Edit

Share via

Casting and Conversion

  • Article

When transferring values between the host application and an effect parameter, data is written as expected when the source data type (in the host application) matches the destination data type (in the effect parameter). When the data types differ, casting will occur as the source data is rearranged to fit the destination.

DXSAS defines the following type conversion rules. These are a superset of the effect (.fx) and HLSL type conversion rules. DXSAS also defines some Parameter Value Modifiers that can affect the value being transferred from the host application to a bound parameter.

Type Casting

The following table lists the casting that will occur when one data type is transferred to another data type:

Source Type Destination Type Behavior
float int Round towards zero.
float, int bool Values not equal to 0 - based on a not-equal-to comparison to 0 (int) or 0.0 (float) - are considered to be true, otherwise the value is considered to be false.
int float
bool int, float False is converted to zero.True is converted to one.
texture1D, texture2D, texture3D, textureCUBE texture Destination texture is treated as the source texture type.
string texture1D, texture2D, texture3D, textureCUBE String value is treated as a resource address and resolved in the same way as the Parameter Initialization Annotation. If the resource address cannot be resolved, the cast is invalid and the host applications must return an error. If the resource is resolved properly, the type of the expression is treated as the same type as the resolved resource.

 

Class Casting

In addition to the type casting rules described above, DXSAS defines the set of casting rules necessary to convert between class types. Column matrices (N x 1), row matrices (1 x N), and numeric structures are treated as vectors.

Effect matrix parameters and HLSL matrix variables can define whether the value is a row-major or column-major matrix; however, the DirectX APIs always treat D3DMATRIX and D3DXMATRIX as row-major.

Source Class Destination Class Behavior
Scalar Scalar Apply Type Casting.
Scalar Vector Replicate the scalar source value into every component of the destination vector after applying the type casting and conversion behavior described in Type Casting.
Scalar Matrix Replicate the scalar source value into every component of the destination matrix after applying the type casting and conversion behavior described in Type Casting.
Scalar Object Invalid cast. Host applications must return an error.
Scalar Structure Valid only if the destination structure contains only numeric elements. If valid, replicate the scalar source value into every component of the destination structure after applying the type casting and conversion behavior described in Type Casting.
Vector Scalar The destination scalar receives the value of the first component of the source vector after applying the type casting and conversion behavior described in Type Casting.
Vector Vector Invalid cast if the destination vector has more components than the source vector. The destination vector receives the left-most values of the source vector after applying the type casting and conversion behavior described in Type Casting. The remaining right-most components of the source vector are lost.
Vector Matrix Invalid cast unless the source vector has the same number of components as the destination matrix. If the cast is valid, the behavior of the vector-to-vector cast is then applied.
Vector Object Invalid cast. Host applications must return an error.
Vector Structure Valid only if the destination structure has only numeric elements and does not contain more elements than the source vector has components. The destination structure's elements receive the left-most components of the source vector after applying the type casting and conversion behavior described in Type Casting.
Matrix Scalar The destination scalar receives the value of the upper-left-most value of the source matrix after applying the type casting and conversion behavior described in Type Casting.
Matrix Vector Invalid cast unless the source matrix has the same number of components as the destination vector. If the cast is valid, the behavior of the vector-to-vector cast above is then applied.
Matrix Matrix Invalid cast if the destination matrix has more components than the source matrix. The destination matrix receives the upper-left-most values of the source matrix after applying the type casting and conversion behavior described in Type Casting. The remaining lower-right-most components of the source matrix are lost.
Matrix Object Invalid cast. Host applications must return an error.
Matrix Structure The size of the structure must be equal to the size of the matrix and all components of the structure must be numeric.
Object Scalar Invalid cast. Host applications must return an error.
Object Vector Invalid cast. Host applications must return an error.
Object Matrix Invalid cast. Host applications must return an error.
Object Object Valid if the types of the objects are identical and in accordance with the behavior defined in Type Casting.
Structure Scalar Valid if the source structure contains at least one numeric member. The destination scalar receives the value of the source structure's first numeric member after applying the type casting and conversion behavior described in Type Casting.
Structure Vector The source structure must be at least the size of the vector. The first components must be numeric up to the size of the destination vector.
Structure Matrix The source structure must be at least the size of the vector. The first components must be numeric up to the size of the destination vector.
Structure Structure The destination structure must not be larger than the source structure. A valid cast must exist between all respective source and destination components.

 

Parameter Value Modifiers

Parameter modifier annotations add additional information to a parameter so the parameter's data can be interpreted properly. For instance, a vector may need to be expressed with normalized data or a length may be measured in inches. Parameter value modifier annotations express this additional information so that host applications can transform values correctly when data is transferred to an effect parameter.

These are the parameter modifiers:

Parameter Value Modifier Annotations Description
SasNormalize Specify whether a vector is normalized or not.
SasUnits Specify the units of measure for a parameter.

 

SasNormalize

The SasNormalize annotation denotes that the associated parameter should be a normalized value whenever it is assigned. This annotation only affects float2, float3, and float4 parameters.

string SasNormalize = "Value";

where Value is either True or False.

Here is an example:

float3 UpNormal
<
  bool SasNormalize = "True";
>;

SasUnits

The effect parameter data is in the following units:

string SasUnits = "Value";

where Value is one of the following:

Measurement Type Value Description
No Units empty string No units
Distance mm Millimeters
cm Centimeters
m Meters
km Kilometers
Angle rad Radians
Time ms Milliseconds
sec Seconds
min Minutes
hr Hours
Linear Velocity mm/sec Millimeters per second
cm/sec Centimeters per second
m/sec Meters per second
m/hr Meters per hour
km/hr Kilometers per hour
Linear Acceleration mm/sec² Millimeters per second squared
cm/sec² Centimeters per second squared
m/sec² Meters per second squared
m/hr² Meters per hour squared
km/hr² Kilometers per hour squared
Angular Velocity rad/sec Radians per second
Angular Acceleration rad/sec² Radians per second squared
Area mm² Millimeters squared
cm² Centimeters squared
m² Meters squared
km² Kilometers squared
Volume mm³ Millimeters cubed
cm³ Centimeters cubed
m³ Meters cubed
km³ Kilometers cubed

 

DirectX Standard Annotations and Semantics Reference