ZBasic Language Reference
106
ZBasic Microcontrollers
relationships between values of multiple data members. Such control would be difficult or impossible to
enforce if outside entities were allowed to directly modify the data members. Moreover, requiring the use
of mutator methods to change data members also allows the freedom to later change the way that the
data members are stored without affecting any of the object’s users. An example of enforcing data value
ranges is shown below in the re-written SetTime() method. The class definition has also been
augmented with accessor methods for the three data members.
Class MyTime
Private hour as Byte
Private minute as Byte
Private seconds as Single
Public Sub SetTime(ByVal h as Byte, ByVal m as Byte, ByVal s as Single)
hour = Min(h, 23)
minute = Min(m, 59)
Seconds = Min(s, 59.999)
End Sub
Public Function GetHour() as Byte
GetHour = hour
End Function
Public Function GetMinute() as Byte
GetMinute = minute
End Function
Public Function GetSeconds() as Single
GetSeconds = seconds
End Function
End Class
4.4 Object Creation Issues
When an object is created, the data members have initial values that depend on where the object is
defined. This is exactly the same situation as with non-object variables. For example, if an Integer
variable is defined at the module level, its value is guaranteed to be initially zero. In contrast, if an
Integer variable is defined within a subroutine or function (without the Static attribute), its initial value
is indeterminate. String variables (both the allocated type and BoundedString type) have slightly
different rules in that they are guaranteed initally to have zero-length string values no matter where the
variable is defined. Similarly, fixed-length strings are guaranteed to have an initial value of a space-filled
string with the specified length. These rules apply to the data members of objects (just as they do with
structures), again depending on where the object is defined.
Beyond these fundamental initialization guarantees, when you define a class you can also define optional
constructor methods that initialize the values of some or all of the data members. A constructor method is
a subroutine that has the special name _Create. You may define several different constructor methods,
all having the name _Create, but each of the constructor overloads must have different formal
parameter lists so that the compiler can distinguish between them.
The expanded MyTime class definition below includes several constructor methods and the additional
code shows how a particular constructor can be specified when an object is instantiated. It is important to
note that when a particular constructor executes, the fundamental initialization (described above) will
already have been performed.
Class MyTime
Private hour as Byte
Private minute as Byte
Private seconds as Single
 |
 |
 |