Make explicit that we expect co-ordinates to be finite

The various spatial-vector, line, point, region and margin types have
all long taken for granted that their co-ordinates are finite and, in
particular, not NaN. Make this expectation explicit in the
documentation. Added assertions where (chiefly) noexcept and (where
the assertion would be a simple qIsFinite() call) constexpr didn't
preclude doing so. Also assert against zero divisors that would lead
to non-finite results.

Make minor clean-ups to docs in the process. QMarginsF had several
methods whose declaration, definition and docs weren't consistent in
their parameter names.

Task-number: QTBUG-89010
Change-Id: I601a830959d13a73dcb17ce31a1202a1486c8f7f
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>

1 files changed, 28 insertions, 2 deletions

diff --git a/src/gui/math3d/qvectornd.h b/src/gui/math3d/qvectornd.h
index 8d2789338c..153f3743d8 100644
--- a/src/gui/math3d/qvectornd.h
+++ b/src/gui/math3d/qvectornd.h
@@ -147,11 +147,14 @@ QT_WARNING_POP
 
     constexpr friend inline QVector2D operator/(QVector2D vector, float divisor)
     {
+        Q_ASSERT(divisor < 0 || divisor > 0);
         return QVector2D(vector.v[0] / divisor, vector.v[1] / divisor);
     }
 
     constexpr friend inline QVector2D operator/(QVector2D vector, QVector2D divisor)
     {
+        Q_ASSERT(divisor.v[0] < 0 || divisor.v[0] > 0);
+        Q_ASSERT(divisor.v[1] < 0 || divisor.v[1] > 0);
         return QVector2D(vector.v[0] / divisor.v[0], vector.v[1] / divisor.v[1]);
     }
 
@@ -288,12 +291,17 @@ QT_WARNING_POP
 
     constexpr friend inline QVector3D operator/(QVector3D vector, float divisor)
     {
+        Q_ASSERT(divisor < 0 || divisor > 0);
         return QVector3D(vector.v[0] / divisor, vector.v[1] / divisor, vector.v[2] / divisor);
     }
 
     constexpr friend inline QVector3D operator/(QVector3D vector, QVector3D divisor)
     {
-        return QVector3D(vector.v[0] / divisor.v[0], vector.v[1] / divisor.v[1], vector.v[2] / divisor.v[2]);
+        Q_ASSERT(divisor.v[0] > 0 || divisor.v[0] < 0);
+        Q_ASSERT(divisor.v[1] > 0 || divisor.v[1] < 0);
+        Q_ASSERT(divisor.v[2] > 0 || divisor.v[2] < 0);
+        return QVector3D(vector.v[0] / divisor.v[0], vector.v[1] / divisor.v[1],
+                         vector.v[2] / divisor.v[2]);
     }
 
     friend Q_GUI_EXPORT bool qFuzzyCompare(QVector3D v1, QVector3D v2) noexcept;
@@ -422,12 +430,18 @@ QT_WARNING_POP
 
     constexpr friend inline QVector4D operator/(QVector4D vector, float divisor)
     {
+        Q_ASSERT(divisor < 0 || divisor > 0);
         return QVector4D(vector.v[0] / divisor, vector.v[1] / divisor, vector.v[2] / divisor, vector.v[3] / divisor);
     }
 
     constexpr friend inline QVector4D operator/(QVector4D vector, QVector4D divisor)
     {
-        return QVector4D(vector.v[0] / divisor.v[0], vector.v[1] / divisor.v[1], vector.v[2] / divisor.v[2], vector.v[3] / divisor.v[3]);
+        Q_ASSERT(divisor.v[0] > 0 || divisor.v[0] < 0);
+        Q_ASSERT(divisor.v[1] > 0 || divisor.v[1] < 0);
+        Q_ASSERT(divisor.v[2] > 0 || divisor.v[2] < 0);
+        Q_ASSERT(divisor.v[3] > 0 || divisor.v[3] < 0);
+        return QVector4D(vector.v[0] / divisor.v[0], vector.v[1] / divisor.v[1],
+                         vector.v[2] / divisor.v[2], vector.v[3] / divisor.v[3]);
     }
 
     friend Q_GUI_EXPORT bool qFuzzyCompare(QVector4D v1, QVector4D v2) noexcept;
@@ -576,6 +590,7 @@ constexpr inline QVector2D &QVector2D::operator*=(QVector2D vector) noexcept
 
 constexpr inline QVector2D &QVector2D::operator/=(float divisor)
 {
+    Q_ASSERT(divisor < 0 || divisor > 0);
     v[0] /= divisor;
     v[1] /= divisor;
     return *this;
@@ -583,6 +598,8 @@ constexpr inline QVector2D &QVector2D::operator/=(float divisor)
 
 constexpr inline QVector2D &QVector2D::operator/=(QVector2D vector)
 {
+    Q_ASSERT(vector.v[0] > 0 || vector.v[0] < 0);
+    Q_ASSERT(vector.v[1] > 0 || vector.v[1] < 0);
     v[0] /= vector.v[0];
     v[1] /= vector.v[1];
     return *this;
@@ -736,6 +753,7 @@ constexpr inline QVector3D &QVector3D::operator*=(QVector3D vector) noexcept
 
 constexpr inline QVector3D &QVector3D::operator/=(float divisor)
 {
+    Q_ASSERT(divisor < 0 || divisor > 0);
     v[0] /= divisor;
     v[1] /= divisor;
     v[2] /= divisor;
@@ -744,6 +762,9 @@ constexpr inline QVector3D &QVector3D::operator/=(float divisor)
 
 constexpr inline QVector3D &QVector3D::operator/=(QVector3D vector)
 {
+    Q_ASSERT(vector.v[0] > 0 || vector.v[0] < 0);
+    Q_ASSERT(vector.v[1] > 0 || vector.v[1] < 0);
+    Q_ASSERT(vector.v[2] > 0 || vector.v[2] < 0);
     v[0] /= vector.v[0];
     v[1] /= vector.v[1];
     v[2] /= vector.v[2];
@@ -947,6 +968,7 @@ constexpr inline QVector4D &QVector4D::operator*=(QVector4D vector) noexcept
 
 constexpr inline QVector4D &QVector4D::operator/=(float divisor)
 {
+    Q_ASSERT(divisor < 0 || divisor > 0);
     v[0] /= divisor;
     v[1] /= divisor;
     v[2] /= divisor;
@@ -956,6 +978,10 @@ constexpr inline QVector4D &QVector4D::operator/=(float divisor)
 
 constexpr inline QVector4D &QVector4D::operator/=(QVector4D vector)
 {
+    Q_ASSERT(vector.v[0] > 0 || vector.v[0] < 0);
+    Q_ASSERT(vector.v[1] > 0 || vector.v[1] < 0);
+    Q_ASSERT(vector.v[2] > 0 || vector.v[2] < 0);
+    Q_ASSERT(vector.v[3] > 0 || vector.v[3] < 0);
     v[0] /= vector.v[0];
     v[1] /= vector.v[1];
     v[2] /= vector.v[2];