinTheta2; + } +} static int _numberCount(char cmd) { @@ -170,12 +347,12 @@ static void _processCommand(vector* cmds, vector* pts, char } case 'q': case 'Q': { - tvg::Point p[3]; + Point p[3]; float ctrl_x0 = (cur->x + 2 * arr[0]) * (1.0 / 3.0); float ctrl_y0 = (cur->y + 2 * arr[1]) * (1.0 / 3.0); float ctrl_x1 = (arr[2] + 2 * arr[0]) * (1.0 / 3.0); float ctrl_y1 = (arr[3] + 2 * arr[1]) * (1.0 / 3.0); - cmds->push_back(tvg::PathCommand::CubicTo); + cmds->push_back(PathCommand::CubicTo); p[0] = {ctrl_x0, ctrl_y0}; p[1] = {ctrl_x1, ctrl_y1}; p[2] = {arr[2], arr[3]}; @@ -217,12 +394,8 @@ static void _processCommand(vector* cmds, vector* pts, char } case 'a': case 'A': { - //TODO: Implement arc_to - break; - } - case 'E': - case 'e': { - //TODO: Implement arc + _pathAppendArcTo(cmds, pts, arr, cur, curCtl, arr[5], arr[6], arr[0], arr[1], arr[2], arr[3], arr[4]); + *cur = {arr[5] ,arr[6]}; break; } default: {