RslArg Class Reference

#include <RslPlugin.h>

Inheritance diagram for RslArg:

Public Member Functions
virtual	~RslArg ()
	Virtual destructor.
Public Member Functions inherited from RslArg_v7
virtual bool	IsFilterRegion () const =0
	Returns true if this argument is a filter region.
virtual	~RslArg_v7 ()
	Virtual destructor.
Public Member Functions inherited from RslArg_v6
virtual	~RslArg_v6 ()
	Virtual destructor.
Public Member Functions inherited from RslArg_v5
virtual bool	IsStruct () const =0
	Returns true if this argument is a struct (or struct array).
virtual const char *	GetName () const =0
virtual bool	IsResizable () const =0
	Returns true if this argument is a resizable array.
virtual class RslResizer *	GetResizer () const =0
	Get resizer for this argument (returns NULL if not resizable).
virtual	~RslArg_v5 ()
	Virtual destructor.
Public Member Functions inherited from RslArg_v3
virtual bool	IsNormal () const =0
	Returns true if this argument is a normal (or normal array).
virtual bool	IsWriteable () const =0
virtual	~RslArg_v3 ()
	Virtual destructor.
Public Member Functions inherited from RslArg_v2
virtual	~RslArg_v2 ()
	Virtual destructor.
virtual bool	IsFloat () const =0
	Returns true if this argument is a float (or float array).
virtual bool	IsPoint () const =0
	Returns true if this argument is a point (or point array).
virtual bool	IsVector () const =0
	Returns true if this argument is a vector (or vector array).
virtual bool	IsColor () const =0
	Returns true if this argument is a color (or color array).
virtual bool	IsString () const =0
	Returns true if this argument is a string (or string array).
virtual bool	IsMatrix () const =0
	Returns true if this argument is a matrix (or matrix array).
virtual bool	IsArray () const =0
	Returns true if this argument is an array.
virtual bool	IsVarying () const =0
	Returns true if this argument is varying.
virtual int	GetArrayLength () const =0
virtual unsigned int	NumValues () const =0
virtual void	GetData (float **data, int *stride) const =0

Additional Inherited Members
Static Public Member Functions inherited from RslArg_v2
static unsigned int	NumValues (int argc, const class RslArg **argv)

Detailed Description

An array of RslArg pointers is passed to the RSL plugin function. The result parameter is always the zeroth argument (even in a void function).

Various methods can be used to query the type of an RslArg (e.g. IsFloat(), IsArray()), its detail (IsVarying()), and other information (GetArrayLength(), IsWriteable()).

An iterator (RslIter) must be constructed to access the data represented by an RslArg. The type of iterator depends on the underlying data type. For example:

assert(argv[0].IsFloat());
assert(argv[1].IsArray() && argv[1].IsColor());
RslFloatIter num(argv[0]);
RslColorArrayIter colors(argv[1]);

If an argument might be varying, the NumValues() method is used to determine the number of iterations required to process it. Any arguments that are varying require the same number of iterations. If the function returns a result (i.e. does not have a void return type), the shader compiler guarantees that the result argument (argv[0]) is varying whenever any of the other arguments are varying. It's usually sufficient to use "argv[0]->NumValues()" as the required number of iterations. For example:

int n = argv[0]->NumValues();
for (int i = 0; i < n; ++i) {
     a = *b + *c;
    ++a; ++b; ++c;
}

Void plugin functions must take the detail of all the arguments into account, however:

int n = RslArg::NumValues(argc, argv);

Note that incrementing an iterator for a uniform argument has no effect. Also, the number of iterations might be one if all the arguments are uniform. In our experience, it is usually not worth optimizing for such cases. If performance is a concern, overloading can be used to define a specialized plugin function that operates on all uniform arguments.

---

The documentation for this class was generated from the following file:

• RslPlugin.h