29823 05 pp115-178 r4jm.

ps 8/6/03 3:53 PM Page 115

C H A P T E R 5

The Photo Editor

Application
Unified Process:
Elaboration Phase and Third Iteration

This chapter discusses the third iteration of the project, which corresponds
to the elaboration phase of the Unified Process. In this phase the goals to
be achieved are to provide an architectural baseline and to formulate a
project agreement with the customer to further pursue the project.
The architectural baseline implements a working application with lim-
ited functionality. The implemented features in our example include ex-
ception handling, loading and saving images, and basic image operations.
The project agreement that is signed at the end of this phase should
include the time frame, equipment, staff, and cost. This leads to the goals
that are set for this iteration:

■ The vision and business case are updated.

■ The requirements are refined and analyzed.
■ The risk assessment is updated.
■ A detailed project plan is developed.
■ The executable program, implementing the architectural framework
with UML documentation, is developed.
■ The project agreement is signed by the stakeholders to continue the
project.

Most of the milestones are based on the work products that were
started in the inception phase. Based on those work products, the work
breakdown into the five core workflows is as follows:
■ Requirements: Refine the requirements and system scope.
■ Analysis: Analyze the requirements by describing what the system
does.

115
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 116

116 Chapter 5 The Photo Editor Application

■ Design: Develop a stable architecture using UML.

■ Implementation: Implement the architectural baseline.
■ Test: Test the implemented architectural baseline.

In the elaboration phase the main focus lies on the requirements and
analysis workflows. But as explained in Chapter 2, all five core workflows
are to be worked on in all phases. This means that in this iteration we spend
a considerable amount of time in the design, implementation, and test
workflows as well.

5.1 The Refined Project Vision and Business Case

The refined vision is based on the analysis discussed in Chapter 4 and adds
new information.
The goal, or vision, for this project is to provide software components
to a printing and embossing business that allows online ordering of prints
and other items, such as cups, postcards, T-shirts, and mouse pads, that can
be customized with digital images.
To accomplish this task, the user must download a photo editor applica-
tion that is then installed on the customer’s computer. After installation the
user can alter images, adding graphics, text, and special effects. Users can
then save the altered images on their local computers and later upload the
images via the Online Photo Shop’s Web site to order prints or customized
products.
The new software will open the business to a new and larger customer
base, allowing for future growth.

5.2 Refined Requirements for Online Photo Shop

5.2.1 Refined Use Case Diagram

The first task in refining the requirements is to reevaluate the use case
R diagram. Based on the initial requirements, analysis, and design decisions
described in Chapter 4, we update the use case diagram as shown in Fig-
ure 5.1.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 117

5.2 Refined Requirements for Online Photo Shop 117

Online Photo Shop

Web Order Application
Browse
Catalog

Customize with
Picture
Add to Shopping
List
Order Products

Checkout
Collect Shipping
Information

Upload Pictures
Collect Payment
Select Information
Options
Print options
Summarize
Customer Order Printing
Business

Photo Editor Application

Crop

Prepare Picture Brightness

for Printing Correction

Contrast
Adjustment

Load and Save Text and Graphics

Pictures Annotations

Apply Special
Effects

Color Adjustment

Figure 5.1 Refined Use Case Diagram for Online Photo Shop

The refined use case diagram clearly distinguishes the different parts of
Online Photo Shop, as discussed in Chapter 4. Online Photo Shop provides
a Web-based interface for ordering products, whereas the photo editor
application is used to edit pictures on the local computer. Because of that,
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 118

118 Chapter 5 The Photo Editor Application

the photo editor and the Web order application are in different processes.
This also means that they are two independent applications and have no
anticipated dependencies. In addition, the new diagram shows additional
use cases for loading and saving pictures in the photo editor, as well as a use
case for uploading images to the Web application.

5.2.2 Refined Requirements List

The next task is to update the requirements. In addition to adding the new
requirements, it is important to put more details into the requirements.
These extra details include pre- and postconditions, functional descrip-
tions, and constraints. Refinement of requirements will not be finished in
this iteration, but the goal is to put all details of the use cases known at this
point into the requirements to support the following activities:

■ Negotiating the deliverable functionality with the customer, includ-

ing the constraints
■ Analyzing the requirements and developing a good design for the
implementation
■ Planning the project and providing reasonably accurate effort esti-
mates (the effort estimates will be refined throughout development)
■ Enabling project management to develop an accurate contract
agreement with the customer

Table 5.1 shows the refined requirements for the current iteration.
For the remainder of the book, we show only the requirements that are
to be implemented in the described iteration. The alternative would be to
refine all the requirements in the first construction iteration, but we have
not chosen that approach for several reasons. First, it is easier to focus on
the functionality that needs to be implemented in the specified iteration.
Second, if requirements must be changed because of design or implemen-
tation issues, the changes usually will affect only the current iteration’s
requirements, and the requirements in later iterations can take all these
changes into account from the beginning. (If complete planning is done up
front, it is like the Waterfall model. Changes in later phases could trigger
many changes.) For reference, the complete list of refined requirements
can be found in the Requirement Specification XML file of each chapter
on the accompanying CD.
The refined requirements include more details on the actual use case
functionality and consider known constraints. In every iteration, we refine
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 119

5.2 Refined Requirements for Online Photo Shop 119

Table 5.1 Refined Base Requirement Description

Requirement Type Summary

F:photo_ Functional Customers shall be able to perform basic photo post-

editor processing on their digital images. The GUI shall be
similar to the GUI model provided in the file Photo
EditorGUI.vsd. The GUI model shows the screen
layout and the approximate position of the necessary
buttons. The buttons to be used are not defined. The
buttons should be chosen so that they can be used
intuitively.
F:error_ Functional Errors shall be reported via exceptions. The exception-
handling handling mechanism is based on the Microsoft Excep-
tion Management Application Block (described in
detail later in this chapter) and needs to be extended
to show error messages in a window on the screen.
Every exception shall be displayed on the screen
within a dialog window. A meaningful error message
shall be displayed. For fatal errors, from which the
application is not able to recover, the exception is writ-
ten into the event log and the application is shut down.
F:picture_ Functional It shall be possible to load pictures into the photo
load_and_save editor application, alter them, and save them either
to the same or to a new file. An error message is dis-
played if the image format is not known (see also
image_format).
F:image_crop Functional If an image is loaded, then the photo editor shall pro-
vide functionality to select a region of interest and
crop the image to that region.
F:image_ Functional If an image is loaded, the photo editor shall be able to
rotate rotate an image by 90 degrees in both directions.
F:image_flip Functional If an image is loaded, the photo editor shall be able to
flip an image vertically and horizontally.
F:image_ Functional The supported image formats are JPG and GIF. The
format supported image sizes range from 100 pixels to 2,048
pixels for both width and height. An error is reported
if the image format is not known.
C:platform_os Constraint The target platform operating system is Windows XP
Professional. No special hardware requirement is
necessary.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 120

120 Chapter 5 The Photo Editor Application

the requirements that are to be implemented in that iteration. This allows

us to choose the best possible design and therefore the best possible imple-
mentation for the system.

5.3 Analysis of the Photo Editor Requirements

Next, we analyze the refined requirements. The goal in the analysis work-
A flow is to describe what the system is supposed to do without defining how
it does it (the time to define how the system is expected to be implemented
is in the design workflow). For example, the photo_editor requirement
key defines what the graphical user interface should look like without
describing how it is implemented.
In this book, the analysis and design workflows are done just in time,
meaning they are done within the iteration that implements the correspond-
ing requirements. This approach is chosen because the team members are
highly experienced programmers in the domain and technology we are
using. Because of that, the effort estimates are based solely on the require-
ments and depend heavily on the experience of the team members. In
addition, this approach provides a clearer structure for this book in describ-
ing the analysis, design, and implementation of the features within the
chapter in which they are to be implemented. However, if a project has
less-experienced team members or an unknown domain, it might be neces-
sary to analyze all the requirements in more detail up front and to develop
the complete design in the elaboration phase in order to provide a reason-
able estimate. The estimates are the basis of the project agreement.
In the following subsections, we analyze, design, implement, and test
the requirement keys implemented in this iteration and the requirement
keys that define dependencies between the system’s modules.

5.3.1 The photo_editor Requirement

The photo editor application will run in its own namespace. For this itera-
tion, the photo editor application will provide basic functionality. The
appearance of the GUI is defined in Figure 5.2 as a guideline.
The figure shows the screen layout and the controls used. It does not
define the exact appearance or shapes of the resources that are used to
implement the GUI. We’re using a toolbar menu containing the items
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 121

5.3 Analysis of the Photo Editor Requirements 121

Photo Editor Application

File
Open
Save
Save as…
Exit

Basic image operations

Rotate 90 clockwise Flip horizontal

Image crop
Rotate 90 counter clockwise Flip vertical

Figure 5.2 The Photo Editor GUI

Open, Save, Save As, and Exit. For the image-processing functionality, a
TabControl with buttons is provided. The idea is to group related function-
ality on one tab and to provide other tabs for other groups of functionality.

5.3.2 The error_handling Requirement

Errors are reported via the exception-handling mechanism. Various mes-
sages will be available to identify the error that led to the message. The
errors will be reported via message boxes and will contain meaningful error
messages.

5.3.3 The picture_load_and_save Requirement

Users will be able to load and save rectangular pictures. The load and save
option is part of the File menu, as shown in Figure 5.2. Users can browse
for files or specify a file manually, and we use the standard Windows Forms
Open/Save File dialog.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 122

122 Chapter 5 The Photo Editor Application

5.3.4 The image_crop Requirement

Image crop allows users to extract a rectangular part of the image. When
the Crop button is pressed, a dialog window opens and the user is asked to
type in the width and height of the cropped image. The image is cropped to
the size specified by the user if the size specified is smaller than the actual
image; otherwise, it is ignored. The cropping is done in a way that the mid-
point of the original image is the midpoint of the cropping rectangle
defined by the user. The area outside the cropping rectangle is deleted, and
the new image in the new size is shown. If the defined cropping rectangle is
larger than the image, no cropping is done.

5.3.5 The image_rotate Requirement

The image can be rotated in 90 degree steps clockwise and counterclock-
wise. The image will be shown in the same location and with the same mid-
point as the original image.

5.3.6 The image_flip Requirement

By pressing the image flip buttons, users can flip the image either horizon-
tally or vertically. By “flip,” we mean mirroring the image. The position of
the midpoint of the image stays the same. The image will be shown in the
same location and with the same midpoint as the original image.

5.3.7 The image_format Requirement

The standard formats for pictures are supported. If an image with another
format is selected, an error message will be shown.

5.3.8 The platform_os Requirement

All development and testing are done on Windows XP computers. Even
though the application might be able to run on other versions of Windows
operating systems, this will not be supported.

5.4 Design of the Photo Editor Application

We have analyzed the requirements and have defined what the system is
D supposed to do. The next step is to decide how the system will be imple-
mented—in other words, to design the system.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 123

5.4 Design of the Photo Editor Application 123

A standard way to document software architectures is to use the Uni-

fied Modeling Language (UML). We mainly use UML class diagrams to
define the structure of the photo editor software system. The design shows
classes, attributes, operations, and the relationships among them. If you
have a working knowledge of UML, you may want to skip the following
brief introduction to UML class diagrams and go immediately to the pro-
vided class diagram of the photo editor in section 5.4.2.
If you’re not familiar with UML, we offer a very short introduction to
the class diagram in UML. This introduction is not intended as a tutorial
on UML class modeling but rather aims to build a common vocabulary to
help you understand the design of the system throughout the book. There-
fore, the goal of this section is to give you just enough details to under-
stand the class diagrams developed for the application. For more detailed
information on UML and class diagrams, see the references section in
Chapter 4.

