/* * Copyright 2012 Google Inc. * * Use of this source code is governed by a BSD-style license that can be * found in the LICENSE file. */ #include "PathOpsExtendedTest.h" #include "PathOpsThreadedCommon.h" static void testOpCubicsMain(PathOpsThreadState* data) { SkASSERT(data); PathOpsThreadState& state = *data; char pathStr[1024]; // gdb: set print elements 400 sk_bzero(pathStr, sizeof(pathStr)); for (int a = 0 ; a < 6; ++a) { for (int b = a + 1 ; b < 7; ++b) { for (int c = 0 ; c < 6; ++c) { for (int d = c + 1 ; d < 7; ++d) { for (int e = SkPath::kWinding_FillType ; e <= SkPath::kEvenOdd_FillType; ++e) { for (int f = SkPath::kWinding_FillType ; f <= SkPath::kEvenOdd_FillType; ++f) { SkPath pathA, pathB; char* str = pathStr; pathA.setFillType((SkPath::FillType) e); str += sprintf(str, " path.setFillType(SkPath::k%s_FillType);\n", e == SkPath::kWinding_FillType ? "Winding" : e == SkPath::kEvenOdd_FillType ? "EvenOdd" : "?UNDEFINED"); pathA.moveTo(SkIntToScalar(state.fA), SkIntToScalar(state.fB)); str += sprintf(str, " path.moveTo(%d,%d);\n", state.fA, state.fB); pathA.cubicTo(SkIntToScalar(state.fC), SkIntToScalar(state.fD), SkIntToScalar(b), SkIntToScalar(a), SkIntToScalar(d), SkIntToScalar(c)); str += sprintf(str, " path.cubicTo(%d,%d, %d,%d, %d,%d);\n", state.fC, state.fD, b, a, d, c); pathA.close(); str += sprintf(str, " path.close();\n"); pathB.setFillType((SkPath::FillType) f); str += sprintf(str, " pathB.setFillType(SkPath::k%s_FillType);\n", f == SkPath::kWinding_FillType ? "Winding" : f == SkPath::kEvenOdd_FillType ? "EvenOdd" : "?UNDEFINED"); pathB.moveTo(SkIntToScalar(a), SkIntToScalar(b)); str += sprintf(str, " pathB.moveTo(%d,%d);\n", a, b); pathB.cubicTo(SkIntToScalar(c), SkIntToScalar(d), SkIntToScalar(state.fB), SkIntToScalar(state.fA), SkIntToScalar(state.fD), SkIntToScalar(state.fC)); str += sprintf(str, " pathB.cubicTo(%d,%d, %d,%d, %d,%d);\n", c, d, state.fB, state.fA, state.fD, state.fC); pathB.close(); str += sprintf(str, " pathB.close();\n"); for (int op = 0 ; op <= kXOR_PathOp; ++op) { outputProgress(state.fPathStr, pathStr, (SkPathOp) op); testPathOp(state.fReporter, pathA, pathB, (SkPathOp) op); } } } } } } } } static void PathOpsOpCubicsThreadedTest(skiatest::Reporter* reporter) { int threadCount = initializeTests(reporter, "cubicOp"); PathOpsThreadedTestRunner testRunner(reporter, threadCount); for (int a = 0; a < 6; ++a) { // outermost for (int b = a + 1; b < 7; ++b) { for (int c = 0 ; c < 6; ++c) { for (int d = c + 1; d < 7; ++d) { *testRunner.fRunnables.append() = SkNEW_ARGS(PathOpsThreadedRunnable, (&testOpCubicsMain, a, b, c, d, &testRunner)); } } if (!reporter->allowExtendedTest()) goto finish; } } finish: testRunner.render(); } #include "TestClassDef.h" DEFINE_TESTCLASS_SHORT(PathOpsOpCubicsThreadedTest)