The function pfNewShaderObject() creates and returns a handle to a pfShaderProgram. The value arena specifies a malloc() arena out of which the pfShaderProgram is allocated or NULL for allocation off the process heap. You can delete pfShaderPrograms with pfDelete().

The function new(arena) allocates a pfShaderProgram from the specified memory arena, or from the heap if arena is NULL. The function allocates a pfShaderProgram from the default memory arena (see the pfGetSharedArena man page). Like other pfObjects, pfShaderPrograms cannot be automatically created statically on the stack or in arrays. Delete pfShaderPrograms with pfDelete() rather than with the delete operator.

The function pfGetShaderObjectClassType() returns the pfType* for the class pfShaderProgram. The pfType* returned by pfGetShaderObjectClassType() is the same as the pfType* returned by invoking pfGetType(), the virtual function getType() on any instance of class pfShaderProgram. Because OpenGL Performer allows subclassing of built-in types when decisions are made based on the type of an object, use pfIsOfType() the member function isOfType() to test if an object is of a type derived from an OpenGL Performer type rather than to test for strict equality of the pfType*s.

In order to use a pfShaderProgram as a piece of state, you must specify it as an attribute for a pfGeoState and enable that mode, as shown in the following code:

pfGeoState *gState = pfNewGState(arena);
pfShaderProgram *sProg = pfNewSProg(arena);

pfGStateMode(gState, PFSTATE_ENSHADPROG, PF_ON);
pfGStateAttr(gState, PFSTATE_SHADPROG, sProg);

Creating a valid pfShaderProgram involves the following steps:

int uniformIndex;
GLfloat scale = 0.5f;
pfShaderProgram *sProg = pfNewSprog(arena);

uniformIndex = pfSProgAddUniform(sProg, "scaleFactor", PFUNI_FLOAT1, 1, &scale);

GLfloat minValue = 0.0f;
GLfloat maxValue = 1.0f;

pfSProgUniformMin(sProg, uniformIndex, &minValue);
pfSProgUniformMax(sProg, uniformIndex, &maxValue);

You can apply a pfShaderProgram using pfShaderProgramApply(), but only in the draw process. When this operation is performed, OpenGL Performer will determine if the shader program needs to be recompiled and perform that operation if required. It is also possible to force compilation to occur by calling pfSProgLink(). This method will return the following:

In the case where one would like to force OpenGL Performer to relink a pfShaderProgram once, call pfSProgForceRelink().

One can verify if a given pfShaderProgram is in a state where it is ready to be applied by calling pfSProgValidate(), which will return 1 if the program can be applied given the current state or 0, otherwise.

The OpenGL handle associated with a given pfShaderProgram can be retrieved using pfGetSProgHandle().

The function pfNewShaderObject() creates a new shader object and returns a handle to a pfShaderObject. The value arena specifies a malloc() arena out of which the pfShaderObject is allocated or NULL for allocation off the process heap. You can delete pfShaderObjects with pfDelete().

The function call new(arena) allocates a pfShaderObject from the specified memory arena or from the heap if arena is NULL. The function allocates a pfShaderObject from the default memory arena (see the pfGetSharedArena man page). Like other pfObjects, pfShaderObjects cannot be automatically created statically on the stack or in arrays. Delete pfShaderObjects with pfDelete() rather than the delete operator.

The function pfGetShaderObjectClassType() returns the pfType* for the class pfShaderObject. The pfType* returned by pfGetShaderObjectClassType() is the same as the pfType* returned by invoking pfGetType(), the virtual function getType() on any instance of class pfShaderObject. Because OpenGL Performer allows subclassing of built-in types when decisions are made based on the type of an object, use  pfIsOfType(), the member function isOfType(), to test if an object is of a type derived from an OpenGL Performer type rather than to test for strict equality of the pfType*s.

A pfShaderObject can be specified either by loading the source explicitly or by simply specifying the filename parameter with pfShaderObjectName(). The location for shader objects specified by filename corresponds to the semantics of pfFindFile() and, hence, the PFPATH environment variable can be used to specify the location of source files. In order to retreive the name of the current shader source, you can use pfGetShaderObjectName(). If the return srting is NULL, it means that the source is inlined, not loaded from an external file. If the shader source is loaded from a file, you must also call pfShaderObjectLoad() in order to load the source code into the shader object.

The source code for the shader object can also be specified explicitly with pfShaderObjectSource(). The corresponding get method is pfGetShaderObjectSource().

In addition to setting the source in the form of either a filename or an ASCII string, you must also specify the shader type for a pfShaderObject. By default, this is set to –1 (invalid) and it must be set to either PFSHAD_FRAGMENT_SHADER or PFSHD_VERTEX_SHADER with pfShaderObjectShaderType(). You can also retrieve the shader type for a given pfShaderObject with pfGetShaderObjectShaderType().

If either the source code for the shader object or the type for the shader object have changed, the compilation status for the shader object will change. One can determine the necessity for recompiling a given pfShaderObject by calling pfGetShaderObjectCompileStatus(), which will return 1 if recompilation is required and 0, otherwise. This is used internally to determine if a pfShaderProgram needs to be re-linked and, hence, should not normally be called from a user-specified program.

A pfShaderObject may be compiled with pfShaderObjectCompile(). The log for the compilation process will be stored in the log parameter. If successful, the compilation will return 1 and 0, otherwise.

Once a pfShaderObject has been bound to the current graphics context, you can retrieve its GL handle with pfGetShaderObjectHandle(). The handle is created as needed during the compilation process. If the pfShaderObject has not yet been compiled, then the handle will be NULL.