5.4.1 Brief Introduction to UML Class Diagrams

The basic element of a class diagram is a class, which you model using a
rectangular box. This box can contain the class name by itself and can also
contain the class attributes and the operations defined by the class. Figure
5.3 shows sample class definitions.
Figure 5.3A shows a class that has sample attributes and an operation
definition. The sampleOperation call returns a uint value, which is shown
after the operation name and separated from it by a colon. In addition, the
visibility of attributes and operations is shown in the class model by means
of the provided prefix. Table 5.2 lists the supported visibility prefixes and
their meanings. As you can see, in SampleClass the attributes are private
and protected, whereas sampleOperation is defined as public.

A B C D
T
SampleClass SampleAbstractClass «interface»
InterfaceClass SampleParameterClass
-sampleAttribute -size : long = 0
#anotherAttribute +ISampleInterface() -example : T
+GetSize() : long
+sampleOperation() : uint +Operation() : short

E
SimpleClass

Figure 5.3 Sample Class Models

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 124

124 Chapter 5 The Photo Editor Application

Table 5.2 Class Diagram Visibility Prefixes

Prefix Visibility

- (hyphen) private: Can be seen only within the class that defines it

# protected: Visible within the class that defines it and by subclasses of

the defining class
+ public: Visible to any class within the program

The sample class shown in Figure 5.3B is an abstract class. Abstract

class names are shown in italics to distinguish them in the class dia-
gram. Note, too, that the public operation GetSize() returns a long value,
which actually is the private attribute size of SampleAbstractClass. The
private attribute size is initialized to 0 (zero) and of type long.
An abstract class without any implementation in UML is called an
interface class. Figure 5.3C shows a sample interface class identified by the
stereotype <<interface>> above the class name. Interface classes provide
no implementation, and this means that no section for attributes is needed
in the class model. The operations in an interface class are shown in italics
because they do not provide any implementation and must be imple-
mented by the derived classes.
UML also lets you define parameterized classes. In C++ this feature is
known as a template class (C# will provide support for parameterized or
generalized classes in a coming version). Parameterized classes are identi-
fied by the parameter definition in the upper-right corner of the class
model. Figure 5.3D shows a sample parameterized class.
The class shown in Figure 5.3D is a simple class showing only the
class name. This type of model is usually used in overview diagrams of large
systems.
To model the system, we must define relationships between the classes.
The next section introduces the class relationship models in UML.

Class Relationships
Figure 5.4 shows the basic dependency principles among classes.
Figure 5.4A shows a dependency between Class1 and Class2. The
dependency indicates that changes to Class2 could trigger changes in
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 125

5.4 Design of the Photo Editor Application 125

A Dependency
«call»
Class1 Class2

B Navigability
-role C5 -role C6
Class5 Class6
1 0..1

Navigable in both directions

C Object Association
role C3role C4
Object1 : Class1 Object2 : Class2

Use notes to provide additional information

D Multiplicities

Class7 Class8
1 *

Class9 Class10
0..1 1..*

Class11 Class12
0..* 2..9

Figure 5.4 Class Dependency

Class1. Figure 5.4B shows navigability in both directions. Navigability

shows the directions (unidirectional or bidirectional) in which the program
can traverse the class tree. In the example, Class6 “knows” about Class5,
and vice versa. The navigability diagram also shows how to use notes to add
information to the class diagram if necessary. At both ends of the navigation
relationship, you can show any multiplicity of the relation (see Figure
5.4D) for more detail).
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 126

126 Chapter 5 The Photo Editor Application

Figure 5.4C shows an association relationship between two class

instances (or objects) in addition to a note. The class instances are identi-
fied by the prefix object and the class name, which are separated by a colon
and underlined. Both ends of an association can be labeled. The label is
called a role name.
Another important part of class diagrams is the ability they give you to
express multiplicities, as shown in Figure 5.4D. Multiplicity is expressed by
a number on each side of the dependency connection. If no number is sup-
plied, then the default, which is 1, is assumed. The meanings of the multi-
plicities numbers are as follows:

1 Exactly one
* Many (zero or more)
0..1 Optional (zero or one)
1..* One or more
0..* Optional, one or many
2..9 Example of an m..n relationship; in this case, a two to nine
dependency

Now let’s look at the more advanced class dependencies, as shown in

Figure 5.5.
Aggregation (shown in Figure 5.5A) is a part-of relationship. For
example, a point is part of a circle. This sounds very simple, but then what
is the difference between dependency and aggregation? Aggregation can
be seen as a placeholder. In other words, you can use aggregation as an
indication of a relationship without specifying the details of the relation-
ship. This is possible because UML defines very little of the semantics of
aggregation.
In addition to aggregation, UML offers a more defined variation called
composition. Composition implies that the part-of object belongs only to
one whole, and the part-of object’s lifetime responsibility is with the object
that contains it. Figure 5.5B shows that Class2 is part of Class1 and that
Class1 is responsible for creating and deleting objects of Class2.
Another dependency defined in UML is the generalization, shown in
Figure 5.5C. Generalization can be seen as an is-a relationship. For
example, a diesel engine is an engine, and a gasoline engine is an engine.
This holds true if “engine” is defined as the commonalities of diesel and
gasoline engines. Every diesel engine and every gasoline engine can then
be defined by deriving from the general engine class and adding the
specifics according to the engine type.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 127

5.4 Design of the Photo Editor Application 127

A Aggregation D Interface and Implementation

Class3 Class4 «interface» Class8
1 * Interface1

B Composition realization

Class1 Class2 «implementation class»

1 * ImplementationClass

C Generalization Class9
Class5 Interface
«implementation class»
AnotherImplementationClass

Class6

E Parameter Class F Parameter Class

T AssociationClass1
SampleParameterClass
-example : T
+Operation() : short * *
Person Company
«bind»(short)

BoundElement

Figure 5.5 Advanced Class Dependencies

Interfaces and their implementation can be modeled within UML in

two ways, as shown in Figure 5.5D. The first technique is to define an inter-
face class that does not contain any implementation. This class is then con-
nected to the implementation class via the realization dependency. In the
example, Class8 calls on the Interface method, which is implemented in
the ImplementationClass. The second way to show an interface in a UML
diagram is by using an Implementation class that shows the provided inter-
face with the lollipop representation. In the example, Class9 calls the
Interface method of AnotherImplementationClass. In the first tech-
nique, you can explicitly show the interface definition and the implementa-
tion, whereas the second technique shows only the implementation class.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 128

128 Chapter 5 The Photo Editor Application

The use of parameterized classes is called bound elements and is known

to C++ programmers as template classes. In Figure 5.5E, BoundElement is
bound by a short parameter to SampleParameterClass. Parameterized
classes cannot be extended by the bound class. BoundElement can use com-
pletely specified types from the parameter class. The only thing added to
the parameter class is the restricting type information.
Association classes are used to add an extra constraint to a relation of
two objects. This is shown in Figure 5.5F. Only one instance of the associa-
tion class can be used between any two participating objects. In the
example, a Person is working for at most one Company at a time (at least in
theory). An association class is used to keep information about the date
range over which each employee is working for each company. Because
persons sometimes switch jobs, this information should not be kept within
the person’s class directly but rather in an association class.
In object-oriented programming, you usually try to limit dependencies
to the absolute minimum. In this way, changes made to one class do not
trigger endless changes to other classes. Necessary dependencies are usu-
ally kept on the interface level.

5.4.2 Design of the Photo Editor Application

The design of the photo editor application is done in Microsoft Visio. The
decision to use Visio was made because in addition to providing tools to
generate UML diagrams, Visio lets you generate code from class models.
Also, Visio supports extracting class models from existing code, a capability
that can be very useful when you’re reengineering poorly documented
software. Unfortunately, Visio has no support for the creation of XML doc-
umentation from the UML models. Therefore, during document review
we must trace requirement keys manually through the design process.
The class diagram shows that the main application class, called Photo
EditorForm, is derived from the Windows.Forms class. The Windows.Forms
class provides Windows-related functionalities that are used by the photo
editor. In addition, PhotoEditorForm creates an instance of Exception
Publisher, which is derived from the IExceptionPublisher interface pro-
vided by the Microsoft application block. In addition, an instance of the
Picture class is created. The Picture class instance holds the loaded
image and provides the basic image-manipulating operations (such as rotat-
ing and flipping).
Figure 5.6 shows the class diagram for the photo editor application in
this iteration.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 129

5.5 The Detailed Project Schedule 129

System.Windows.Forms.Form

+OnPaint() «interface»
IExceptionPublisher
+Publish()

PhotoEditorForm
-Picture : Picture -exception «implementation class»
-Exception : ExceptionPublisher 1 ExceptionPublisher
+OnPaint() : System.Windows.Forms.Form 1 1
+Publish()

-image Picture
-size
0..1
+FromFile() : Picture
+Dispose()
+RotateFlip()
+Save()
+SaveAdd()
+Size()

Figure 5.6 The Photo Editor Class Diagram

5.5 The Detailed Project Schedule

As mentioned earlier, it is important to track the project and update the

plan constantly. In this iteration we develop a more refined project sched-
ule, as shown in Figure 5.7. The project schedule is kept in Microsoft Visio
as a Gantt chart, showing the various tasks, their dependencies, and the
project milestones. Online Photo Shop is the overall project, and we indent
all other tasks to show that they are defined as subtasks.
For easier readability, the requirements are grouped into blocks of
related functionality. The name of the developer who is responsible for a
block is shown in the resource column. Most of the blocks contain depen-
dent or very small functionalities (or both). By combining these items into
blocks, we make it easier to do tracking on a daily instead of an hourly basis.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 130

130 Chapter 5 The Photo Editor Application

Detailed Project Schedule: Online Photo Shop

Nov 3 2002 Nov 10 2002

Resource
ID Task Name Start Finish Duration
Names 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

1 Milestone 1 (progress review) 11/11/2002 11/11/2002 0d

2 Milestone 2 (progress review) 11/25/2002 11/25/2002 0d

3 Milestone 3 (progress review) 12/19/2002 12/19/2002 0d

4 Online Photo Shop 10/10/2002 1/31/2003 82d Online Photo Shop

5 photo_editor 10/10/2002 1/7/2003 64d GL, GL,

KC KC photo_editor

6 order_online 10/10/2002 1/7/2003 64d TM TM order_online

error_handling, picture_load_and_save, image_format,

7 10/22/2002 11/8/2002 14d GL
error_handling, GL
picture_load_and_save, image_format, image_crop, image_rotate
image_crop, image_rotate
8 Image_flip 11/4/2002 11/8/2002 5d Reader Reader Image_flip

9 image_brightness, image_contrast, image_color, 11/11/2002 11/20/2002 8d GL GL image_brightness, image_contrast, image_color,

10 image_graphics_annotations, image_text_annotations 11/11/2002 11/20/2002 8d KC, Reader KC, Readerimage_graphics_annotations, image_text_annotations

order_products, product_browse, product_options,
11 11/11/2002 11/20/2002 8d TM, Reader TM,order_products,
Reader product_browse, product_options, product_custo
product_customize
12 Risk Buffer 11/20/2002 11/22/2002 3d

13 image_processing_direct_3d, editor _optimizations 11/25/2002 12/13/2002 14d GL GL

14 image_graphics_special_effects, image_special_effects 11/25/2002 12/13/2002 14d KC, Reader KC, Reader

product_shopping_cart, order_checkout, checkaout_shipping,
15 11/25/2002 12/13/2002 14d TM, Reader TM, Reader
checkout_payment, checkout_summerize.
16 Risk Buffer 12/13/2002 12/18/2002 4d
GL, TM, KC,
17 editor_system_test 12/19/2002 1/20/2003 22d GL, TM, KC, Reader
Reader
18 Risk Buffer 1/20/2003 1/31/2003 10d

Figure 5.7 Detailed Schedule

In addition to the tasks themselves, including their estimated duration,

other milestones are defined and connected to the related tasks (as indi-
cated by the lines that go from the milestones to the tasks). The milestones
are review points at which the status of the project is reviewed and neces-
sary adjustments are made. In the case shown in Figure 5.7, the milestones
coincide with the end (or the beginning) of the iterations. This works well if
the requirements are broken into small enough pieces (no longer than two
weeks’ worth of effort). If you cannot do this, you must either break the
tasks into smaller pieces or define interim milestones.
From the detailed schedule, you can see that even though the itera-
tions are described in a serialized fashion, iterations are actually worked on
in parallel. This practice shortens the overall duration of the project. In our
case, working on development tasks in parallel works very well because the
photo editor application and the online shop have no anticipated depen-
dency. The schedule shown is constantly refined and updated throughout
the development cycle. Also, we have added risk buffers to the schedule.
These buffers give the project additional time to react to unforeseen prob-
lems. Nevertheless, the risk buffers should not be used for adding new
requirements or implementing new features.
Tracking is done in small meetings. We hold one weekly meeting to dis-
cuss problems, ideas, and management-related items; in addition, stand-up
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 131

5.6 Implementation of the Photo Editor Application 131

meetings (which last about 12 minutes) are held every other day for fre-
quent status updates (no discussions are allowed at these meetings). Stand-
up meetings often lead to the calling of a regular meeting at another time
so that team members can quickly discuss problems found.
From the Gantt chart we developed in Visio, we can export a variety of
diagrams, such as a timeline diagram. These exported diagrams are very
useful for reports to the customer or upper management (assuming that
the plan is kept up-to-date).

