CppUnit - C++ port of JUnit

Update of /cvsroot/cppunit/cppunit
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv13323

Modified Files:
	ChangeLog 
Log Message:
* doc/cookbook.dox: changed suite() to return a TestSuite instead
  of a Test to avoid introducing unnecessary complexity.

Index: ChangeLog
===================================================================
RCS file: /cvsroot/cppunit/cppunit/ChangeLog,v
retrieving revision 1.263
retrieving revision 1.264
diff -C2 -d -r1.263 -r1.264
*** ChangeLog	25 Feb 2007 03:54:34 -0000	1.263
--- ChangeLog	25 Feb 2007 09:21:45 -0000	1.264
***************
*** 1,2 ****
--- 1,7 ----
+ 2007-02-25  Baptiste Lepilleur <blep@...>
+ 
+    * doc/cookbook.dox: changed suite() to return a TestSuite instead
+      of a Test to avoid introducing unnecessary complexity.
+ 
  2007-02-24  Steve M. Robbins  <smr@...>

Update of /cvsroot/cppunit/cppunit/doc
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv13323/doc

Modified Files:
	cookbook.dox 
Log Message:
* doc/cookbook.dox: changed suite() to return a TestSuite instead
  of a Test to avoid introducing unnecessary complexity.

Index: cookbook.dox
===================================================================
RCS file: /cvsroot/cppunit/cppunit/doc/cookbook.dox,v
retrieving revision 1.8
retrieving revision 1.9
diff -C2 -d -r1.8 -r1.9
*** cookbook.dox	27 Nov 2005 18:32:31 -0000	1.8
--- cookbook.dox	25 Feb 2007 09:21:45 -0000	1.9
***************
*** 275,279 ****
  \code
  public: 
!   static CppUnit::Test *suite()
    {
      CppUnit::TestSuite *suiteOfTests = new CppUnit::TestSuite( "ComplexNumberTest" );
--- 275,279 ----
  \code
  public: 
!   static CppUnit::TestSuite *suite()
    {
      CppUnit::TestSuite *suiteOfTests = new CppUnit::TestSuite( "ComplexNumberTest" );

Update of /cvsroot/cppunit/cppunit/examples/cppunittest
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv1787/examples/cppunittest

Modified Files:
	TestAssertTest.cpp 
Log Message:
- removed wrong comment.

Index: TestAssertTest.cpp
===================================================================
RCS file: /cvsroot/cppunit/cppunit/examples/cppunittest/TestAssertTest.cpp,v
retrieving revision 1.14
retrieving revision 1.15
diff -C2 -d -r1.14 -r1.15
*** TestAssertTest.cpp	24 Feb 2007 21:13:04 -0000	1.14
--- TestAssertTest.cpp	25 Feb 2007 08:49:12 -0000	1.15
***************
*** 233,237 ****
    CPPUNIT_ASSERT_ASSERTION_PASS( CPPUNIT_ASSERT_DOUBLES_EQUAL( inf, inf, 1.0 ) );
    // NaN tests 
!   CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, 0.0, 1.0 ) ); // this one fails
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, nan, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, inf, 1.0 ) );
--- 233,237 ----
    CPPUNIT_ASSERT_ASSERTION_PASS( CPPUNIT_ASSERT_DOUBLES_EQUAL( inf, inf, 1.0 ) );
    // NaN tests 
!   CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, 0.0, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, nan, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, inf, 1.0 ) );

Update of /cvsroot/cppunit/cppunit/include/cppunit/portability
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv23570/include/cppunit/portability

Modified Files:
	FloatingPoint.h 
Log Message:
portability/FloatingPoint.h must include Portability.h

Index: FloatingPoint.h
===================================================================
RCS file: /cvsroot/cppunit/cppunit/include/cppunit/portability/FloatingPoint.h,v
retrieving revision 1.1
retrieving revision 1.2
diff -C2 -d -r1.1 -r1.2
*** FloatingPoint.h	24 Feb 2007 21:13:04 -0000	1.1
--- FloatingPoint.h	25 Feb 2007 03:54:35 -0000	1.2
***************
*** 2,5 ****
--- 2,6 ----
  #define CPPUNIT_PORTABILITY_FLOATINGPOINT_H_INCLUDED
  
+ #include <cppunit/Portability.h>
  #include <math.h>

Update of /cvsroot/cppunit/cppunit
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv23570

