[Ada-tut] Functions and procedures

[wtd]

Some background

Functions and procedures are a basic way of giving a name to meaningful and often repeated bits of code.

A small surprise

As we've seen, a basic Ada95 program is contained within a procedure. It may strike you as odd then that we can define functions and procedures within this procedure. In Ada95, a procedure or function may contain other functions and procedures.

A few simple examples

Let's define a function which returns 42.

code:

procedure Simple_Program is    function Forty_Two return Integer is    begin       return 42;    end Forty_Two; begin end Simple_Program;

Functions are easy, since they don't alter anything outside of them. Procedures act on external things, so we'll need to create a variable for a procedure to act on.

code:

procedure Simple_Program is    function Forty_Two return Integer is    begin       return 42;    end Forty_Two;    My_Int : Integer; begin end Simple_Program;

Now let's create a procedure which sets My_Int to 42.

code:

procedure Simple_Program is    function Forty_Two return Integer is    begin       return 42;    end Forty_Two;    My_Int : Integer;    procedure Set_My_Int is    begin       My_Int := Forty_Two;    end Set_My_Int; begin     end Simple_Program;

And calling the Set_My_Int procedure, which calls the Forty_Two function is as simple as:

code:

procedure Simple_Program is    function Forty_Two return Integer is    begin       return 42;    end Forty_Two;    My_Int : Integer;    procedure Set_My_Int is    begin       My_Int := Forty_Two;    end Set_My_Int; begin    Set_My_Int; end Simple_Program;

Parameters

Functions or procedures which can't accept data and act on it are rather useless. Fortunately Ada95 provides simple and easy syntax for passing data to functions and procedures.

Let's rewrite our previous example with a function Double_It, which returns the integer passed in, multiplied by two, and a procedure Double_Me which passes its parameter to Double_It and assigns the result to that variable.

Of course, we'll also need a vriable to pass to the procedure.

code:

procedure Simple_Program is    function Double_It(X : in Integer) return Integer is    begin       return X * 2;    end Double_It;    procedure Double_Me(X : in out Integer) is    begin       X := Double_It(X);    end Double_Me;    My_Int : Integer := 42; begin    Double_Me(My_Int); end Simple_Program;

What's with "in" and "out"?

If a parameter is specified as "in", the parameter's value can be read. If "out" is specified, then the variable can be assigned to.

In the Double_It function, X can only be read. An attempt to assign a new value to X would result in the program not compiling and a very descriptive error message.

In the same way, if only "out" is specified, then any attempt to read the value of X will result in a similar compiler error.