5.6 Implementation of the Photo Editor Application

After all the ground work is set, the implementation can start. To begin the
I implementation, a project workspace must be set up and added to configu-
ration management. As mentioned in Chapter 4, the configuration man-
agement administrator is the person who should set up the directory
structure. Based on the directory structure (see Figure 4.3), a solution is
created that will contain all the subprojects of the software developed dur-
ing the course of this book.
To create an empty Visual Studio solution, open Visual Studio.NET
(see Figure 5.8). Then select Visual Studio Solutions and Blank Solution,
and name the solution Photo Editor. The solution should be crated in the

Figure 5.8 Creating an Empty Solution

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 132

132 Chapter 5 The Photo Editor Application

src directory, so browse to src in your project directory tree and click on
the OK button. The Visual Studio solution will be created.

5.6.1 The Basic Photo Editor Application

Now that we have an empty Visual Studio solution, we can add other
projects to it.

Creating a Basic Application

We will develop the GUI and most of the functionality of the photo editor
application in C# using Microsoft Visual Studio 2003. The first implemen-
tation step is to create a Windows application project; we will then extend it
to provide exception handling and image loading, saving, and displaying.
The application is created as a project within the Photo Editor solution.
Therefore, with the Photo Editor solution opened in Visual Studio,
click on the File menu and then Add Project. Then choose New Project. In
the dialog window that appears, choose Visual C# Projects and Windows
Application. In the Name field, type Photo Editor Application. Click on
OK, and Visual Studio.NET creates a new project within the Photo Editor
solution.
The Microsoft wizard in the solution generates three files: App.ico,
AssemblyInfo.cs, and Form1.cs. You can explore all three files by double-
clicking on the file name in Solution Explorer. Obviously, App.ico is the
icon associated with the photo editor application and can be customized.
The AssemblyInfo.cs file contains assembly-related information such as
version and binding. The version information is updated before all releases
(including intermediate releases); binding will be discussed when the prod-
uct is deployed to the customer in the final release. The more interesting
file that was generated is Form1.cs. This file contains the form design and
the code for the photo editor application. To see the generated code, go to
the Form1.cs [Design] view, choose the form, right-click on the form, and
choose View Code.
The top of the file shows various namespaces that are used in the appli-
cation (similar to the kind of information you see in #include statements in
C++). These using statements define shortcuts for the compiler to use in
resolving externally defined namespaces. Extern in this case relates to
objects that are not defined in the project’s namespace. If a call to a method
cannot be resolved, the compiler will try to resolve the call by checking the
namespaces defined by the using statement.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 133

5.6 Implementation of the Photo Editor Application 133

Listing 5.1 shows the namespaces used by the code generated by Visual
Studio.NET. The System namespace contains the basic .NET Framework
types, classes, and second-level namespaces. In contrast, the second- and
third-level namespaces contain types, classes, and methods to support vari-
ous kinds of development, such as GUI, the runtime infrastructure, .NET
security, component model, and Web services development, to name only a
few categories.
In the photo editor application, you can see that several second-level
namespaces are automatically included. The System.Drawing namespace,
for example, provides rich two-dimensional graphics functionality and
access to Microsoft’s GDI+ functionalities. (GDI+ is explained in more
detail in Chapter 6.) For the remainder of this chapter, we use GDI+ to
provide a memory location where the image can be stored and then dis-
played; in addition, we use some GDI+ methods for image operations.
The System.Collections namespace holds collections of objects, such
as lists, queues, arrays, hash tables, and dictionaries. In addition, System.
ComponentModel implements components, including licensing and design-
time adaptations. For a rich set of Windows-based user interface features,
we also include the System.Windows.Forms namespace. Last but not least,
the System.Data namespace lets us access and manage data and data
sources. For more information on the namespaces provided by the .NET
Framework, please refer to the MSDN help.

Listing 5.1 Using Externally Defined Namespaces

using System;
using System.Drawing;
using System.Collections;
using System.ComponentModel;
using System.Windows.Forms;
using System.Data;

The next section in the source file defines the namespace for the appli-
cation, the basic application classes, and the methods. Listing 5.2 shows the
source code that is created.

Listing 5.2 The Photo Editor Namespace

namespace Photo_Editor_Application
{
/// <summary>
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 134

134 Chapter 5 The Photo Editor Application

/// Summary description for Form1.

/// </summary>
public class Form1 : System.Windows.Forms.Form
{
/// <summary>
/// Required designer variable.
/// </summary>
private System.ComponentModel.Container components = null;

public Form1()
{
//
// Required for Windows Form Designer support
//
InitializeComponent();

//
// TODO: Add any constructor code after
// InitializeComponent call
//
}

/// <summary>
/// Clean up any resources being used.
/// </summary>
protected override void Dispose( bool disposing )
{
if( disposing )
{
if (components != null)
{
components.Dispose();
}
}
base.Dispose( disposing );
}

#region Windows Form Designer generated code

/// <summary>
/// Required method for Designer support - do not modify
/// the contents of this method with the code editor.
/// </summary>
private void InitializeComponent()
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 135

5.6 Implementation of the Photo Editor Application 135

{
this.components = new System.ComponentModel.Container();
this.Size = new System.Drawing.Size(300,300);
this.Text = "Form1";
}
#endregion

/// <summary>
/// The main entry point for the application.
/// </summary>
[STAThread]
static void Main()
{
Application.Run(new Form1());
}
}
}

First, we define the namespace Photo_Editor_Application. This is

the namespace that refers to all the classes, types, and methods defined by
the photo editor application. Next, class Form1 is defined as public and is
derived from System.Windows.Forms.Form. This class implements the
application window, which is called Form1, at least for now. The first prop-
erty defined in the Form1 class is declared private, named components,
and defined to be of type System.ComponentModel.Container. The value
is set to null to indicate that no initialization has yet been done.
Next, we define the public constructor of Form1. The implementation
first calls the InitializeComponent method created by Visual Studio De-
signer. This method takes care of the necessary Windows Forms property
initialization. You should not modify the Designer-generated part of the
code directly, but you can do so through the properties window of Form1.
cs[Design].
After the constructor, the dispose method is defined. The Dispose()
method is used to free system resources if they are no longer needed. Even
though Visual Studio.NET provides garbage collection for allocated mem-
ory, we must explicitly delete other system resources at the time they are no
longer used. Resources other than memory should be disposed of in order
to keep the resource footprint (memory, disk space, handles, and so on) as
small as possible. In the example, the components are disposed of if they
were allocated (not null); to do this, we call the base class’s dispose method.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 136

136 Chapter 5 The Photo Editor Application

Next comes a block specified with the #region and #endregion key-
words. This block allows the developer to write code that can be expanded
or collapsed within the Visual Studio.NET development environment. In
the example, the #region-#endregion block encloses the initializing
method of the Designer-generated form, as described earlier, and this code
should not be altered. You can collapse the code by pressing the “-” symbol,
or expand it by selecting the “+” symbol next to the keyword. Usually the
IDE provides the expand-collapse feature automatically for multiline com-
ments, class definitions, and method definitions, to name only a few. The
developer can define additional collapsible and expandable regions. The
region statements can be nested. In that case, the #endregion matches
the last defined #region statement that has not yet been matched.
The final part of the code defines the static main entry point for the
application. The application is defined to be running in a single-threaded
apartment (STA), and the main entry point then creates and runs an
instance of Form1.
The next step is to change the output directory of the compiler to the
bin and bind directories. As mentioned in Chapter 4, the bin directory
holds all assemblies necessary to run the photo editor application in re-
lease configuration, whereas the bind directory holds the same files but
compiled in debug configuration. To change the output directory, choose
the project in the Solution Explorer window and then choose Project |
Properties | Configuration Properties | Build | Output Path; change the
path to the bin directory (for release configuration) and bind (for debug
configuration).
Before you check the source files into the configuration management
system, you need to make some additional changes. For easier readability,
maintainability, and understanding, it is worthwhile to rename some of the
generated source files to more meaningful names. We do this before
checkin because renaming files already under configuration management is
not always easy.
Therefore, we change the name of the application source file from
Form1.cs to PhotoEditor.cs. We do this by right-clicking on the file name
Form1.cs in the Solution Explorer window and then going to the Proper-
ties window below Solution Explorer and changing the name in the File
Name field. After the name is changed, we adjust other properties of
PhotoEditor.cs[Design]. We select the corresponding tab and click on
the form. We change the Text field to Photo Editor Application, and
change the (Name) field to PhotoEditorForm.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 137

5.6 Implementation of the Photo Editor Application 137

To finish the cosmetics, click on the photo editor form and choose View
Code (by right-clicking on the form and choosing the option). Change
Form1()to PhotoEditorForm(), as shown in Listing 5.3.

Listing 5.3 PhotoEditorForm Creation

static void Main()

{
Application.Run(new PhotoEditorForm());
}

Before you put the project into the source control system, make sure
that it compiles. Go to the Build menu in Visual Studio, and choose Build
Project (or use the shortcut by pressing Ctrl+Shift+B). If there are any
errors during compilation, they will be shown in the output window below
the main window. Double-clicking on the error message will open the cor-
rect source file to the approximate position of the error.

5.6.2 Using Version Control

The Visual Source Safe source control system integrates smoothly with the
Visual Studio.NET IDE. There are several ways to communicate with it.
The following assumes that Visual Source Safe is installed on the computer
(even though working with other source control systems may be quite simi-
lar). The first possibility is to use the File | Source Control menu, which is
shown in Figure 5.9.
Another method of communicating with the version control system is
to configure a toolbar menu. Go to View | Toolbars | Source Control. You’ll
get a toolbar (shown in Figure 5.10) that can be added to the toolbar sec-
tion of Visual Studio.
There is yet another way to check the project into the source control
system. Right-click on the project in Solution Explorer and choose Add
Solution to Source Control.
In our sample project, we add the project to the source control system
by using the latter method. The system asks for the login credentials before
the Visual Source Safe dialog appears, as shown in Figure 5.11.
Now we choose the name and location for storing the project in the
source control system. For this project, we take the defaults suggested by
Source Safe (location root and project name Photo Editor Application), so
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 138

138 Chapter 5 The Photo Editor Application

Figure 5.9 Visual Source Safe Menu

Figure 5.10 Visual Source Safe Toolbar

Figure 5.11 Source Safe Dialog

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 139

5.6 Implementation of the Photo Editor Application 139

we just click OK. The system asks whether it should create the new project,
and again we simply confirm by clicking OK. As you can see, all files in
Visual Studio Explorer now have a little lock symbol next to them; this
means that the files are read-only (checked in) under source control.
Before a developer can make any change to the files, he or she must
check out the files to get write permissions to them. Checkout means that
the developer will work on a private copy of the file. The developer works
on the private copy, keeping it checked out, until the change is (partially)
complete and compiled; then the file must be checked in. At checkin time
the private file is added to the version control as the newest version of the
file; the file becomes read-only again, and all developers on the project can
see the the new file. Changes can be made only by the developer who cur-
rently has the file checked out. It is also possible to undo a checkout,
thereby reverting to the original version of the file and discarding the
changes made. (Nobody will ever know about the changes. Usually it is a
good practice to save the changes under a different file name in a private
file before undoing a checkout and losing the changes.)
Visual Source Safe provides many other useful tools, such as forcing the
system to get the latest version of all files and comparing the changes in dif-
ferent file versions. Most of the functionality is self-explanatory. For more
detailed information, you can consult the Visual Source Safe help files.