Modified Files:
	ChangeLog 
Log Message:
portability/FloatingPoint.h must include Portability.h

Index: ChangeLog
===================================================================
RCS file: /cvsroot/cppunit/cppunit/ChangeLog,v
retrieving revision 1.262
retrieving revision 1.263
diff -C2 -d -r1.262 -r1.263
*** ChangeLog	24 Feb 2007 21:13:03 -0000	1.262
--- ChangeLog	25 Feb 2007 03:54:34 -0000	1.263
***************
*** 1,2 ****
--- 1,6 ----
+ 2007-02-24  Steve M. Robbins  <smr@...>
+ 
+ 	* include/cppunit/portability/FloatingPoint.h: Include Portability.h.
+ 
  2007-02-24  Baptiste Lepilleur <blep@...>

Update of /cvsroot/cppunit/cppunit
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv9552

Modified Files:
	ChangeLog 
Log Message:
* src/cppunit/TestAssert.cpp (assertDoubleEquals): Moved finite & NaN 
  tests to include/cppunit/portability/FloatingPoint.h. Changed 
  implementation assertDoubleEquals to explicitly test for NaN 
  in case of non-finite values to force equality failure in the 
  presence of NaN. Previous implementation failed on Microsoft 
  Visual Studio 6 (on this platform: NaN == NaN).
* examples/cppunittest/TestAssertTest.cpp: Add more unit tests to
  test the portable floating-point primitive. Added missing 
  include <limits>.

* include/cppunit/portability/Makefile.am:
* include/cppunit/portability/FloatingPoint.h: Added file. Extracted
  isfinite() from TestAssert.cpp.

* include/cppunit/config-evc4:
* include/cppunit/config-msvc6: Added support for _finite().

Index: ChangeLog
===================================================================
RCS file: /cvsroot/cppunit/cppunit/ChangeLog,v
retrieving revision 1.261
retrieving revision 1.262
diff -C2 -d -r1.261 -r1.262
*** ChangeLog	31 Jan 2007 05:27:06 -0000	1.261
--- ChangeLog	24 Feb 2007 21:13:03 -0000	1.262
***************
*** 1,2 ****
--- 1,21 ----
+ 2007-02-24  Baptiste Lepilleur <blep@...>
+ 
+ 	* src/cppunit/TestAssert.cpp (assertDoubleEquals): Moved finite & NaN 
+      tests to include/cppunit/portability/FloatingPoint.h. Changed 
+      implementation assertDoubleEquals to explicitly test for NaN 
+      in case of non-finite values to force equality failure in the 
+      presence of NaN. Previous implementation failed on Microsoft 
+      Visual Studio 6 (on this platform: NaN == NaN).
+ 	* examples/cppunittest/TestAssertTest.cpp: Add more unit tests to
+      test the portable floating-point primitive. Added missing 
+      include <limits>.
+ 
+    * include/cppunit/portability/Makefile.am:
+    * include/cppunit/portability/FloatingPoint.h: Added file. Extracted
+      isfinite() from TestAssert.cpp.
+ 
+    * include/cppunit/config-evc4:
+    * include/cppunit/config-msvc6: Added support for _finite().
+ 
  2007-01-30  Steve M. Robbins  <smr@...>

Update of /cvsroot/cppunit/cppunit/include/cppunit/portability
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv9552/include/cppunit/portability

Modified Files:
	Makefile.am 
Added Files:
	FloatingPoint.h 
Log Message:
* src/cppunit/TestAssert.cpp (assertDoubleEquals): Moved finite & NaN 
  tests to include/cppunit/portability/FloatingPoint.h. Changed 
  implementation assertDoubleEquals to explicitly test for NaN 
  in case of non-finite values to force equality failure in the 
  presence of NaN. Previous implementation failed on Microsoft 
  Visual Studio 6 (on this platform: NaN == NaN).
* examples/cppunittest/TestAssertTest.cpp: Add more unit tests to
  test the portable floating-point primitive. Added missing 
  include <limits>.

* include/cppunit/portability/Makefile.am:
* include/cppunit/portability/FloatingPoint.h: Added file. Extracted
  isfinite() from TestAssert.cpp.

* include/cppunit/config-evc4:
* include/cppunit/config-msvc6: Added support for _finite().

