Extensions to DrawTools

Introduction

Alex Fr provided an excellent set of drawing tools in his DrawTools article
and these tools serve as a basis for this article, which expands on the original toolset in the following ways:

In addition to the basic Rectangle, Ellipse, Line and Scribble tools, this version adds PolyLine, Filled Ellipse, Filled Rectangle, Text and Image tools
Multiple drawing Layers
Zooming
Panning
Rotation

In this article, I will describe how Layerswere implemented, as well as the Text and Image tools.

Background

See the original DrawTools article for details on how the basic application is
built, class structure, etc.

It is also assumed that the reader has a working understanding of GDI+ fundamentals, including Matrices. For an excellent introduction to GDI+, see www.bobpowell.net.

Implementing Layers

Adding Layersto the application involved adding two classes,

Layer

and

Layers

,
where

Layer

defines a single

Layer

and

Layers

defines
the collection of

Layers

in an

ArrayList

.

Each

Layer

exposes the following properties:

Hide Copy Code

private string _name;
private bool _isDirty;
private bool _visible;
private bool _active;
private GraphicsList _graphicsList;

Note that the

Layer

contains the

GraphicsList

-
this is the key to the whole thing - each

Layer

contains its own list of drawing objects instead of

DrawArea

.

DrawArea

is
modified to declare a

Layers

collection instead of a

GraphicsList

collection:

Hide Copy Code

// Define the Layerscollection
private Layers_layers;

When

DrawArea

is initialized, the

Layers

are
initialized by creating the first

Layer

and setting it

Active

and

Visible

:

Hide Copy Code

public DrawArea()
{
// create list of Layers, with one default active visible layer
_layers = new Layers();
_layers.CreateNewLayer("Default");
_panning = false;
_panX = 0;
_panY = 0;
// This call is required by the Windows.Forms Form Designer.
InitializeComponent();
}

In the

Layers

class, the

CreateNewLayer()

method
actually creates the new

Layer

:

Hide Copy Code

/// <summary>
/// Create a new layer at the head of the layers list and set it
/// to Activeand Visible.
/// </summary>
/// <param name="theName">The name to assign to the new layer</param>
public void CreateNewLayer(string theName)
{
// Deactivate the currently active Layer
if(layerList.Count > 0)
((Layer)layerList[ActiveLayerIndex]).IsActive= false;
// Create new Layer, set it visible and active
Layerl = new Layer();
l.IsVisible = true;
l.IsActive= true;
l.LayerName = theName;
// Initialize empty GraphicsList for future objects
l.Graphics = new GraphicsList();
// Add to Layerscollection
this.Add(l);
}

Note that any one or all

Layers

can be visible at the same time, but only one

Layer

may
be active at any time.

You can control the

Layers

in the sample application by clicking on the Current

Layer

:
name at the bottom of the application window - Click on the name ("

Default

") to open the

Layers

dialog:



From this dialog, you can Add new

Layers

, change the names of the

Layer

(s),
and change the

Layer

(s) visibility and which

Layer

is

Active

.
The "New Layer" column is checked whenever you click the "Add Layer" button. To delete Layer(s), simply check the "Deleted" column and close the dialog with the "Close" button. Remember only one Layermay be active at any one time. You will be reminded of
this if you attempt to have more than one

Layer

active. Also note the Active

Layer

must
be

Visible

.

When the application runs, each object that is drawn is added to the

GraphicsList

maintained by the active

Layer

.
Note this relationship is preserved through saving and re-opening a drawing file.

Layerscome in very handy when you want to draw "on top of" another image. For example, the image at the top of this article contains two layers. The following image shows the same picture with the Background Layerturned off:



Here is the same drawing with the Drawing Layerinvisible and the Background Layervisible:



Objects on Layerswhich are visible but not active cannot be selected, moved, deleted, etc.

Each drawing object is added to the correct

Layer

by the

AddNewObject()

method
in the

ToolObject

class:

Hide Copy Code

protected void AddNewObject(DrawAreadrawArea, DrawObject o)
{
int al = drawArea.TheLayers.ActiveLayerIndex;
drawArea.TheLayers[al].Graphics.UnselectAll();
o.Selected = true;
o.Dirty = true;
drawArea.TheLayers[al].Graphics.Add(o);
drawArea.Capture = true;
drawArea.Refresh();
} 

Implementing Zooming, Panning, and Rotation

Zooming, Panning, and Rotation are implemented by adding a few variables and some code to the

MainForm

and

DrawArea

classes.