5.6.3 The Exception-Handling Application Block

The first “real” functionality that we will implement is exception manage-
ment. As discussed earlier, exception handling in the photo editor applica-
tion is based on the Microsoft Exception Management Application Block.
You can download this application block from http://msdn.microsoft.
com/downloads/list/bda.asp. You install the files by following the in-
structions on the screen. Alternatively, you can simply take the installed
source files from the sample solutions provided on the CD under Chapter5\
src\Exception Management Application Block.
Next, go to the installation directory of the application block and
open the Code directory with a double-click in Windows Explorer. There
are two versions of the exception management block installed. The first
version is written in Visual Basic (VB), and the second version is imple-
mented using C# (CS). We are interested in the C# version, so open the CS
subdirectory and double-click on the Exception Management Application
Block (CS) solution. This will open a solution with two projects: Microsoft.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 140

140 Chapter 5 The Photo Editor Application

ApplicationBlocks.ExceptionManagement.Interfaces and Microsoft.

ApplicationBlocks.ExceptionManagement in Visual Studio.NET.
To use the application blocks, we need to build both projects. But
before we actually build them, we first set the output path of both projects
to the bin and bind directories of the photo editor project. To set the out-
put path of the compiler-generated assemblies, right-click on the project in
Solution Explorer. Choose Properties and go to Configuration Properties |
Build | Outputs | Output Path. Make sure that the configuration of the
selected project is set to Release. Choose the table entry for Output Path,
and navigate to the photo editor bin directory. Then choose Apply. If the
project is now being built in release mode, then the assemblies will be
saved in the bin directory of the photo editor project. Do the same for the
Debug configuration of both projects by specifying the bind directory as
the output path.
After that, build the solution by going to the Build menu and choosing
Build Solution (or by using one of the previously mentioned shortcuts).
Then change the build configuration in the Active Solution Configuration
list box and build this configuration as well. (Either use the menu bar or go
to the Build menu, choose Configuration Management, and choose Debug
or Release depending on the configuration you just built.) You can check
the success of the build by checking the bin and bind directories for the
created assemblies. You should be able to see ExceptionManagement.dll
and ExceptionManagement.Interfaces.dll in both directories.
After the assemblies are built, we add references to Microsoft.
ApplicationBlock.ExceptionManagement.dll and Microsoft.Application
Block.ExceptionManagement.Interfaces.dll to the photo editor appli-
cation project. To do this, you choose the menu item Project | Add Refer-
ence. Alternatively, you can right-click on the Photo Editor Application
project in Solution Explorer and choose Add Reference. A dialog box opens.
Choose Browse and navigate to the bin directory, where the assemblies for
the Microsoft Exception Management Application Block reside. (By refer-
encing the assemblies in the bin directory we are referencing the assembly
that was built in release configuration. Alternatively, we could reference
the debug version. But we do not intend to debug the application block, so
we reference the release version.) Select the two assemblies and press OK.
Another Source Safe dialog box opens that lets you choose whether to
check out the project file for editing. Because we will add references to the
project, we need to update the file Photo Editor Application.csproj to
reflect these changes. Click OK to check out the file. You can add a com-
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 141

5.6 Implementation of the Photo Editor Application 141

ment to the history of the file before checkout. The rule for the photo edi-
tor application is to add checkin comments that explain the changes made.
In Solution Explorer, you can now see the newly added references in the
Reference section. In addition to the references, we need to add a using
statement to the photo editor application to indicate the use of the exter-
nally defined functionality. Again, the source control will ask whether the
PhotoEditor.cs file should be checked out; we acknowledge this by
choosing checkout. The added code can be seen in Listing 5.4.

Listing 5.4 Using the Microsoft Exception Management Application Block

using System;
using System.Drawing;
using System.Collections;
using System.ComponentModel;
using System.Windows.Forms;
using System.Data;

using Microsoft.ApplicationBlocks.ExceptionManagement;

Now that the photo editor application is aware of the exception man-
agement class, the classes and methods provided by it can be used. The
provided default publisher logs the exceptions to the event log. This is not
the behavior we intend for the photo editor application, so we must create a
custom publisher. Before continuing, make sure that all changes are saved
and the project compiles.

5.6.4 Customized Exception Handling

You create custom publishers by implementing the IExceptionPublisher
Interface, which is defined in the Microsoft.ApplicationBlocks.
ExceptionManagement.Interfaces.dll assembly. To keep the exception
publisher code separate from the application code, add a new file to the
photo editor application project. To add a file, right-click on Photo Editor
Application in Solution Explorer and choose Add | New Item (see Figure
5.12).
Select Class in the dialog box and type the name ExceptionPublisher.
cs for the file. Click on Open; this will open a new file with the added name
selected. The file contains the using system statement and shows that it is
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 142

142 Chapter 5 The Photo Editor Application

Figure 5.12 Adding the ExceptionPublisher.cs File

part of the Photo_Editor_Application namespace. In addition, a class

definition for class ExceptionPublisher and its constructor are provided.
The new file is automatically added to Visual Source Safe and is marked as
checked out (marked with a red check mark next to the file name in Solu-
tion Explorer).
Next, we add the using statement for the Exception Management
Application Block in the same way as was shown for the PhotoEditor.cs
file. In addition, we add the using System.Windows.Forms statement for
Windows Forms support. Then we must derive the ExceptionPublisher
class from IExceptionPublisher. Therefore, we change the class defini-
tion to public class ExceptionPublisher : IExceptionPublisher.
When we’re finished typing, a stub for the IExceptionPublisher interface
can be added automatically by pressing the Tab key (otherwise, we simply
type it in as shown in Listing 5.5). The finished Publish method is shown in
Listing 5.5. It also shows the three arguments the Publish method takes.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 143

5.6 Implementation of the Photo Editor Application 143

Listing 5.5 Publish Interface Implementation with XML Documentation

#region IExceptionPublisher Members

/// <summary>
/// Custom Publisher, displays error message on screen.
/// </summary>
/// <param name="exception">Exception, containing meaningful
/// error message</param>
/// <param name="additionalInfo">Provides additional info
/// about the exception</param>
/// <param name="configSettings">Describes the config
/// settings defined in the app.config file</param>
public void Publish(Exception exception,
System.Collections.Specialized.NameValueCollection
additionalInfo,
System.Collections.Specialized.NameValueCollection
configSettings)
{
// TODO: Add ExceptionPublisher.Publish implementation
string caption = "Photo Editor";
DialogResult result;

// Displays the MessageBox.

result = MessageBox.Show( exception.Message, caption,

MessageBoxButtons.OK,
MessageBoxIcon.Error, MessageBoxDefaultButton.Button1,
MessageBoxOptions.RightAlign);
}
#endregion

To generate nicely formatted documentation from the source code, we

add the XML description as specified in Chapter 4. You generate the docu-
mentation from the code by selecting the Tools menu in Visual Studio and
choosing Build Comment Web Pages. Then select the radio button Build
for the entire solution, and specify Save Web Pages in the doc directory
of your project (in the sample solution this is Chapter5\doc) and click OK.
The documentation is generated, and an Explorer window opens that
shows the generated solution’s comment Web pages. Click on the name-
space to navigate down the program documentation tree and open the next
lower level.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 144

144 Chapter 5 The Photo Editor Application

According to the requirements, exceptions are to be published via a

message box displaying meaningful error messages. This leads us to the next
implementation step, which is to implement the message box. The easiest
way to display a message box is to use a Windows message box. The System.
Windows.Forms namespace, which we have already added to the using
statement section, provides the functionality to display a simple message
box. Several overloaded types of message boxes are supported by the .NET
Framework. The one that we use here is as follows:

MessageBox.Show( text, caption, buttons, icon, defaultButton, options);

The message box used in this example takes six parameters, which are
explained in Table 5.3. For other variants of the message box, please refer
to the MSDN help.
To use the customized exception publisher, the final step is to provide a
configuration file. The configuration file makes the exception application
block aware of the custom publisher that is to be used (this is similar to reg-
istration of the custom publisher). To add an application configuration file,
right-click on Photo Editor Application in Solution Explorer and choose

Table 5.3 MessageBox.Show Parameters

Parameter Type Name Description

string text Text to be displayed in the dialog box,

which is the error message in this case.
string caption Text displayed in message box title.
MessageBoxButtons buttons The buttons to be displayed in the dia-
log window. For the error message dia-
log window, this is just the OK button.
MessageBoxIcon icon This defines the icon displayed in the
message box. For an error message the
displayed icon is a red circle with a
white x inside.
MessageBoxDefaultButton defaultButton The default button. In this case there is
only one button displayed, so button1
is the only and default button.
MessageBoxOptions options The text in the message box is right-
aligned.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 145

5.6 Implementation of the Photo Editor Application 145

Add | New item. In the dialog window that opens, choose Application Con-
figuration File and press Open. A configuration file is added to the solution.
Change the configuration file to correspond with Listing 5.6.

Listing 5.6 App.config: The Application Configuration File

<?xml version="1.0" encoding="utf-8" ?>

<configuration>
<configSections>
<section name="exceptionManagement"
type="Microsoft.ApplicationBlocks.ExceptionManagement.
ExceptionManagerSectionHandler,
Microsoft.ApplicationBlocks.ExceptionManagement" />
</configSections>

<exceptionManagement mode="on">
<publisher assembly="Photo Editor
Application"type=
"Photo_Editor_Application.ExceptionPublisher"
fileName="c:\PhotoEditorExceptionLog.xml"/>
</exceptionManagement>
</configuration>

As you can see, the configuration file provides information regarding the
customized publisher. The configuration entry <publisher assembly=...
defines the name of the assembly in which the customized exception pub-
lisher is defined. The type="..." defines the namespace and the method
name for the publisher, whereas the file name specifies the log file name to
which exceptions are logged in case the defined publisher cannot be found.
If an exception occurs, the Exception Manager Application Block will now
know about the customized exception publisher and will call the specified
exception publisher.
Make sure that the project compiles, and check in all the changes by
choosing the Pending Checkins tab below the main window. When you
check in a file, usually it is good practice to provide a meaningful comment.
The comment for the checkin at this point might read, “Added custom
exception handling using the Microsoft Exception Manager Application
Block. Exceptions are published in a window on the screen.” After typing
the comment, choose Check In. A dialog opens if the files really should be
checked in. Click on OK, and all the changes are available in the repository,
visible to everybody on the team.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 146

146 Chapter 5 The Photo Editor Application

Now the exceptions can be used in the photo editor application. All
code that could possibly throw an exception should be put in a try-catch-
finally block:

try
{
// Some code here
*
*
*
//in case a problem is found, an exception can be thrown
throw(new Exception("Some information here"));
}
catch(Exception exception)
{
ExceptionManager.Publish(exception);
}
finally
{
// Code here will always be executed, whether
//there was an exception thrown or not.
}

The example also shows how an exception is thrown within a method

and how it provides additional information as a string. C# defines a finally
block in addition to the catch block. The finally statement can be put after
the catch block. The code in the block is always executed, either after the
catch is finished or if the try block is finished.