Index: Makefile.am
===================================================================
RCS file: /cvsroot/cppunit/cppunit/include/cppunit/portability/Makefile.am,v
retrieving revision 1.3
retrieving revision 1.4
diff -C2 -d -r1.3 -r1.4
*** Makefile.am	25 Jun 2004 10:54:47 -0000	1.3
--- Makefile.am	24 Feb 2007 21:13:04 -0000	1.4
***************
*** 2,9 ****
  
  libcppunitinclude_HEADERS = \
! 	CppUnitDeque.h \
! 	CppUnitMap.h \
! 	CppUnitSet.h \
! 	CppUnitStack.h \
  	CppUnitVector.h \
!   Stream.h
--- 2,10 ----
  
  libcppunitinclude_HEADERS = \
! 	CppUnitDeque.h  \
! 	CppUnitMap.h    \
! 	CppUnitSet.h    \
! 	CppUnitStack.h  \
  	CppUnitVector.h \
!    FloatingPoint.h \
!    Stream.h

--- NEW FILE: FloatingPoint.h ---
#ifndef CPPUNIT_PORTABILITY_FLOATINGPOINT_H_INCLUDED
#define CPPUNIT_PORTABILITY_FLOATINGPOINT_H_INCLUDED

#include <math.h>

CPPUNIT_NS_BEGIN

/// \brief Tests if a floating-point is a NaN.
// According to IEEE-754 floating point standard, NaN comparison should always
// be 'false'. 
// At least Microsoft Visual Studio 6 is known not to implement this test correctly.
// It emits the following code to test equality:
//  fcomp       qword ptr [nan]
//  fnstsw      ax                        // copie fp (floating-point) status register to ax
//  test        ah,40h                    // test bit 14 of ax (0x4000) => C3 of fp status register
// According to the following documentation on the x86 floating point status register,
// the C2 bit should be tested to test for NaN value. 
// http://webster.cs.ucr.edu/AoA/Windows/HTML/RealArithmetic.html#1000117
// In Microsoft Visual Studio 2003 & 2005, the test is implemented with:
//  test        ah,44h         // Visual Studio 2005 test both C2 & C3...
//
// To work around this, a NaN is assumed to be detected if no strict ordering is found.
inline bool floatingPointIsUnordered( double x )
{
   // x != x will detect a NaN on conformant platform
   // (2.0 < x  &&  x < 1.0) will detect a NaN on non conformant platform:
   // => no ordering can be found for x.
   return  (x != x) ||  (2.0 < x  &&  x < 1.0);
}


/// \brief Tests if a floating-point is finite.
/// @return \c true if x is neither a NaN, nor +inf, nor -inf, \c false otherwise.
inline bool floatingPointIsFinite( double x )
{
#if defined(CPPUNIT_HAVE_ISFINITE)
   return (bool)isfinite( x );
#elif defined(CPPUNIT_HAVE_FINITE)
   return (bool)finite( x );
#elif defined(CPPUNIT_HAVE__FINITE)
   return _finite(x);
#else
   double testInf = x * 0.0;  // Produce 0.0 if x is finite, a NaN otherwise.
   return testInf == 0.0  &&  !floatingPointIsUnordered(testInf);
#endif
}

CPPUNIT_NS_END

#endif // CPPUNIT_PORTABILITY_FLOATINGPOINT_H_INCLUDED

Update of /cvsroot/cppunit/cppunit/include/cppunit/config
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv9552/include/cppunit/config

Modified Files:
	config-evc4.h config-msvc6.h 
Log Message:
* src/cppunit/TestAssert.cpp (assertDoubleEquals): Moved finite & NaN 
  tests to include/cppunit/portability/FloatingPoint.h. Changed 
  implementation assertDoubleEquals to explicitly test for NaN 
  in case of non-finite values to force equality failure in the 
  presence of NaN. Previous implementation failed on Microsoft 
  Visual Studio 6 (on this platform: NaN == NaN).
* examples/cppunittest/TestAssertTest.cpp: Add more unit tests to
  test the portable floating-point primitive. Added missing 
  include <limits>.

* include/cppunit/portability/Makefile.am:
* include/cppunit/portability/FloatingPoint.h: Added file. Extracted
  isfinite() from TestAssert.cpp.

* include/cppunit/config-evc4:
* include/cppunit/config-msvc6: Added support for _finite().

