©2008 Felleisen, Proulx, et. al.

3  Designing Methods — Part 1

Methods for Simple Classes and Classes with Containment

For this series of exercises use the classes from Lab 1 that represented photo images and included the date and time information. Recall the class diagram:

| Photo          |
| String name    |
| String kind    |
| int width      |
| int height     |
| int bytes      |
| Date date      |--------------------+
| ClockTime time |-+                  |
+----------------+ |                  |
                   v                  v
          +------------+       +------------+
          | ClockTime  |       | Date       |
          +------------+       +------------+
          | int hour   |       | int year   |
          | int minute |       | int month  |
          +------------+       | int day    |

Recall some examples of the information we wish to represent:

You can work with these data definitions, or, if you used different names in the previous lab, you can use the code from that lab, as long as you can represent all the information given in the examples.

Note: Make sure you understand the data definitions, can translate these examples to data, and converesely, translate any instance of data defined for these classes into the information the data represents.

Design Recipe for a simple method definition

Recall from the lectures that in a class based language every method is defined in a class that is most relevant, it is then invoked by the instance of that class, and this instance becomes the first argument for the method.

Below is an example of the design of a method that computes the number of pixels in a photo image:

  1. Design the method timeToDownload that determines how long it will take to download this image, if we know the number of bytes we can download in one second.

  2. Design the method takenBefore that determines whether this picture was taken before another picture.

    Remember: One task, one method. Make each class responsible for its data.

  3. In the class ClockTime design the method that advances the time by the given number of minutes. Use helper methods as needed.

4  Quiz

5  Designing Methods – Part 2

Methods for Unions

For this part use the data for camera shots (photos or videos) from the previous lab. When designing methods for unions, we first need to define the purpose and the header in the interface that represents the union, then work on designing the method body in each class that implements the interface following the desing recipe.

  1. Design the method timeToDownload that determines how long it will take to download this shot, if we know the number of bytes we can download in one second.

  2. Design the method takenOn that determines whether this shot has been taken on a given day.

6  Designing Methods — Part 3

We continue with the the theme of the photo images. Our goal is to design methods that answer questions about list of images and manipulate these lists.

We will also work with the geometric shapes and learn to draw their images on the Canvas.

6.1  Methods for Self-Referential Data

In the previous lab you designed a list of photo images. We first design the method that counts the images in our list. (We use the simpler version of the class Photo that was defined in Lab 1.)

Note: Of course, you will quickly realize that this method will look the same regardless of what are the pieces of data contained in the list. We will address that issue later on, once we are comfortable with dealing with lists that contain specific items.

Recall that the class diagram for a list of photo images looked as follows:

               | IListOfPhoto |<----------------+
               +--------------+                 |
               +--------------+                 |
                      |                         |
                     / \                        |
                     ---                        |
                      |                         |
          ------------------------              |
          |                      |              |
  +---------------+    +-------------------+    |
  | MTListOfPhoto |    | ConsListOfPhoto   |    |
  +---------------+    +-------------------+    |
  +---------------+  +-| Photo first       |    |
                     | | IListOfPhoto rest |----+
                     | +-------------------+ 
            | Photo          |
            | String name    |
            | String kind    |
            | int width      |
            | int height     |
            | int bytes      |

Design Recipe for a method definition for unions of data

Below is an example of the design of a method that counts the number of pictures in a list of photo images.

The method deals with IListOfPhotos. We have an interface IListOfPhotos and two classes that implement the interface, MTListOfPhotos and ConsListOfPhotos. When the DESIGN RECIPE calls for the method purpose statement and the header, we include the purpose statement and the header in the interface IListOfPhotos and in all the classes that implement the interface.

Including the method header in the interface serves as a contract that requires that all classes that implement the interface define the method with this header. As the result, the method can be invoked by any instance of a class that implement the interface - without the need for us to distinguish what is the defined type of the object.

We can now proceed with the DESIGN RECIPE.

Design the methods that will help you in dealing with your photo collection:

  1. Before burning a CD of your photos, you want to know what is the total size in bytes of all photos in the list of photos. The method totalSize should help you. Design it.

  2. You now want to go over the list of photos and select only the photos in the jpeg format. Design the method onlyJpeg to help you with this task.

  3. Finally, you want to sort the list of photos by the name of the image (as typically these are generated by your camera and represent the date and time when the photo was taken). Of course, you design the method sortByDateTime.

    Note: If you are running out of time, sort only by date.

More Methods for Self-Referential Data Graphics

You may finish this part at home as a part of your Portfolio.

In the previous lab you defined classes that represent different geometric shapes - a circle, a square, and a shape that is a combination of two shapes, the top and the bottom one. Recall the data definition as given by the class diagram:

                        | Shape |<--------------------------+
                        +-------+                           |
                        +-------+                           |
                            |                               |
                           / \                              |
                           ---                              |
                            |                               |
         ---------------------------------------            |
         |                  |                  |            |
  +-------------+    +-------------+    +--------------+    |
  | Square      |    | Circle      |    | Combo        |    |
  +-------------+    +-------------+    +--------------+    |
+-| Posn nw     |  +-| Posn center |    | Shape top    |----+
| | int size    |  | | int radius  |    | Shape bottom |----+
| | Color color |  | | Color color |    +--------------+ 
| +-------------+  | +-------------+                     
+----+ +-----------+
     | |
     v v
  | Posn  |
  | int x |
  | int y |

  1. Design the method totalArea that computes the total area of a shape. For the shape that consists of two components add the areas - as if you were measuring how much paint is needed to paint all the components.

    You will need to use math functions, such as square root. The following example show how you can use the math function, and how to test doubles for equality. (You can only make sure they are different only within some given tolerance.)

    class Foo{
      double x;
      Foo(double x){
        this.x = x;
      double squareRoot(){
        return Math.sqrt(this.x);
    class Examples {
      Examples () {}
      Foo f = new Foo(16.0);
      boolean testSquared = 
        check this.f.squareRoot() expect 4.0 within 0.01;

  2. Design the method moveBy that produces a new shape moved by the given distance in the vertical and horizontal direction.

  3. Design the method isWithin that determines whether the given point is within this shape.

  4. Of course, we would like to draw the shapes on a canvas. Design the method drawShape that draws this shape on the given Canvas. The following code (that can be written within the Examples class) shows how you can draw one circle:

    import draw.*;
    import colors.*;
    import geometry.*;
    class Examples{
      Examples() {}
      Canvas c = new Canvas(200, 200);
      boolean makeDrawing = 
        this.c.show() && 
        this.c.drawDisk(new Posn(100, 150), 50, new Red());

    The three import statements on the top indicate that we are using the code programmed by someone else and available in the libraries named draw, colors, and geometry. Open the Help Desk and look under the Teachpacks for the teachpacks for How to Design Classes to find out more about the drawing and the Canvas.

Save all your work — the next lab will build on the work you have done here!

Last modified: Tuesday, January 22nd, 2008 12:24:51pm