An Impor tant Note on Code Conventions From this point on, it is

your task to provide the try-catch statements during the implementa-
tion of the code, even if it is not explicitly mentioned in the descriptions
of the implementation. In addition, for the remainder of this book it is
assumed that you will add the XML comments according to the coding
guidelines while implementing the functionalities.
It’s also up to you to save and check in files whenever you achieve
an intermediate result. It is good practice to check in only code that is
compiling and that has the added XML documentation.
The sample solutions that are provided on the accompanying CD
contain comments and exception code.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 147

5.6 Implementation of the Photo Editor Application 147

5.6.5 Loading an Image

After the implementation of the custom exception publisher is completed,
we start to implement the basic functionality of the photo editor applica-
tion: loading an image. The .NET Framework provides standard file
dialogs for loading and for saving files. The standard implementation of the
Windows File dialog provides a window that enables the user to browse,
select, name, and filter files. Because this functionality corresponds to what
was specified in the requirements, the photo editor application uses the
.NET-provided standard file dialogs. To start the implementation, we click
on the PhotoEditor.cs[Design] tab and choose the Photo Editor Applica-
tion form.
Next, we set the WindowState in the Layout section of the properties
to Maximized. This sets the main window to maximized when the applica-
tion is started. Compile and run the solution to see the described effect.
After that, we add the File menu to the form. Go to the Toolbox, choose
MainMenu, and drop the main menu on the form. (The Toolbox is usually
shown on the left border of the screen. If the Toolbox is not visible, go to
the View menu and select Toolbox, or simply press Ctrl+Alt+X.)
This action adds an empty menu bar to the form. Rename the menu to
MainMenu in the properties window. To do that, first make sure that main
Menu1 is selected. Then go to the menu bar and type &File where the text
Type Here is shown, and press Return. This will add a top-level menu item
called File. The prefix & indicates that the shortcut to access the menu is
the Alt+F key combination. After that, change the name of the menu in the
properties section to FileMenu. Add a submenu item for the Open File
menu item, as shown in Figure 5.13.

Figure 5.13 Creating the File Menu

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 148

148 Chapter 5 The Photo Editor Application

Change the name in the property section of the submenu item to Open
File. To add functionality to the menu, double-click on the OpenFile
menu item. The PhotoEditor.cs source file will be opened, and the
Designer adds a stub implementation of the OpenFile_Click event han-
dler. The Designer also adds code to make the system aware of the function
that handles the event of the specified type. Listing 5.7 shows the gener-
ated code that initializes some properties of the OpenFile menu item and
registers the event handler method.

Listing 5.7 Adding an Event Handler Method

//
// OpenFile
//
this.OpenFile.Index = 0;
this.OpenFile.Text = "&Open File";
this.OpenFile.Click +=
new System.EventHandler(this.OpenFile_Click);

To load an image, we must add a new field to the PhotoEditorForm

class. To add new fields to a class using the Visual Studio class wizard, click
on the Class View tab, right-click on the class, and select Add | Add Field.
In the dialog window that opens, you specify the properties of the field. To
create the field, specify private as the field access, OpenFileDialog as the
field type, and loadFileDialog as the field name. Alternatively, simply add
the field to the class manually by adding the following line to the class:

private OpenFileDialog loadFileDialog;

At initialization time the loadFileDialog object is created and mem-

ory is allocated for it. To do that, add the following line to the constructor
PhotoEditorForm():

//
// TODO: Add any constructor code after InitializeComponent call
//
loadFileDialog = new OpenFileDialog();

Finally, you implement the loadFileDialog in the OpenFile_Click( )

method, as shown in Listing 5.8.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 149

5.6 Implementation of the Photo Editor Application 149

Listing 5.8 The File Open Dialog

loadFileDialog.Filter = " jpg files (*.jpg)|*.jpg| gif files

(*.gif)|*.gif| bmp files (*.bmp)|*.bmp| All files (*.*)|*.*";
loadFileDialog.ShowDialog();
loadedImage = new Bitmap(loadFileDialog.FileName);
this.Invalidate();

First, you define a file filter whose task it is to show only files of certain
file types within the chosen directory. Files of other types (or file exten-
sions) than the ones specified are not shown in the dialog window. This fil-
ter operation is provided by the OpenFileDialog class, which defines a
Filter property that can be used to filter files by their types. In our
example, the files that are of interest are image files of various types. The
supported types are images with the extensions .jpg, .gif, and .bmp. In
addition, we want to show all files in a directory when the file name is spec-
ified as “*.*”.
After the file filter is defined, the standard Windows File Open dialog
box is shown by calling its method ShowDialog(). A file dialog window
appears that enables the user to browse directories and select a file. After
the user has selected a file and clicked the Open button, the selected file
name can be extracted from the file dialog object using the FileName()
method. The .NET Framework provides converters to load and save the
most commonly used image types. For all image types supported by the
photo editor application, converters are provided by the .NET Framework.
Thus, we need no customized functionality to work with the various image
types. For supported formats and available conversion types, refer to the
MSDN help pages.
To work with the loaded image, we must create a Bitmap object and
allocate memory for it. We do this by calling the Bitmap constructor with
the file name as a parameter and assigning the image to loadedImage. The
loaded image field is not yet defined. Therefore, we add the following line
to the PhotoEditorForm class:
private Bitmap loadedImage;

As you can see, the application uses a bitmap type image to work with
rather than using the type under which the image was actually stored. It is
at this point that Windows’ automatic image type conversion saves a lot of
work. To show the loaded image on the screen, we next force a refresh,
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 150

150 Chapter 5 The Photo Editor Application

invalidating the current Windows form. Invalidating the window (or parts
of it) sends a paint message to the appropriate window: either the control
or the child window. As a result, the part of the screen that has been invali-
dated is redrawn using the Paint event handler method.
Before an image is loaded by the user, a default image will be shown.
The image provided is Hawaii.jpg and should be located in the bin and
bind directories if the debug version is run. To initialize the bitmap, simply
add a line to the PhotoEditorForm constructor:

loadedImage = new Bitmap(@"Hawaii.JPG");

The last step tells Windows that custom drawing is needed for this
form. To do that, you overridee the OnPaint method. To implement that
method, you can either add the event handler manually by typing or create
a stub implementation using Visual Studio.NET. To automatically create a
stub, click on PhotoEditorForm in the PhotoEditor.cs[Design] view, and
go to the properties section. If you press the yellow lightning symbol
underneath the Solution Explorer window, you will see a tab with all the
events listed. Double-click on the Paint event. This will create a stub
implementation. Then implement the code as shown in Listing 5.9. The
implementation reveals that the Graphics class is used to save the image in
memory, which is then displayed on the screen.

Listing 5.9 Overriding the OnPaint Method

/// <summary>
/// Custom paint method.
/// </summary>
protected override void OnPaint(PaintEventArgs e)
{
try
{
Graphics deviceContext = e.Graphics;
deviceContext.DrawImage(loadedImage, 0,0);
}
catch(Exception exception)
{
ExceptionManager.Publish(exception);
}
}
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 151

5.6 Implementation of the Photo Editor Application 151

Compile and run the project. The default image will be displayed as
the background image when the application is first started. To display
another image, go to File | Open File. The Open File dialog will be shown,
and you can browse the file system. Select an image, and it will be displayed
in the application window.

5.6.6 Scrolling and Basic Image Operations

Loading a large image or resizing the application window shows a drawback
of the current implementation: If the image is larger than the application
window, only part of the image is shown. A better behavior for a Windows
application is to show scrollbars if the image is larger than the window.
Because the photo editor should be a well-behaved application, we need to
add scrolling.
Depending on the needs of an application, there are several ways to
add scrolling capabilities. Table 5.4 gives a short overview of the various
techniques, describing how they are implemented and when they should
be used.

Table 5.4 Comparison of Scrolling Techniques

Scrolling
Technique Description When to Use

Using a scrollable These are controls directly or Use this scrolling technique
control indirectly derived from System. when there is no need to
Windows.Forms.Scrollable draw in the control with
Control. They support scrolling, GDI+, the control is com-
provided by the .NET Framework. posed of this custom con-
For example, TextBox, ListBox, and trol and other controls, and
the Form class itself support scrolling. the virtual space is limited.
Placing a non- An instance of the Picture control Use this technique if you
scrollable control in is created, and the Picture control want to be able to draw into
the Panel control is placed in a Panel control. You the image with GDI+, the
then create a new image in which custom control is not com-
you can draw (possibly using GDI+). posed of other custom con-
Then the background image of the trols, and the virtual space
custom control is set to the new is limited.
image (including the graphics you
were drawing). (continued)
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 152

152 Chapter 5 The Photo Editor Application

Table 5.4 (Cont.)

Scrolling
Technique Description When to Use

Using the Derive the control from the Use this technique if you
UserControl class UserControl class, build an image, need to draw into the lim-
with child controls draw into the image, and set the ited virtual space (possibly
BackgroundImage property of this using GDI+) and you’re
control to the build image. using child controls (con-
stituent controls) of the
custom controls.

Smooth scrolling Create a custom control that derives This technique creates a
from the User control, add vertical polished and professional
and horizontal scrollbars as child appearance. The image
(constituent) controls, and write a scrolls smoothly to the
Paint event handler that draws the desired position rather
image incrementally according to than jumping to the new
the scrollbar’s position. In addition, position, as it does when
the background color of the part of you use the build in
the control that is not covered by the scrolling.
image can be drawn in a defined
background color.

For the photo editor application, we choose smooth scrolling to give

the application a professional and polished appearance. This also lets us
support graphics drawn with GDI+, something that is necessary for the
supported graphics and text overlays described in the later chapters.

5.6.7 Refactoring of the Current Code

Before we begin implementing scrolling, we must do some refactoring of
the existing project code. Even though the implemented code works and
shows a quick result, we need to adapt it to meet the approved design and
to give us the flexibility to accommodate the implementation of functional
requirements to be added in later iterations.
According to the design specification, the Picture class should be split
out as an independent class. This makes perfect sense, because the image
and all its properties should be self-contained and separate from the GUI
code. To add a new class, go to Solution Explorer, right-click on Photo Edi-
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 153

5.6 Implementation of the Photo Editor Application 153

tor Application, and select Add | Add Class. Type Picture for the name of
the class to be created, and press Open. After that, add a reference to the
System.Drawing namespace to the new file by adding the following state-
ment:

using System.Drawing;

The reference is needed to access the type Bitmap, which is used to store
the image.
To implement the functionality to load an image, to the Picture class
we add a public method named LoadImage with the return type void and
no parameters. Switch to the class view and right-click on the Picture class
and Add | Add Method. Or simply add the following line:

public void LoadImage() {...}

We also add a field for storing the image data. Add the following line to
the Picture class:

private Bitmap loadedImage;

In C#, properties (also called accessors) are used to access private

field data. To add the property, right-click on the Picture class (in the Class
View tab of Solution Explorer) and select Add | Property. This will display a
dialog box as shown in Figure 5.14.
Enter the data as it is shown in Figure 5.14, and press Finish. The prop-
erty functionality with public accessor methods is generated automatically
and is added to the class. To actually return the loaded image when the get
method is called, change the return value from null to loadedImage. For
now, the set part of the property is not used. In theory you could set the
image by assigning the loaded image to the provided value:

loadedImage = value;

The next step is to add to the PhotoEditorForm class a public field that
is of type Picture and is named PictureObject. This field will hold a refer-
ence to an instance of a Picture class object. The PictureObject then
needs to be initialized with an allocated Picture class instance. This is
done in the constructor of the PhotoEditorForm. Add the following line to
the PhotoEditorForm() constructor:

PictureObject = new Picture();

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 154

154 Chapter 5 The Photo Editor Application

Figure 5.14 Adding the loadedImage Property

The previously added lines in the PhototEditorForm constructor need

to be deleted because the OpenFileDialog and the loadedImage function-
alities are now responsibilities of the Picture class. Therefore, we delete
the following lines in the PhotoEditorForm constructor:
loadFileDialog = new OpenFileDialog();
loadedImage = new Bitmap(@"Hawaii.JPG");

Next, in the OnPaint event handler method, the PictureObject.

