Fixed all issues detected by the test cases under MinGW - such as usage of abs() -> changed to qAbs(), rounding issues detected during calculations, and changed streaming methods with qDebug() (QDebug vs &QDebug issue).

src/blackmisc/basestreamstringifier.h

@@ -20,12 +20,27 @@ template <class UsingClass> class CBaseStreamStringifier
      * \param uc
      * \return
      */
-    friend QDebug &operator<<(QDebug &debug, const UsingClass &uc)
+    friend QDebug operator<<(QDebug debug, const UsingClass &uc)
     {
-        const CBaseStreamStringifier &s = uc;
-        debug << s.stringForStreaming();
+        const CBaseStreamStringifier &sf = uc; // allows to acces protected method
+        debug << sf.stringForStreaming();
         return debug;
     }
+    // msvc2010: friend QDebug &operator<<(QDebug &debug, const UsingClass &uc)
+    // MinGW: No reference
+
+    /*!
+     * \brief Operator << based on text stream
+     * \param textStream
+     * \param uc
+     * \return
+     */
+    friend QTextStream &operator<<(QTextStream &textStream, const UsingClass &uc)
+    {
+        const CBaseStreamStringifier &sf = uc; // allows to acces protected method
+        textStream << sf.stringForStreaming();
+        return textStream;
+    }
 
     /*!
      * \brief Operator << when there is no debug stream
@@ -46,8 +61,8 @@ template <class UsingClass> class CBaseStreamStringifier
      */
     friend QDataStream &operator<<(QDataStream &stream, const UsingClass &uc)
     {
-        const CBaseStreamStringifier &s = uc;
-        stream << s.stringForStreaming();
+        const CBaseStreamStringifier &sf = uc; // allows to acces protected method
+        stream << sf.stringForStreaming();
         return stream;
     }
 
@@ -59,8 +74,8 @@ template <class UsingClass> class CBaseStreamStringifier
      */
     friend CLogMessage &operator<<(CLogMessage &log, const UsingClass &uc)
     {
-        const CBaseStreamStringifier &s = uc;
-        log << s.stringForStreaming();
+        const CBaseStreamStringifier &sf = uc; // allows to acces protected method
+        log << sf.stringForStreaming();
         return log;
     }
 
@@ -72,8 +87,8 @@ template <class UsingClass> class CBaseStreamStringifier
      */
     friend std::ostream &operator<<(std::ostream &ostr, const UsingClass &uc)
     {
-        const CBaseStreamStringifier &s = uc;
-        ostr << s.stringForStreaming().toStdString();
+        const CBaseStreamStringifier &sf = uc; // allows to acces protected method
+        ostr << sf.stringForStreaming().toStdString();
         return ostr;
     }
 

src/blackmisc/coordinatetransformation.cpp

@@ -66,8 +66,6 @@ CCoordinateEcef CCoordinateTransformation::toEcef(const CCoordinateNed &ned)
  */
 CCoordinateEcef CCoordinateTransformation::toEcef(const CCoordinateGeodetic &geo)
 {
-    // TODO: Clarify the comparisons with fixed angles (==90, ==180) -> what happens here
-
     CLatitude lat = geo.latitude();
     CLongitude lon = geo.longitude();
 
@@ -96,10 +94,9 @@ CCoordinateEcef CCoordinateTransformation::toEcef(const CCoordinateGeodetic &geo
  */
 CCoordinateNed CCoordinateTransformation::toNed(const CCoordinateEcef &ecef, const CCoordinateGeodetic &referencePosition)
 {
-
     CLatitude lat = referencePosition.latitude();
     CLongitude lon = referencePosition.longitude();
-    double angleRad = - (lat.value(CAngleUnit::rad())) - CMath::PI() / 2;
+    double angleRad = - (lat.value(CAngleUnit::rad())) - CMath::PIHALF();
 
     CMatrix3x3 dcm1;
     CMatrix3x3 dcm2(0.0);

src/blackmisc/coordinatetransformation.h

@@ -46,7 +46,7 @@ private:
      */
     static const double &Flattening()
     {
-        static double f = 1 / 298.257223563;
+        static double f = 1.0 / 298.257223563;
         return f;
     }
 
@@ -76,7 +76,7 @@ private:
      */
     static const double &e2abs()
     {
-        static double e2abs = abs(e2());
+        static double e2abs = qAbs(e2());
         return e2abs;
     }
 
@@ -86,7 +86,7 @@ private:
      */
     static const double &e2m()
     {
-        static double e2m = BlackMisc::Math::CMath::square(1 - Flattening());
+        static double e2m = BlackMisc::Math::CMath::square(1.0 - Flattening());
         return e2m;
     }
 