Index: config-evc4.h
===================================================================
RCS file: /cvsroot/cppunit/cppunit/include/cppunit/config/config-evc4.h,v
retrieving revision 1.2
retrieving revision 1.3
diff -C2 -d -r1.2 -r1.3
*** config-evc4.h	25 Jun 2004 12:15:24 -0000	1.2
--- config-evc4.h	24 Feb 2007 21:13:04 -0000	1.3
***************
*** 61,64 ****
--- 61,69 ----
  #endif
  
+ /* define to 1 if the compiler has _finite() */
+ #ifndef CPPUNIT_HAVE__FINITE
+ #define CPPUNIT_HAVE__FINITE 1 
+ #endif
+ 
  // Uncomment to turn on STL wrapping => use this to test compilation. 
  // This will make CppUnit subclass std::vector & co to provide default

Index: config-msvc6.h
===================================================================
RCS file: /cvsroot/cppunit/cppunit/include/cppunit/config/config-msvc6.h,v
retrieving revision 1.10
retrieving revision 1.11
diff -C2 -d -r1.10 -r1.11
*** config-msvc6.h	30 Jul 2005 07:02:38 -0000	1.10
--- config-msvc6.h	24 Feb 2007 21:13:04 -0000	1.11
***************
*** 65,68 ****
--- 65,73 ----
  #define CPPUNIT_HAVE_CPP_CAST 1
  
+ /* define to 1 if the compiler has _finite() */
+ #ifndef CPPUNIT_HAVE__FINITE
+ #define CPPUNIT_HAVE__FINITE 1 
+ #endif
+ 
  
  // Uncomment to turn on STL wrapping => use this to test compilation.

Update of /cvsroot/cppunit/cppunit/src/cppunit
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv9552/src/cppunit

Modified Files:
	TestAssert.cpp cppunit.dsp 
Log Message:
* src/cppunit/TestAssert.cpp (assertDoubleEquals): Moved finite & NaN 
  tests to include/cppunit/portability/FloatingPoint.h. Changed 
  implementation assertDoubleEquals to explicitly test for NaN 
  in case of non-finite values to force equality failure in the 
  presence of NaN. Previous implementation failed on Microsoft 
  Visual Studio 6 (on this platform: NaN == NaN).
* examples/cppunittest/TestAssertTest.cpp: Add more unit tests to
  test the portable floating-point primitive. Added missing 
  include <limits>.

* include/cppunit/portability/Makefile.am:
* include/cppunit/portability/FloatingPoint.h: Added file. Extracted
  isfinite() from TestAssert.cpp.

* include/cppunit/config-evc4:
* include/cppunit/config-msvc6: Added support for _finite().

Index: cppunit.dsp
===================================================================
RCS file: /cvsroot/cppunit/cppunit/src/cppunit/cppunit.dsp,v
retrieving revision 1.54
retrieving revision 1.55
diff -C2 -d -r1.54 -r1.55
*** cppunit.dsp	13 Oct 2005 20:35:50 -0000	1.54
--- cppunit.dsp	24 Feb 2007 21:13:04 -0000	1.55
***************
*** 248,251 ****
--- 248,255 ----
  # Begin Source File
  
+ SOURCE=..\..\include\cppunit\portability\FloatingPoint.h
+ # End Source File
+ # Begin Source File
+ 
  SOURCE=..\..\include\cppunit\Portability.h
  # End Source File

Index: TestAssert.cpp
===================================================================
RCS file: /cvsroot/cppunit/cppunit/src/cppunit/TestAssert.cpp,v
retrieving revision 1.13
retrieving revision 1.14
diff -C2 -d -r1.13 -r1.14
*** TestAssert.cpp	27 Jan 2007 05:45:08 -0000	1.13
--- TestAssert.cpp	24 Feb 2007 21:13:04 -0000	1.14
***************
*** 1,18 ****
  #include <cppunit/TestAssert.h>
! 
! #include <math.h>
! 
! #if !defined(CPPUNIT_HAVE_ISFINITE)
! 
!     static inline bool isfinite( double x )
!     {
! #if defined(CPPUNIT_HAVE_FINITE)
! 	return finite( x );
! #else
!         return ( x * 0.0 ) == 0.0;
! #endif
!     }
! 
! #endif
  
  CPPUNIT_NS_BEGIN
--- 1,4 ----
  #include <cppunit/TestAssert.h>
! #include <cppunit/portability/FloatingPoint.h>
  
  CPPUNIT_NS_BEGIN
***************
*** 31,38 ****
  
    bool equal;
!   if ( isfinite(expected) && isfinite(actual) )
        equal = fabs( expected - actual ) <= delta;
    else
!       equal = expected == actual;
  