LoadedImage should be painted instead of the default image. We imple-
ment this by changing the DrawImage call in the OnPaint event handler to
look like this:
deviceContext.DrawImage(PictureObject.LoadedImage, 0,0);

To load an image via the Picture class, we change the OpenFile_Click

event handler method to call the LoadImage method of the Picture class
by adding this line:
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 155

5.6 Implementation of the Photo Editor Application 155

PictureObject.LoadImage();

From the same method, we move the code for opening the file dialog
window to the LoadImage method of the Picture class. Basically this is all
the code except the following line:

this.Invalidate();

Now add the following statement to the Picture.cs file:

using System.Windows.Forms;

Then add a field for OpenFileDialog to the Picture class by adding

this line:

private OpenFileDialog loadFileDialog;

Now it is time to load the default image at startup and to create an

instance of OpenFileDialog in the Picture class constructor. Therefore,
we add the following two lines to the Picture class constructor:

string defaultImage =
PhotoEditorForm.GetApplicationDirectory + @"\Hawaii.jpg";
loadFileDialog = new OpenFileDialog();

You can see in this code that we have introduced an additional method
to the PhotoEditorForm with the name GetApplicationDirectory. When
you call this method, the path to the directory in which the photo editor
application was started is returned. This is necessary in case a user starts
the application from a different directory via the command line (in that
case, if we would search the current directory for the default image we
would search in the directory the command line shows and not the direc-
tory where the application was started). To make this work, add the follow-
ing lines to the PhotoEditorForm class:

// Get the directory in which the application was started.

// Note: In C# you can initialize a member at definition time.
// The compiler will take care of initializing the member in the
// constructor of the corresponding class.
private static string applicationDirectory = Application.StartupPath;
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 156

156 Chapter 5 The Photo Editor Application

/// <summary>
/// Accessor to the Application directory.
/// </summary>
public static string GetApplicationDirectory
{
get
{
return applicationDirectory;
}
}

To complete the refactoring, we delete the fields for loadFileDialog

and loadedImage from the PhotoEditorForm class. Check the implemen-
tation by compiling and running the application.
Even though the changes made in this section are not very large, they
show that refactoring can take a substantial amount of time. We think that
this time is well invested if developers consistently try to improve the exist-
ing code. This does not mean that the implementation should necessarily
become more complex by refactoring, but if developers identify possibili-
ties to improve existing code with regard to maintainability and extensibil-
ity (if needed), the refactoring should be implemented (in fact, refactoring
should simplify the code by making it easier for other developers to read and
understand). We also strongly recommend that you do refactoring in very
small steps. Refactoring can be a powerful tool if done consistently over the
life cycle of a project. On the other hand, refactoring can become a night-
mare if it is not done consistently throughout the whole life cycle of the
project and if it is not done in small chunks with thorough testing in place.

5.6.8 Creating a Custom Control for Smooth Scrolling

Even though Visual Studio provides a wide variety of controls and wizards,
sometimes it is necessary to develop controls with different, customized
behavior. For the scrolling in this project, we want to provide smooth
scrolling, something that the Visual Studio controls do not provide. There-
fore, we’ll develop a custom control. Before we start, it’s a good idea to
check whether a control that satisfies the needs of the project has already
been developed and is available on the Internet.
As mentioned earlier, we want to implement the smooth scrolling con-
trol to give our application a professional feel when users scroll an image.
The advantage of developing a custom control for this is that we can use the
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 157

5.6 Implementation of the Photo Editor Application 157

control in other applications if needed. Another advantage is certainly that

implementation of a custom control is a common task in application devel-
opment, and we want to show how to develop and use a custom control to
extend the features provided by Visual Studio.
For the implementation details of CustomScrollableControl, please
refer to the sample solution on the CD. The project can be found in the
Chapter5\src\Photo Editor directory. Instead of showing all the imple-
mentation details here, we explain the functionalities implemented and the
necessary properties without the implementation details. You can imple-
ment the custom control based on the description of the functionality, or
simply read through the text while checking the source code for the imple-
mentation details. If you try to implement the functionality, the Web com-
ment report, which is available in the doc directory, might be helpful. It
shows all members of the CustomScrollableControl class along with com-
ments explaining the functionality.
To create custom controls, add a new C# project to the photo editor
solution. The type of the project is a Windows control library, and the name
of the control is CustomScrollableControl. After the project and its files
are generated, change the name of the generated file and the class name
from UserControl1 to CustomScrollableControl and change the output
directories to bin and bind.

Do It Yourself Try to implement the CustomScrollableControl

feature. Use the description here, the comment Web pages in the
doc directory, and the sample solution to guide you through the imple-
mentation.

5.6.9 Implementation of CustomScrollableControl

All fields are defined as private, and accessor methods are provided for
fields that need to be accessed by other classes.
The Image scrollingImage field is used to store the image that is dis-
played in the control. The get property for this field returns the image; the
set method sets the scrollable image and calls a method to adjust the
scrollbars.
Another field needed is a point that specifies the viewport coordinates.
Point viewportCoords represents the coordinates of the image relative to
the control. The viewport defines the coordinates of the image relative
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 158

158 Chapter 5 The Photo Editor Application

to the window. If the image is not scrolled, the pixel at position (0/0) of the
picture is shown at position (0/0) of the control. If the image is scrolled by
100 pixels in y direction, the picture position (0/100) is shown at the (0/0)
position of the custom control.
Rectangle ScrollingImageArea is a convenience accessor that returns
a rectangle whose size is measured from the origin of the control to the
x-coordinate of the vertical scrollbar and the y-coordinate of the horizontal
scrollbar. This is equivalent to the area of the control that is available for
drawing the image and is defined as the client area minus the area that is
taken by the scrollbars.
The base functionality of this control is to scroll through an image
smoothly. Therefore, we add vertical and horizontal scrollbars to the Custom
ScrollableControl form. The scrollbars can be dragged from the toolbox
onto the form. The scrollbars are positioned and docked in the form to the
right and the bottom of the form, as in other Windows applications.
private void drawImage is a helper method that is used to calculate and
draw the correct portion of the bitmap in the custom control. This method is
called directly by the scrollbars whenever a change is detected. The method
clips the region of the image to the area of the image that is visible in the con-
trol and draws the image. To clip the region, we use a Rectangle method
that is defined in the Windows.System.Drawing namespace.
We customize the Paint event handler using private void Custom
ScrollableControl_Paint so that we can repaint parts or the entire image
(in case, for example, the image is restored after the control was minimized).
The GDI+ drawing surface, provided as a parameter to PaintEventArgs,
is stored in a local variable called graphics. Then a solidBrush is created
to fill the client area with a solid color. Next, we check whether the
scrollingImage exists. If it does not, then there is no image and the com-
plete client area is filled by the solid brush.
After that, a local variable of type Rectangle is created. The rectangle
to the right of the image and left of the scrollbar is calculated and stored in
the local variable rect. If the calculated rectangle is not empty, this area
will be filled with the solid brush. After that, we do the same thing for the
area below the image and above the horizontal scrollbar. Then the small rect-
angle in the lower-right corner is calculated and filled with the solid brush.
The private void adjustScrollBars method dimensions and sets
certain properties for the scrollbars. This method does not take any param-
eters. A constant field is defined that is used to calculate the number of
incremental steps for each scroll request. Then we check whether an image
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 159

5.6 Implementation of the Photo Editor Application 159

exists. If it does, the minimum and maximum values of the scrollbars are set
to 0 and the width or height of the image. In addition, we define the behav-
ior of small and large changes to the scrollbars. The actual values of the
scrollbars are set to the corresponding value of the viewport (meaning the
coordinates of the upper-left corner in the actual image).
The private void scroll method is the heart of the custom scrolling
functionality. This is the code that actually does the smooth scrolling. This
method handles the scrolling from the given previous position to the cur-
rent position of the scrollbars. To achieve a smooth scrolling effect, the
viewport is incrementally changed until it is in the new end position. In
between the incremental steps, the method is sleeping for a short time to
simulate the effect of a sliding image. Therefore, two constants are defined.
The first constant is used for the time period during which the control
sleeps before displaying the next image position relative to the viewport,
and the second is a divisor for the difference calculation of the previous and
the current position of the scrollbar.
We also create a local variable that holds the drawing context; this vari-
able is checked to see whether the previous value of the scrollbar is the
same as the current value. If it is not, we must apply horizontal scrolling. A
Boolean local variable indicates that the direction the scrollbar was moved,
and the defined integer divides the absolute change that was made into the
smaller, incremental steps. The incremental steps are then checked to see
whether they are smaller than 1. If they are, then the value is set to 1 for
scrolling up, or to –1 for scrolling down. Following that, the loop in which
the incremental scrolling over the image is executed.
Then some checks are added to make sure that scrolling is stopped if
the image is shown according to the scrollbar position and that the stepping
did not go too far (if it did, the values are set to the final position). Before
the image is drawn at its new position (with respect to the control), the con-
trol sleeps for a specified amount of time. Then the image is drawn, and the
next image position is calculated and displayed. This continues until the
image is shown in its final position, in which case a break statement is exe-
cuted what makes the program jump out of the while loop.
private void hScrollBar_Scroll and private void vScrollBar_
Scroll are the event handlers for the scrollbars. The Scroll event is trig-
gered whenever the user clicks on the scrollbar and changes its position. The
parameters that are passed to the event handler methods are references to
the sender’s object and ScrollEventArgs. The ScrollEventArgs object
provides information on the scroll type. If the user clicks on the small
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 160

160 Chapter 5 The Photo Editor Application

arrows of the scrollbar, either a ScrollEventType.SmallDecrement or a

ScrollEventType.SmallIncrement is provided. If the user clicks inside
the scrollbar or drags the scrollbar, a ScrollEventType.LargeDecrement or
ScrollEventType.LargeIncrement type is provided. The event handlers
for the scrollbars usually contain a switch statement, depending on the
scroll type. In the case of customScrollableControl, the event handler
stores the previous position of the viewport in a local variable before it
assigns the current position to the viewport. Then the scrolling method is
called with the previous position and the new position.
private void CustomScrollableControl_Resize implements the
event handler for the resize event. The resize event is triggered when-
ever the control window is resized. In that case, the scrollbar position must
be recalculated and the viewport may have to be updated. The implemen-
tation checks to see whether there is an image, and, if there is, the new
viewportCoords are calculated. The Math.Min method is used to return
the value of the smaller number that was provided. The Math.Max method
is used to return the maximum value of the provided parameters.
Build the custom control from the sample solution so that it can be
used in Visual Studio.NET.

5.6.10 Configuring the Custom Control for Use in Visual

Studio.NET
It’s easy to configure Visual Studio.NET to use the custom control. Just go
to the Tools menu. Choose Add/Remove Toolbox Items, and browse to the
directory in which the control was built (navigate to the bind directory if
the debug version is used; otherwise, go to the bin directory, assuming we
set the output path correctly) and choose the control, as shown in Figure
5.15. The control is then shown in the Toolbox. Like any other control, it
can be used by dragging it from the Toolbox onto the form.
To implement the custom scrolling in the photo editor application, add
the control to the form PhotoEditorForm[Design]. Position it as shown
in the requirements and dock it to the top, left, and right of the form (by
using the Dock property) and rename it customScrollableControl. Then
double-click on the new control to create the Load event handler.
The next step is to draw the image to the custom control instead of the
Form. In order to do that, remove the following line from the Paint event
handler:
deviceContext.DrawImage(PictureObject.LoadedImage, 0,0);
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 161

5.6 Implementation of the Photo Editor Application 161

Figure 5.15 Adding a Custom Control to the Toolbox

Change the OpenFile_Click event handler method to display the

image in the custom control, and invalidate the control to force a paint
event. This will show the image within the new control. Listing 5.10 shows
the new event handler.