@@ -95,7 +95,6 @@ private:
      */
     CCoordinateTransformation() {}
 
-
 public:
     /*!
      * \brief NED to ECEF

src/blackmisc/mathematics.cpp

@@ -1,10 +1,10 @@
-/*  Copyright (C) 2013 VATSIM Community / authors
+/*  Copyright (C) 2013 VATSIM Community / contributors
  *  This Source Code Form is subject to the terms of the Mozilla Public
  *  License, v. 2.0. If a copy of the MPL was not distributed with this
  *  file, You can obtain one at http://mozilla.org/MPL/2.0/. */
 
 #include "blackmisc/mathematics.h"
-#include <algorithm>    // std::max
+#include <algorithm> // std::max
 #include <cmath>
 
 namespace BlackMisc
@@ -17,8 +17,8 @@ namespace Math
  */
 double CMath::hypot(double x, double y)
 {
-    x = abs(x);
-    y = abs(y);
+    x = qAbs(x);
+    y = qAbs(y);
     double max = std::max(x, y);
     double min = std::min(x, y);
     double r = min / max;
@@ -31,7 +31,7 @@ double CMath::hypot(double x, double y)
 double CMath::cubicRootReal(double x)
 {
     double result;
-    result = std::pow(std::abs(x), (double)1 / 3);
+    result = std::pow(qAbs(x), (double)1.0 / 3.0);
     return x < 0 ? -result : result;
 }
 

src/blackmisc/mathematics.h

@@ -70,6 +70,16 @@ public:
      */
     static double roundEpsilon(double value, double epsilon);
 
+    /*!
+     * \brief PI
+     * \return
+     */
+    static const double &PIHALF()
+    {
+        static double pi = 2.0 * qAtan(1.0);
+        return pi;
+    }
+
     /*!
      * \brief PI
      * \return
@@ -86,7 +96,7 @@ public:
      */
     static const double &PI2()
     {
-        static double pi2 = PI();
+        static double pi2 = 8.0 * qAtan(1.0);
         return pi2;
     }
 

src/blackmisc/pqphysicalquantity.cpp

@@ -57,26 +57,29 @@ template <class MU, class PQ> bool CPhysicalQuantity<MU, PQ>::operator ==(const
     // some special cases for best quality
     double diff;
     const double lenient = 1.001; // even diff already has a rounding issue to be avoided
+    bool eq = false;
     if (this->m_unit == otherQuantity.m_unit)
     {
         // same unit
         if (this->m_isIntegerBaseValue && otherQuantity.m_isIntegerBaseValue)
         {
             // pure integer comparison, no rounding issues
-            return this->m_unitValueI == otherQuantity.m_unitValueI;
+            eq = this->m_unitValueI == otherQuantity.m_unitValueI;
         }
         else
         {
-            diff = abs(this->m_unitValueD - otherQuantity.m_unitValueD);
-            return diff <= (lenient * this->m_unit.getEpsilon());
+            // same unit, comparison based on double
+            diff = qAbs(this->m_unitValueD - otherQuantity.m_unitValueD);
+            eq = diff <= (lenient * this->m_unit.getEpsilon());
         }
     }
     else
     {
         // based on SI value
-        diff = abs(this->m_convertedSiUnitValueD - otherQuantity.m_convertedSiUnitValueD);
-        return diff <= (lenient * this->m_unit.getEpsilon());
+        diff = qAbs(this->m_convertedSiUnitValueD - otherQuantity.m_convertedSiUnitValueD);
+        eq = diff <= (lenient * this->m_conversionSiUnit.getEpsilon());
     }
+    return eq;
 }
 
 /*
@@ -89,7 +92,7 @@ template <class MU, class PQ> bool CPhysicalQuantity<MU, PQ>::operator !=(const
 }
 
 /*
- * Assigment operator =
+ * Assignment operator =
  */
 template <class MU, class PQ> CPhysicalQuantity<MU, PQ>& CPhysicalQuantity<MU, PQ>::operator=(const CPhysicalQuantity<MU, PQ> &otherQuantity)
 {
@@ -238,9 +241,11 @@ template <class MU, class PQ> PQ CPhysicalQuantity<MU, PQ>::operator /(double di
  */
 template <class MU, class PQ> bool CPhysicalQuantity<MU, PQ>::operator <(const CPhysicalQuantity<MU, PQ> &otherQuantity) const
 {
-    if (this == &otherQuantity) return false;
+    if ((*this) == otherQuantity) return false;
+
+    // == considers epsilon, so we now have a diff > epsilon here
     double diff = this->m_convertedSiUnitValueD - otherQuantity.m_convertedSiUnitValueD;
-    return (diff < 0 && abs(diff) >= this->m_unit.getEpsilon());
+    return (diff < 0);
 }
 
 /*
@@ -249,7 +254,7 @@ template <class MU, class PQ> bool CPhysicalQuantity<MU, PQ>::operator <(const C
 template <class MU, class PQ> bool CPhysicalQuantity<MU, PQ>::operator >(const CPhysicalQuantity<MU, PQ> &otherQuantity) const
 {
     if (this == &otherQuantity) return false;
-    return otherQuantity < *this;
+    return otherQuantity < (*this);
 }
 
 /*