    Asserter::failNotEqualIf( !equal,
--- 17,37 ----
  
    bool equal;
!   if ( floatingPointIsFinite(expected)  &&  floatingPointIsFinite(actual) )
        equal = fabs( expected - actual ) <= delta;
    else
!   {
!     // If expected or actual is not finite, it may be +inf, -inf or NaN (Not a Number).
!     // Value of +inf or -inf leads to a true equality regardless of delta if both
!     // expected and actual have the same value (infinity sign).
!     // NaN Value should always lead to a failed equality.
!     if ( floatingPointIsUnordered(expected)  ||  floatingPointIsUnordered(actual) )
!     { 
!        equal = false;  // expected or actual is a NaN
!     }
!     else // ordered values, +inf or -inf comparison
!     {
!        equal = expected == actual;
!     }
!   }
  
    Asserter::failNotEqualIf( !equal,

Update of /cvsroot/cppunit/cppunit/examples/cppunittest
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv9552/examples/cppunittest

Modified Files:
	CppUnitTestMain.dsp TestAssertTest.cpp 
Log Message:
* src/cppunit/TestAssert.cpp (assertDoubleEquals): Moved finite & NaN 
  tests to include/cppunit/portability/FloatingPoint.h. Changed 
  implementation assertDoubleEquals to explicitly test for NaN 
  in case of non-finite values to force equality failure in the 
  presence of NaN. Previous implementation failed on Microsoft 
  Visual Studio 6 (on this platform: NaN == NaN).
* examples/cppunittest/TestAssertTest.cpp: Add more unit tests to
  test the portable floating-point primitive. Added missing 
  include <limits>.

* include/cppunit/portability/Makefile.am:
* include/cppunit/portability/FloatingPoint.h: Added file. Extracted
  isfinite() from TestAssert.cpp.

* include/cppunit/config-evc4:
* include/cppunit/config-msvc6: Added support for _finite().

Index: TestAssertTest.cpp
===================================================================
RCS file: /cvsroot/cppunit/cppunit/examples/cppunittest/TestAssertTest.cpp,v
retrieving revision 1.13
retrieving revision 1.14
diff -C2 -d -r1.13 -r1.14
*** TestAssertTest.cpp	31 Jan 2007 05:27:07 -0000	1.13
--- TestAssertTest.cpp	24 Feb 2007 21:13:04 -0000	1.14
***************
*** 1,5 ****
--- 1,7 ----
  #include "CoreSuite.h"
  #include "TestAssertTest.h"
+ #include <cppunit/portability/FloatingPoint.h>
  #include <algorithm>
+ #include <limits>
  
  /*
***************
*** 193,206 ****
  }
  
  void 
  TestAssertTest::testAssertDoubleNonFinite()
  {
    double inf = std::numeric_limits<double>::infinity();
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( inf, 0.0, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( 0.0, inf, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_PASS( CPPUNIT_ASSERT_DOUBLES_EQUAL( inf, inf, 1.0 ) );
! 
!   double nan = std::numeric_limits<double>::quiet_NaN();
!   CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, 0.0, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, nan, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, inf, 1.0 ) );
--- 195,237 ----
  }
  
+ 
  void 
  TestAssertTest::testAssertDoubleNonFinite()
  {
    double inf = std::numeric_limits<double>::infinity();
+   double nan = std::numeric_limits<double>::quiet_NaN();
+   // test our portable floating-point primitives that detect NaN values
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsUnordered( nan ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( inf ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( -inf ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( 1.0 ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( 1.5 ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( 2.0 ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( 2.5 ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( 0.0 ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( -1.0 ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsUnordered( -2.0 ) );
+   // test our portable floating-point primitives that detect finite values
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsFinite( 0.0 ) );
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsFinite( 0.5 ) );
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsFinite( 1.0 ) );
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsFinite( 1.5 ) );
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsFinite( 2.0 ) );
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsFinite( 2.5 ) );
+   CPPUNIT_ASSERT( CPPUNIT_NS::floatingPointIsFinite( -1.5 ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsFinite( nan ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsFinite( inf ) );
+   CPPUNIT_ASSERT( !CPPUNIT_NS::floatingPointIsFinite( -inf ) );
+   // Infinity tests
+   CPPUNIT_ASSERT( inf == inf );
+   CPPUNIT_ASSERT( -inf == -inf );
+   CPPUNIT_ASSERT( -inf != inf );
+   CPPUNIT_ASSERT( -inf < inf );
+   CPPUNIT_ASSERT( inf > -inf );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( inf, 0.0, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( 0.0, inf, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_PASS( CPPUNIT_ASSERT_DOUBLES_EQUAL( inf, inf, 1.0 ) );
!   // NaN tests 
!   CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, 0.0, 1.0 ) ); // this one fails
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, nan, 1.0 ) );
    CPPUNIT_ASSERT_ASSERTION_FAIL( CPPUNIT_ASSERT_DOUBLES_EQUAL( nan, inf, 1.0 ) );