Listing 5.10 The New OpenFile_Click Event Handler

/// <summary>
/// Opens a file dialog window so the user can select an
/// image to be loaded
/// </summary>
/// <param name="sender">A reference to the object calling
/// this method</param>
/// <param name="e">The event arguments provided by the
/// event handler</param>
/// <requirements>F:editor_load_and_save</requirements>
private void OpenFile_Click(object sender,
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 162

162 Chapter 5 The Photo Editor Application

System.EventArgs e)
{
try
{
PictureObject.LoadImage();
if(PictureObject.LoadedImage == null)
throw(new Exception("Error, image could not be
loaded"));
DisplayImage();
}
catch(Exception exception)
{
ExceptionManager.Publish(exception);
}
}

In addition, a new DisplayImage method is introduced in this example.

This method is implemented in the PhotoEditorForm class.
Its implementation sets the scrolling image of the custom control to
the currently loaded image of the Picture class instance, and it invalidates
the custom control. The implementation of the DisplayImage method is as
follows:
public void DisplayImage();
customScrollableControl.ScrollingImage = PictureObject.LoadedImage;
customScrollableControl.Invalidate();

The application, in its first cut, shows the default image when first
loaded. To get the same result with the custom control, we call the newly im-
plemented DisplayImage() method from within the customScrollable
Control_Load event handler (which was generated by Visual Studio.NET
automatically by the double-click on customScrollableControl in the
PhotoEditor.cs[Design] tab).
This completes the implementation of smooth scrolling. It is now time
to test the implemented functionality. Running the application and loading
an image shows that the scrolling works smoothly, but the image flickers
when the scrollbar is moved. We can prevent this by setting a style property
in the CustomScrollableControl constructor. As a result of the change,
the background is not drawn in the background color before the image is
drawn. The following line is used in the constructor to accomplish this:
this.SetStyle (ControlStyles.Opaque, true);
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 163

5.6 Implementation of the Photo Editor Application 163

After this change, the control works without any noticeable flicker. The
next step is to provide the tab control containing the buttons for the basic
image operations.

5.6.11 Cropping an Image

Cropping an image means to cut a particular region out of the image, keep
the cropped portion, and delete the rest. For the photo editor, users will
crop their images to get them to a standard size and upload them to the
online shop. In a later version, we might provide interactive graphics for
this task, but for the first version we present a dialog box where users spec-
ify the new size of the image.
To implement the crop dialog box, choose PhotoEditorForm and go to
the Toolbox to drag a TabControl onto the form. In the properties of the
tab control, select the following properties:

Properties of TabControl
(Name) tabControl
Dock Bottom

A tab control is used to group related image-processing functionality.

Later in the book we will add other tab controls. Now rename the tab con-
trol to tabControl. Add a tab to the control by selecting and then right-
clicking on the tabControl. Then select Add Tab to add the actual tab. Set
its properties as follows:

Properties of Tab
(Name) basicImageOperations
Text Basic Image Operations

When the photo editor application is now run, the screen layout should
correspond to the GUI shown in the photo_editor requirements, except
that the buttons are still missing. To add a button for the cropping function-
ality, drag a button to the tab. Change the displayed text on the button to
Crop Image, and change the name of it to cropImageButton. Double-click
on the button to add the event handler for the click event. We use a dialog
box to collect the size information for the crop rectangle. To let users open
a dialog box when the crop button is pressed, you must add a new form to
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 164

164 Chapter 5 The Photo Editor Application

the photo editor application. You do this by right-clicking the Photo Editor
Application project in Solution Explorer and selecting Add | Add Windows
Form. Name the new form CropDialog. Then drag two text boxes onto the
form and change their properties as follows:

Properties of TextBox1 (Left) and TextBox2 (Right)

(Name) imageWidth imageHeight
AcceptsReturn True True
AcceptsTab True True
Text Image Width ImageHeight

In addition, add two buttons; change their names to OKBtn and CancelBtn
(and change the text to be displayed on the button accordingly). Also add
two labels next to the text boxes that explain what the text box input is—for
example, “Enter new image height.” The Form should now look approxi-
mately like Figure 5.16.
After adding all the design-related properties, we add the event han-
dlers. A double-click on the image width text box will add an imageWidth_
TextChanged event handler. Before implementing the event handler, add
to the CropDialog class two private integer variables called tempWidth
and tempHeight for the width and height. The event handlers are then
implemented. To extract the new width from the text box entry, the image
Width object provides a property, Text, that represents the text in the text
box. This text is converted to an integer value and stored in the tempWidth
variable. The functionality is shown in Listing 5.11.

Figure 5.16 The Crop Dialog Box

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 165

5.6 Implementation of the Photo Editor Application 165

Listing 5.11 The TextChanged Event Handler for the Width

private void imageWidth_TextChanged(object sender,

System.EventArgs e)
{
tempWidth = Convert.ToInt32(imageWidth.Text);
}

Do It Yourself Implement the height text box event handler method

using Listing 5.11 as a template.

Next, we implement the OK and Cancel buttons. We start with the

implementation of the Cancel button. To add the event handler for the but-
ton, double-click on the Cancel button in the CropDialog.cs[Design]
view. If the Cancel button is selected, no cropping action is done and the
PhotoEditor form should be invalidated (to force a repaint). In addition,
the CropDialog box can be disposed of to indicate to the garbage collector
that the memory is no longer being used. Therefore, we change the con-
structor of CropDialog to accept one argument of type object named
sender. This object will be a reference to the calling PhotoEditorForm
object. The sender object is then cast in the constructor to a PhotoEditor
Form object and is stored in a local private variable of type PhotoEditor
Form named editorForm. The code to be added to the constructor, after
creating the private field, is as follows:

editorForm = (PhotoEditorForm)sender;

The CancelBtn_Click event handler is then implemented as shown in

Listing 5.12.

Listing 5.12 The CropDialog Cancel Button

private void CancelBtn_Click(object sender, System.EventArgs e)

{
editorForm.Invalidate();
this.Dispose();
}
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 166

166 Chapter 5 The Photo Editor Application

After that, we implement the OK button click event handler for the
Crop dialog box. When the OK button is pressed, the stored values for
width and height are sent to PictureObject.CropImage. The Picture
object is then responsible for cropping the loaded image to the specified
size. Therefore, we add the event handler by double-clicking on the OK
button and adding the following lines to the event handler:
editorForm.PictureObject.CropImage(tempWidth, tempHeight);
editorForm.DisplayImage();
this.Dispose();

This will crop the image, assuming that a CropImage method is pro-
vided by the Picture class. Therefore, we must add the CropImage method
to the Picture class in the next step.
Add a new public void method CropImage to the Picture class. This
method takes two integer variables (the width and the height). Now that we
have defined the signature, let’s take care of the implementation. The
CropImage method should check whether the provided parameters specify
a region within the loaded image and whether the parameters are actually
larger than zero. After that, the clip region needs to be calculated as a rect-
angle. We then clone the current image by applying the calculated crop-
ping rectangle, and we store a copy of the cropped image in a temporary
bitmap called croppedImage. The loadedImage is then set to the cropped
image and is displayed. The complete implementation of the CropImage
method is shown in Listing 5.13.

Listing 5.13 The CropImage Method

/// <summary>
/// Method called from CropDialog. The current
/// shown image is cropped to the size provided
/// in the parameters. The cropping is done
/// with a rectangle whose center is put on
/// the center of the image.
/// </summary>
/// <requirement>F:image_crop</requirement>
/// <param name="newHeight>height of the cropped
/// image</param>
/// <param name="newWidth">width of the cropped
/// image</param>
public void CropImage(int newWidth, int newHeight)
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 167

5.6 Implementation of the Photo Editor Application 167

{
// Check that cropping region is
// actually within the image and that
// the values provided are positive
if((newWidth < loadedImage.Size.Width) &&
(newHeight < loadedImage.Size.Height) &&
(newHeight > 0) && (newWidth > 0))
{
int xmin = (loadedImage.Size.Width / 2) -
(newWidth / 2);
int xdim = newWidth;
int ymin = (loadedImage.Size.Height / 2) -
(newHeight / 2);
int ydim = newHeight;
Rectangle rectangle = new Rectangle(xmin, ymin,
xdim, ydim);
if(rectangle.IsEmpty)
{
throw(new Exception("Error, CropImage failed to
allocate clipping rectangle"));
}
Bitmap croppedImage = loadedImage.Clone(rectangle,
System.Drawing.Imaging.PixelFormat.DontCare);
Bitmap oldImage = loadedImage;
loadedImage = new Bitmap(croppedImage,
rectangle.Size);
if(loadedImage == null)
{
throw(new Exception("Error, Image memory allocation
failed"));
}
}
}

The last step is to implement the cropImage button click event handler.
First, create the event handler. Then in the event handler create a new
CropDialog object and show the dialog on the screen by adding the follow-
ing lines:

CropDialog openDialog = new CropDialog(this);

openDialog.Show();
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 168

168 Chapter 5 The Photo Editor Application

This completes the implementation of the cropping functionality.

Compile and run the project to see whether it works. If you added the
XML code documentation while implementing the functionality, then you
can also generate the comment Web pages and you will have a nice descrip-
tion of the newly added functionality.

5.6.12 Rotate and Flip an Image

Now we show how to implement the rotate and flip functionality. To sup-
port this, we add new buttons to the tab card of PhotoEditor.cs[Design].
Add the buttons to match the screen layout shown in Figure 5.17. Then
rename them and change the text of the buttons.
To add the event handler method for horizontal flipping, double-click
on the corresponding button. Before adding the event handler, though, you
should consider that the flipping and rotating of an image are basic image
operations and therefore they should be handled by the Picture class.
Therefore, add to the Picture class a public void method called Rotate
FlipImage that takes a parameter of type RotateFlipType that is named
rotateFlip. (The Bitmap class actually supports rotating and flipping and
provides an enumeration for various rotation and flip values.) Implement
the event handlers for the button click events by passing the correct
RotateFlipType to the RotateFlipImage method of the Picture class.
(For information about RotateFlipType, see the MSDN documentation
or the sample solution.) Then the RotateFlipImage method can be imple-
mented as shown in Listing 5.14.

Listing 5.14 The RotateFlipImage Method

public void RotateFlipImage(RotateFlipType rotateFlip)

{
loadedImage.RotateFlip(rotateFlip);
}

Figure 5.17 The Photo Editor Form Buttons

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 169

5.6 Implementation of the Photo Editor Application 169

Do It Yourself Implement the functionality for vertical flipping, rotating

clockwise, and rotating counterclockwise in the same way the horizontal
flipping functionality was added. Use the enumeration members of
RotateFlipType.

5.6.13 Save an Image

Before testing is started, we need to implement the last feature that is miss-
ing in the implementation. We have not implemented the functionality to
save an image on disk. This task is very similar to the implementation of
image loading. Instead of OpenFileDialog, we use SaveFialeDialog; and
instead of loading the image, we save the image under the user-selected
name.
Go to the [Design] view of PhotoEditor.cs and add a new menu item
in the main menu, underneath the Open File entry. Name the menu item
saveMenu, and change the text to &Save as .... Add the event handler
for the click event. The save file functionality is also implemented in the
Picture class. Therefore, add a public method called SaveFile() to the
Picture class, and call it from the event handler of the Save button.
To make SaveFileDialog work, add a private member variable to the
Picture class of type SaveFileDialog, and in the constructor create the
object for it. The SaveFile method then can be implemented as shown in
Listing 5.15.

Listing 5.15 The SaveFile Method

public void SaveImage()

{
saveFileDialog.Filter = " jpg files (*.jpg)|*.jpg|
gif files (*.gif)|*.gif| bmp files (*.bmp)|*.bmp|
All files (*.*)|*.*";
saveFileDialog.ShowDialog();
loadedImage.Save(saveFileDialog.FileName);
saveFileDialog.Dispose();
}

After this change has been made, compile and run the application to
see the result. The images can now be loaded, rotated, flipped, cropped,
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 170

170 Chapter 5 The Photo Editor Application

and saved. This is all the functionality that is needed for the project in the
elaboration phase. The next task is to write unit tests to validate the imple-
mented functionality.