Zooming is controlled by buttons on the form, and also by the mouse wheel when Ctrl is held down.

Pan is controlled by the Hand button on the form, and can be cancelled by a right-click.

Rotation is controlled by buttons on the form - note Rotation affects the entire drawing.

Here is an example of all three in use:



The heart of this code is the

BackTrackMouse()

method, which takes the "apparent" mouse position and converts it to
a valid point based on the current Zoom level, Pan position, and Rotation:

Hide Copy Code

/// <summary>
/// Back Track the Mouse to return accurate coordinates regardless of
/// zoom or pan effects.
/// Courtesy of BobPowell.net <seealso cref="http://www.bobpowell.net/backtrack.htm"/>
/// </summary>
/// <param name="p">Point to backtrack</param>
/// <returns>Backtracked point</returns>
public Point BackTrackMouse(Point p)
{
// Backtrack the mouse...
Point[] pts = new Point[] { p };
Matrix mx = new Matrix();
mx.Translate(-this.ClientSize.Width / 2, -this.ClientSize.Height / 2,
MatrixOrder.Append);
mx.Rotate(_rotation, MatrixOrder.Append);
mx.Translate(this.ClientSize.Width / 2 + _panX, this.ClientSize.Height / 2 +
_panY, MatrixOrder.Append);
mx.Scale(_zoom, _zoom, MatrixOrder.Append);
mx.Invert();
mx.TransformPoints(pts);
return pts[0];
}

This routine comes from Bob Powell's excellent website. Through the use of the
GDI+ Matrix class, the mouse point passed to this method is moved (Translate), Rotated, and Scaled based on the current PanX, PanY, Zoom, and Rotation values. The important thing to remember is that anytime you need to determine where the mouse pointer actually
is in your drawing, you must call this method. You will see this method used throughout the program in the

DrawArea

class
as well as others. An example of its usage is shown here:

Hide Copy Code

private void DrawArea_MouseDown(object sender, System.Windows.Forms.MouseEventArgs e)
{
lastPoint = BackTrackMouse(e.Location);
if (e.Button == MouseButtons.Left)
tools[(int)activeTool].OnMouseDown(this, e);
else if (e.Button == MouseButtons.Right)
{
if (_panning == true)
_panning = false;
ActiveTool = DrawArea.DrawToolType.Pointer;
}
//OnContextMenu(e);
}

The current zoom level is controlled by the following simple routine:

Hide Copy Code

private void AdjustZoom(float _amount)
{
drawArea.Zoom += _amount;
if (drawArea.Zoom < .1f)
drawArea.Zoom = .1f;
if (drawArea.Zoom > 10)
drawArea.Zoom = 10f;
drawArea.Invalidate();
SetStateOfControls();
}

Then in the

DrawArea.Paint()

method, the zoom, pan, and rotation values are used to alter the way the canvas is painted:

Hide Shrink

Copy
Code

private void DrawArea_Paint(object sender, System.Windows.Forms.PaintEventArgs e)
{
Matrix mx = new Matrix();
mx.Translate(-this.ClientSize.Width / 2, -this.ClientSize.Height / 2,
MatrixOrder.Append);
mx.Rotate(_rotation, MatrixOrder.Append);
mx.Translate(this.ClientSize.Width / 2 + _panX, this.ClientSize.Height / 2 +
_panY, MatrixOrder.Append);
mx.Scale(_zoom, _zoom, MatrixOrder.Append);
e.Graphics.Transform = mx;

SolidBrush brush = new SolidBrush(Color.FromArgb(255, 255, 255));
e.Graphics.FillRectangle(brush,
this.ClientRectangle);
// Draw objects on each layer, in succession so we get the correct layering.
// Only draw layers that are visible
if (_layers != null)
{
int lc = _layers.Count;
for (int i = 0; i < lc; i++)
{
if(_layers[i].IsVisible == true)
if(_layers[i].Graphics != null)
_layers[i].Graphics.Draw(e.Graphics);
}
}
DrawNetSelection(e.Graphics);
brush.Dispose();
}

Update - 8/25/2007 - Individual Object Rotation & Bug Fixes

The primary advancement in this update is the ability to rotate individual objects - when one or more objects are selected, clicking the Rotate tools will rotate those objects instead of the entire drawing surface.

There is one caveat, however - the selection rectangle for the rotated object is not rotated - if someone can help with this, I would greatly appreciate it!

This update also includes several small bug fixes reported by users - thanks to all for reporting!