Remove comments

Signed-off-by: Mateusz Palczuk <[email protected]>

common/math/geometry/test/test_catmull_rom_spline.cpp

@@ -205,16 +205,6 @@ TEST(CatmullRomSpline, GetCollisionPointIn2D)
   }
 }
 
-/**
- * This test is not passing because the getCollisionPointIn2D function calling getCollisionPointIn2D
- * on HermiteCurve, which operates on a normalized lengths. After that the accumulated collision S
- * is calculated as a sum of previous HermiteCurves lengths + this HermiteCurve collision S (which
- * is calculated relative to the total length of this HermiteCurve).
- * 
- * Possible fix:
- * - Add autoscale argument to the HermiteCurve getCollisionPointIn2D function and if it is true
- *   then multiply S value by total length (like in almost every other member function of this class)
- */
 TEST(CatmullRomSpline, getCollisionPointIn2D)
 {
   const std::vector<geometry_msgs::msg::Point> points{
@@ -286,11 +276,11 @@ TEST(CatmullRomSpline, getCollisionPointIn2DPolygon)
 
   const auto collision_s0 = spline.getCollisionPointIn2D(polygon);
   EXPECT_TRUE(collision_s0);
-  EXPECT_NEAR(collision_s0.value(), 0.56727227, EPS);  // TODO check after fix if value is OK
+  EXPECT_NEAR(collision_s0.value(), 0.56727227, EPS);
 
   const auto collision_s1 = spline.getCollisionPointIn2D(polygon, true);
   EXPECT_TRUE(collision_s1);
-  EXPECT_NEAR(collision_s1.value(), 0.56727227, EPS);  // TODO check after fix if value is OK
+  EXPECT_NEAR(collision_s1.value(), 0.56727227, EPS);
 }
 
 TEST(CatmullRomSpline, getCollisionPointIn2DEmpty)
@@ -308,10 +298,6 @@ TEST(CatmullRomSpline, getMaximum2DCurvatureLine)
   EXPECT_DOUBLE_EQ(spline.getMaximum2DCurvature(), 0.0);
 }
 
-/**
- * This test fails, because this function chooses maximum value and not the value with greatest
- * absolute value. The curvature -10 is much tighter than 2, but here 2 would be the maximum.
- */
 TEST(CatmullRomSpline, getMaximum2DCurvatureCurve)
 {
   const std::vector<geometry_msgs::msg::Point> points{
@@ -361,22 +347,9 @@ TEST(CatmullRomSpline, getPolygon)
   EXPECT_POINT_NEAR(polygon[23], makePoint(4.0, 0.5), EPS);
 }
 
-/**
- * Statement spline.getPolygon(1, 0) throws this exception:
- * /home/aorusadmin/autoware_ss2_test_implementation/src/simulator/scenario_simulator/common/math/geometry/src/spline/catmull_rom_spline.cpp:256: failed to calculate curve index
- * This happens because the num_points (= 0) is used in getLeftBounds to calculate step_size (which
- * is then equal to infinity). This leads to an infinite `s` value which propagates this way:
- * getLeftBounds -> getLeftBoundsPoint -> getNormalVector -> getCurveIndexAndS (which throws this exception)
- * 
- * Possible fix:
- * - Catch this exception and throw our own which will have a meaningful description.
- * - Add early return in getPolygon when num_points == 0 so that it does not throw.
- */
 TEST(CatmullRomSpline, getPolygonEdge)
 {
   math::geometry::CatmullRomSpline spline = makeCurve();
-  // FIXME should this throw a propagated error, or catch it and throw its own with a clear message what has happened?
-  // EXPECT_THROW(spline.getPolygon(1, 0), common::SimulationError);
   std::vector<geometry_msgs::msg::Point> polygon0 = spline.getPolygon(1.0, 0);
   EXPECT_EQ(polygon0.size(), static_cast<size_t>(0));
 

common/math/geometry/test/test_catmull_rom_subspline.cpp

@@ -54,10 +54,6 @@ TEST(CatmullRomSubspline, getLength_zero)
   EXPECT_DOUBLE_EQ(spline1.getLength(), 0.0);
 }
 
-/**
- * This test tests function that does not function properly yet. I have opened a github issue about
- * the spline collision calculation (which directly affects subspline collision calculation).
- */
 TEST(CatmullRomSubspline, getCollisionPointIn2D)
 {
   const auto spline_ptr = makeLine();
@@ -88,19 +84,6 @@ TEST(CatmullRomSubspline, getCollisionPointIn2D)
   EXPECT_NEAR(ans11.value(), std::hypot(1.25, 3.75), EPS);
 }
 
-/**
- * This test fails because the subspline logic is incorrect. It obtains the collision with
- * the polygon (either first or lst one depending on the search_backward parameter) and then - if
- * the collision point is outside of subspline - returns nothing. This leads to incorrect behavior
- * when the polygon collides with the spline in more than one point, but some of these points are 
- * not on the subspline. If the polygon collides with the spline 2 times and the first collision 
- * point is outside of the subspline, but the second collision point is inside of the subspline and
- * the search_backward parameter is set to true, the second (valid) collision is ignored and nothing
- * is returned.
- * The possible fix for this may be to add function in spline that returns all collision points, not
- * only the first one, and in subspline check all of them and choose the first one that is in the
- * subspline.
- */
 TEST(CatmullRomSubspline, getCollisionPointIn2D_shiftedBeginning)
 {
   const auto spline_ptr = makeLine();