Do It Yourself The buttons provided are not very nice-looking. You can
make the program more appealing by adding bitmaps that show the user
what the result of a requested image transformation will be. (Customize
the rudimentary buttons provided, or develop custom bitmaps that can be
loaded onto the buttons.) Also, we recommend that you change and play
around with the implementation to see what improvements you can make.

Figure 5.18 shows the working application.

Figure 5.18 The Working Photo Editor Application

29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 171

5.7 Unit Tests 171

5.7 Unit Tests

The goal for the unit tests is to verify that all the required functionality is
T actually implemented and that the various modules are working in isola-
tion. For the photo editor application, most of the required functionality is
implemented in the Picture class. The only exception is the GUI. It is
cumbersome to test the GUI automatically, and it is more an integration
test rather then a unit test task. Therefore, the F:photo_editor key is not
explicitly tested at the unit test level.
To test the load and save functionality, it would be an advantage to have
overloaded methods where the test could specify the file name to be loaded
or saved. With the current implementation of the Picture class, the file
dialog would open every time the load functionality is called, making the
automated tests very difficult. You can use a commercial capture replay tool
to exercise these kinds of actions. Tools such as WinRunner do a good job
on those things. But as mentioned, the easier solution (and probably the
nicer solution as well) is to add two overloaded methods to the Picture
class. The one overloaded method is the load method, which accepts a
parameter of type string; the other is the save method, which also accepts
a string parameter. The overloaded methods then load or save the picture
under the provided file name. The implementation can be seen in Listing
5.16.

Listing 5.16 The Overloaded Save and Load Methods of the Picture
class

/// <summary>
/// Opens the file with the provided name.
/// </summary>
/// <param name="fileName">name of the file to be opened</param>
/// <requirement>F:image_load_and_save</requirement>
public void LoadImage(string fileName)
{
Bitmap oldImage;
oldImage = loadedImage;
loadedImage = new Bitmap(fileName);
if(loadedImage == null)
{
throw(new Exception("Error, LoadImage with file name
failed"));
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 172

172 Chapter 5 The Photo Editor Application

}
oldImage.Dispose();

}
/// <summary>
/// Saves the current image
/// with the provided fileName.
/// </summary>
/// <param name="fileName">Name under which the image is
saved</param>
public void SaveImage(string fileName)
{
loadedImage.Save(fileName);
}

These two hooks can now be used by the test program to actually load
and save an image without having to go through a file load and save dialog
box. Very often, such hooks are implemented for testing. This is another
reason to include the test team (if there is a separate test team available) in
the project planning from the beginning. In this way, hooks for testing can
be discussed in the planning phase of the project and implemented as fea-
tures during development. In the photo editor project, this not as impor-
tant because the development team also does the testing. Nevertheless,
testing is incorporated in the project planning from the beginning.

5.7.1 The NUnit Test Framework

To test the Picture class functionality, an automated test framework would
be very helpful. The framework we are looking for should be capable of
running specified tests automatically, and it should show the result on the
screen. Ideally it would have a GUI we could use to run selected tests, or it
would run the tests from the command line (to let us run the tests every
night using a script). It also should be easy to use so that we don’t have to
spend much time in training and setting up the test framework.
Luckily, such a test framework exists. Called NUnit, this framework is
an automated test framework that is implemented for many programming
languages in a similar form (for example, JUnit for Java, CppUnit for C++,
and many more). The unit test framework was first developed for Smalltalk
by the extreme programming group around Kent Beck. Later, the frame-
work was ported to many languages and is now also available for Microsoft
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 173

5.7 Unit Tests 173

.NET as NUnit. The framework can be downloaded from http://www.

nunit.org. A quick-start guide and other documentation can be found at
the site.

5.7.2 Unit Test Implementation

After downloading and installing the NUnit framework, we develop the
tests (you will also find an installation of the NUnit test framework in the
src directory). First, add a new class to the photo editor application project
with the name UnitTest.
Add a reference to NUnit.framework.dll (which can be found in the
src\NUnit20\bin directory), and add a using statement for NUnit.
Framework. To use the classes provided by the Picture class, also add a
using statement for the System.Drawing namespace that defines Rotate
FlipType. To tell the framework that the class UnitTest contains unit test
methods that are executable by NUnit, we add an attribute called [Test
Fixture] before the class definition. Before implementing the actual test
methods, we implement the setup and tear-down methods of the tests by
inheriting two attributes from the unit test framework. The SetUp method
is called before a test method is called, and the TearDown method is called
after the test method is executed. The attributes used to indicate these
methods are [SetUp] and [TearDown].
For the unit tests of the Picture class, the SetUp method creates a new
Picture object and writes to the standard output that a new test has started.
In the TearDown method, the image is disposed of and an indication of the
test result (passed or failed) is written to the standard output. Listing 5.17
shows the SetUp and TearDown implementation.

Listing 5.17 The Unit Test SetUp and TearDown Methods

namespace UnitTest
{
[SetUp] public void Init()
{
Console.WriteLine("********* New Test-case: *******");
Console.WriteLine("photo_editor");
TestImage = new Photo_Editor_Application.Picture();
}

private Photo_Editor_Application.Picture TestImage;

private bool amIPassed = false;
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 174

174 Chapter 5 The Photo Editor Application

[TearDown] public void Destroy()

{
TestImage.LoadedImage.Dispose();
if(amIPassed)
{
Console.WriteLine("=> PASSED");
}
else
{
Console.WriteLine("%%%%%% Failed");
}
}
}

For the file load test, we load the default image and check whether the
loaded image has the correct dimensions. The save file test crops the image
and saves it under a new file name. Then the regular default image is
loaded again (to make sure the Picture object has been changed and has
the original dimensions again) before the saved image is loaded. The saved
image is checked to see whether its dimensions are the same as the crop-
ping image information that was provided before the image was saved. The
XML tag <requirement> </requirement> identifies which requirement
key is tested in this test. To indicate that a method is a test method, the
attribute [Test] is used before the method definition. The implementation
can be seen in Listing 5.18.

Listing 5.18 Image Load and Save Test

/// <summary>
/// Test for F:image_load_and_save.
/// Windows standard dialogs are used, so really
/// this tests only whether the default image loaded has the
/// dimensions
/// that are expected!
/// </summary>
/// <requirement>F:image_load_and_save</requirement>
[Test] public void LoadandSaveImageTest()
{
const string fileName = "Test.jpg";

Console.WriteLine("image_load_and_save");
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 175

5.7 Unit Tests 175

Assertion.AssertEquals("Load Image, width",

TestImage.LoadedImage.Width, 2048);
Assertion.AssertEquals("Load Image, height",
TestImage.LoadedImage.Height, 1536);

TestImage.CropImage(200, 400);
TestImage.SaveImage(fileName);

TestImage.LoadImage("Hawaii.jpg");

TestImage.LoadImage(fileName);

Assertion.AssertEquals("Load Image, width",

TestImage.LoadedImage.Width, 200);
Assertion.AssertEquals("Load Image, height",
TestImage.LoadedImage.Height, 400);
amIPassed = true;
}

We perform the actual test case by calling an Assertion method with

the pass/fail criteria. If the test case fails, then the test is failed and the rest
of the tests within this method are not executed even if there are more test
cases defined. That is why the flag amIPassed can be set to true in case the
end of the method is actually reached.

Do It Yourself Implement test cases for image cropping, rotating, and

flipping in the same way it was done for loading and saving. Check the
unit test project into Source Safe. A sample solution is provided with the
source code on the accompanying CD.

Now that the test cases have been implemented it is time to run them
to see whether the developed program actually does what it is expected to
do. We can run the test either from the command line or via the NUnit
GUI. In this chapter only the GUI is shown.
To start the tests, go to the Start menu; then go to Programs |
NUnitV2.0 | NUnit-GUI. The NUnit GUI will open. Choose File | Open
and navigate to the directory where the unit tests were built. The test tree
will be shown, and the tests can be run by pressing the Run button. The
result will look like Figure 5.19.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 176

176 Chapter 5 The Photo Editor Application

Figure 5.19 The NUnit GUI

The GUI gives a nice overview of the test result. If the progress bar is
green, it means that all test cases passed. Otherwise, the bar will be red.
The Standard Out tab shows the results that were logged during the
test run. If there are errors, information about the failed cases can be found
in the Errors and Failures tab.

5.8 Conclusion

During the elaboration phase of our sample project, we design, develop,

and implement a functional architectural baseline. We run the implemen-
tation and perform initial unit tests. We also update the project planning.
What remains is to produce the documentation from the source code.
To provide a user’s manual in the form of HTML pages, we produce the
comment Web pages by going to the Tools menu and selecting the option
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 177

5.8 Conclusion 177

Build Comment Web Pages. In the displayed message box, we specify the
destination where the files will be saved to (in our case, the documentation
is stored in the doc folder of the project). Sample comment Web pages can
be found in the doc folder in the sample solution. In addition, we produce
the XML documentation, which is used for tracking (because it shows the
requirement keys that have been implemented).
To produce the XML files, go to Solution Explorer, right-click on Photo
Editor Application, and choose Properties. In the Configuration Properties
dialog, specify the file name to be used as the XML documentation file.
The XML file will be generated with every build, and the file will be saved
in the build directory (bin or bind in the case of the photo editor project).
The compiler will now create warnings if public members of a class do not
specify an XML-style comment.
In addition to the documentation, we apply a label to the software and
the documents (at least it should be done for the manually generated docu-
ments such as the project plan, the requirements document, and so on). To
add a label in Visual Source Safe, open the Source Safe application and
select the project. In the File menu choose Label. A window opens where
you can specify the label. The label will be something like “Version 0.1.0.”
This complies with the Microsoft .NET standard, which uses three num-
bers to identify a specific version. Because the version produced is the
first intermediate result, it is labeled with 0 (zero) for the main release, 1
for the intermediate release, and 0 for the minor releases. In addition, the
AssemblyInfo.cs file of the Photo Editor Application should be adjusted
to correspond to the version in the label before checkin.

5.8.1 Review
We must review the status of the project to decide whether the project will
be continued. If the project is continued and an agreement with the cus-
tomer is signed, then we check whether the project is ready to proceed to
the construction phase. To decide whether the project is ready, we assess
whether we have met the goals for the five core workflows:

■ Requirements: Refine the requirements and system scope.

■ Analysis: Analyze the requirements by describing what the system
does.
■ Design: Develop a stable architecture using UML.
■ Implementation: Implement the architectural baseline.
■ Test: Test the implemented architectural baseline.
29823 05 pp115-178 r4jm.ps 8/6/03 3:53 PM Page 178

178 Chapter 5 The Photo Editor Application

It can be seen that the project meets all goals that were set for this
phase and iteration. Therefore, the project is ready to move on to the next
phase.
To get customer feedback early on, we deliver the project to the cus-
tomer as intermediate V 0.1.0. It is crucial to deliver intermediate results
that are functional but not yet the final product. In this way, the customer
has a better understanding of the progress, and any changes that might be
requested by the customer can be discussed and implemented as early as
the intermediate project is delivered instead of at the end of the project.
Especially with GUI development, this is very important. Another advan-
tage of intermediate deliveries is that errors may be found by the customer
and communicated to the development team while the product is still in
development.

5.9 References for Further Reading

UML
Jim Arlow and Ila Neustadt, UML and the Unified Process (London,
England: Addison-Wesley, 2002)
Martin Fowler, UML Distilled (Reading, MA: Addison-Wesley, 1999)
.NET Programming
www.dotnetexperts.com/resources/
www.gotdotnet.com
www.msdn.microsoft.com

The NUnit Test Framework

http://sourceforge.net/projects/nunit/
http:www.nunit.org
http:/www.xprogramming.com/software.htm