Compiling MOF Code with Floating-Point Values

The MOF compiler accepts a floating-point value specified for a nonfloating-point property. The value is rounded up or down and stored as a nonfloating-point number. This situation can cause some unexpected results.

The following MOF code example defines a class called abc in a namespace called "Test". This MOF code compiles without errors, but you cannot query for the floating-point value defined for the property exampleUint16 in the instance this code creates.

#pragma namespace ("\\\\.\\Root")

instance of __Namespace
    Name = "Test";

#pragma namespace ("\\\\.\\Root\\test")

Class abc
        [KEY] String testID ;
        Uint16 exampleUint16;
        Real64 exampleReal64;

Instance of abc
        TestID ="exampleID";
        exampleUint16 = 1000.4;

If you issue the following query, you get an error code that indicates an invalid query.

SELECT * FROM abc WHERE exampleUint16 = 1000.4

However, the following query finds the indicated instance.

SELECT * FROM abc WHERE exampleUint16 = 1000

Related topics

Compiling MOF Files
Preprocessor Commands