Index: CppUnitTestMain.dsp
===================================================================
RCS file: /cvsroot/cppunit/cppunit/examples/cppunittest/CppUnitTestMain.dsp,v
retrieving revision 1.27
retrieving revision 1.28
diff -C2 -d -r1.27 -r1.28
*** CppUnitTestMain.dsp	13 Oct 2005 20:35:51 -0000	1.27
--- CppUnitTestMain.dsp	24 Feb 2007 21:13:03 -0000	1.28
***************
*** 177,180 ****
--- 177,188 ----
  # Begin Source File
  
+ SOURCE=.\assertion_traitsTest.cpp
+ # End Source File
+ # Begin Source File
+ 
+ SOURCE=.\assertion_traitsTest.h
+ # End Source File
+ # Begin Source File
+ 
  SOURCE=.\ExceptionTest.cpp
  # End Source File

Update of /cvsroot/cppunit/cppunit/include/cppunit
In directory sc8-pr-cvs4.sourceforge.net:/tmp/cvs-serv9552/include/cppunit

Modified Files:
	TestAssert.h 
Log Message:
* src/cppunit/TestAssert.cpp (assertDoubleEquals): Moved finite & NaN 
  tests to include/cppunit/portability/FloatingPoint.h. Changed 
  implementation assertDoubleEquals to explicitly test for NaN 
  in case of non-finite values to force equality failure in the 
  presence of NaN. Previous implementation failed on Microsoft 
  Visual Studio 6 (on this platform: NaN == NaN).
* examples/cppunittest/TestAssertTest.cpp: Add more unit tests to
  test the portable floating-point primitive. Added missing 
  include <limits>.

* include/cppunit/portability/Makefile.am:
* include/cppunit/portability/FloatingPoint.h: Added file. Extracted
  isfinite() from TestAssert.cpp.

* include/cppunit/config-evc4:
* include/cppunit/config-msvc6: Added support for _finite().

Index: TestAssert.h
===================================================================
RCS file: /cvsroot/cppunit/cppunit/include/cppunit/TestAssert.h,v
retrieving revision 1.29
retrieving revision 1.30
diff -C2 -d -r1.29 -r1.30
*** TestAssert.h	6 Nov 2005 16:58:21 -0000	1.29
--- TestAssert.h	24 Feb 2007 21:13:04 -0000	1.30
***************
*** 110,113 ****
--- 110,114 ----
   * Use CPPUNIT_ASSERT_DOUBLES_EQUAL instead of this function.
   * \sa Asserter::failNotEqual().
+  * \sa CPPUNIT_ASSERT_DOUBLES_EQUAL for detailed semantic of the assertion.
   */
  void CPPUNIT_API assertDoubleEquals( double expected,
***************
*** 219,222 ****
--- 220,230 ----
  /*! \brief Macro for primitive double value comparisons. 
   * \ingroup Assertions
+  *
+  * The assertion pass if both expected and actual are finite and
+  * \c fabs( \c expected - \c actual ) <= \c delta.
+  * If either \c expected or actual are infinite (+/- inf), the 
+  * assertion pass if \c expected == \c actual.
+  * If either \c expected or \c actual is a NaN (not a number), then
+  * the assertion fails.
   */
  #define CPPUNIT_ASSERT_DOUBLES_EQUAL(expected,actual,delta)        \
***************
*** 231,234 ****
--- 239,243 ----
   * user-supplied message in case of failure. 
   * \ingroup Assertions
+  * \sa CPPUNIT_ASSERT_DOUBLES_EQUAL for detailed semantic of the assertion.
   */
  #define CPPUNIT_ASSERT_DOUBLES_EQUAL_MESSAGE(message,expected,actual,delta)  \