Echelon Neuron C - Manual (Page 98)

86 Using Configuration Properties to Configure Device Behavior
configuration property, or with the entire configuration network variable array
being handled as a single configuration property. See Instantiation of
Configuration Properties on page 88 for details.
A configuration network variable's declaration can contain an initial-value or an
initializer-list , like any other network variable declaration, as discussed in the
previous chapter. Unlike any other network variable, a configuration network
variable cannot, itself, also have a network variable property list. That is, you
cannot define configuration properties that apply to other configuration
properties.
When using a network variable array as a configuration property or properties,
particular care should be given to the compiler's rules of initialization for that
network variable array. The array elements can be initialized in the declaration,
as is the case with any variable or array variable declaration. If some or all of
the array elements are not initialized, the uninitialized elements default to a zero
initialization. However, each array element can be initialized when it appears as
a property in a properties clause, and this declaration overrides the initialization
in the declaration, but only for the element that appears in that property clause.
Similarly, if the entire network variable array is used as a single configuration
property, the entire array can be initialized when it appears as a property in a
properties clause.
You cannot change the type of a configuration property unless it inherits its type
from a changeable-type network variable.
Declaring Configuration Properties within Files
You can declare a configuration property that is to be implemented within a
configuration file using a CP family declaration. A CP family declaration can be
thought of as a
meta -declaration, defining a type construct for later use in the
program. It can be used to declare a collection of many configuration properties,
identical in type and certain other settings, but individually applying to one or
more different network variables, functional blocks (as described in Chapter 5,
Using Functional Blocks to Implement a Device Interface , on page 101), or the
device itself. A CP family can have zero members, one member, or many
members. No code or data is generated until you declare members of the CP
family, as described later. In this regard, the CP family is similar to a C
language typedef.
The syntax for declaring a CP family is shown below:
[const] type cp_family [ cp - modifiers ] family-ident [= initial - value ] ;
family-ident : identifier [ array-bound ]
identifier
Example:
SCPTgain cp_family cpGain = { 2, 3 };
The type for a CP family cannot be just a standard C type, such as int or char.
Instead, the declaration must use a configuration property type (CPT) from a
resource file. The configuration property type can either be a standard
configuration property type (SCPT) or a user configuration property type (UCPT).
There are over 300 SCPT definitions available today, and you can create your
own manufacturer-specific types using UCPTs. The SCPT definitions are stored