FIX: Floating Point Field Incorrectly Assigned -9.123E+19

The information in this article applies to:

• The Microsoft Foundation Classes (MFC) included with: Microsoft Visual C++ for Windows, versions 1.51 and 1.52

SYMPTOMS

When you use the RFX_Single() function and set a field to null, the field may actually be set to -9.123E+19 rather than null. For example, the following code doesn't set the field to null as expected.

   CYourRecordset rs;
   rs.Open();
   rs.Edit();
   rs.SetFieldNull(&rs.m_floatfield);
   rs.Update();
   rs.Close();

CAUSE

The problem is caused by a bug in the MFC database classes. Specifically, the RFX_Single() function performs an incorrect check to see if a field is null.

Looking at DBFLT.CPP in the MSVC15\MFC\SRC directory, you'll see this:

   case CFieldExchange::MarkForUpdate:
        if (value != AFX_RFX_SINGLE_PSEUDO_NULL)
            pFX->m_prs->ClearFieldFlags(nField,
                AFX_SQL_FIELD_FLAG_NULL, pFX->m_nFieldType);
        goto LDefault;

   #define AFX_RFX_SINGLE_PSEUDO_NULL (-9.123e19)

   #define AFX_RFX_SINGLE_PSEUDO_NULL (-9.123e19f)

RESOLUTION

To work around this problem, follow these steps:

1. Copy the RFX_Single() function from DBFLT.CPP into the CRecordset's .CPP file and rename the function to something like RFX_Single2.

2. Include the following code before the RFX_Single2() function:

   #undef AFX_RFX_SINGLE_PSEUDO_NULL

      #define AFX_RFX_SINGLE_PSEUDO_NULL (-9.123e19f)
   
   

3. Change the call to RFX_Single() in the CRecordset's DoFieldExchange() function to RFX_Single2(), and move the function call out of the ClassWizard tagged section. The ClassWizard tagged section is the region between:

         //{{AFX_FIELD_MAP(CYourRecordset)
   

   and:

         //}}AFX_FIELD_MAP
   

STATUS

Microsoft has confirmed this to be a bug in the Microsoft products listed at the beginning of this article. This problem was corrected in Microsoft Visual C++, 32-bit Edition, version 2.2.