
This
Graphics2D class extends the
Graphics class to provide more sophisticated
control over geometry, coordinate transformations, color management,
and text layout. This is the fundamental class for rendering
2-dimensional shapes, text and images on the Java(tm) platform.
Coordinate Spaces
All coordinates passed to a
Graphics2D object are specified
in a device-independent coordinate system called User Space, which is
used by applications. The
Graphics2D object contains
an
AffineTransform object as part of its rendering state
that defines how to convert coordinates from user space to
device-dependent coordinates in Device Space.
Coordinates in device space usually refer to individual device pixels
and are aligned on the infinitely thin gaps between these pixels.
Some Graphics2D objects can be used to capture rendering
operations for storage into a graphics metafile for playback on a
concrete device of unknown physical resolution at a later time. Since
the resolution might not be known when the rendering operations are
captured, the Graphics2D Transform is set up
to transform user coordinates to a virtual device space that
approximates the expected resolution of the target device. Further
transformations might need to be applied at playback time if the
estimate is incorrect.
Some of the operations performed by the rendering attribute objects
occur in the device space, but all Graphics2D methods take
user space coordinates.
Every Graphics2D object is associated with a target that
defines where rendering takes place. A
GraphicsConfiguration object defines the characteristics
of the rendering target, such as pixel format and resolution.
The same rendering target is used throughout the life of a
Graphics2D object.
When creating a Graphics2D object, the
GraphicsConfiguration
specifies the default transform for
the target of the Graphics2D (a
Component or Image). This default transform maps the
user space coordinate system to screen and printer device coordinates
such that the origin maps to the upper left hand corner of the
target region of the device with increasing X coordinates extending
to the right and increasing Y coordinates extending downward.
The scaling of the default transform is set to identity for those devices
that are close to 72 dpi, such as screen devices.
The scaling of the default transform is set to approximately 72 user
space coordinates per square inch for high resolution devices, such as
printers. For image buffers, the default transform is the
Identity transform.
Rendering Process
The Rendering Process can be broken down into four phases that are
controlled by the
Graphics2D rendering attributes.
The renderer can optimize many of these steps, either by caching the
results for future calls, by collapsing multiple virtual steps into
a single operation, or by recognizing various attributes as common
simple cases that can be eliminated by modifying other parts of the
operation.
The steps in the rendering process are:
-
Determine what to render.
-
Constrain the rendering operation to the current
Clip.
The Clip is specified by a Shape in user
space and is controlled by the program using the various clip
manipulation methods of Graphics and
Graphics2D. This user clip
is transformed into device space by the current
Transform and combined with the
device clip, which is defined by the visibility of windows and
device extents. The combination of the user clip and device clip
defines the composite clip, which determines the final clipping
region. The user clip is not modified by the rendering
system to reflect the resulting composite clip.
-
Determine what colors to render.
-
Apply the colors to the destination drawing surface using the current
Composite attribute in the Graphics2D context.
The three types of rendering operations, along with details of each
of their particular rendering processes are:
-
Shape operations
-
If the operation is a
draw(Shape) operation, then
the createStrokedShape
method on the current Stroke attribute in the
Graphics2D context is used to construct a new
Shape object that contains the outline of the specified
Shape.
-
The
Shape is transformed from user space to device space
using the current Transform
in the Graphics2D context.
-
The outline of the
Shape is extracted using the
getPathIterator method of
Shape, which returns a
PathIterator
object that iterates along the boundary of the Shape.
-
If the
Graphics2D object cannot handle the curved segments
that the PathIterator object returns then it can call the
alternate
getPathIterator
method of Shape, which flattens the Shape.
-
The current
Paint in the Graphics2D context
is queried for a PaintContext, which specifies the
colors to render in device space.
-
Text operations
-
The following steps are used to determine the set of glyphs required
to render the indicated
String:
-
If the argument is a
String, then the current
Font in the Graphics2D context is asked to
convert the Unicode characters in the String into a set of
glyphs for presentation with whatever basic layout and shaping
algorithms the font implements.
-
If the argument is an
AttributedCharacterIterator,
the iterator is asked to convert itself to a
TextLayout
using its embedded font attributes. The TextLayout
implements more sophisticated glyph layout algorithms that
perform Unicode bi-directional layout adjustments automatically
for multiple fonts of differing writing directions.
-
If the argument is a
GlyphVector, then the
GlyphVector object already contains the appropriate
font-specific glyph codes with explicit coordinates for the position of
each glyph.
-
The current
Font is queried to obtain outlines for the
indicated glyphs. These outlines are treated as shapes in user space
relative to the position of each glyph that was determined in step 1.
-
The character outlines are filled as indicated above
under
Shape operations.
-
The current
Paint is queried for a
PaintContext, which specifies
the colors to render in device space.
-
Image Operations
-
The region of interest is defined by the bounding box of the source
Image.
This bounding box is specified in Image Space, which is the
Image object's local coordinate system.
-
If an
AffineTransform is passed to
drawImage(Image, AffineTransform, ImageObserver),
the AffineTransform is used to transform the bounding
box from image space to user space. If no AffineTransform
is supplied, the bounding box is treated as if it is already in user space.
-
The bounding box of the source
Image is transformed from user
space into device space using the current Transform.
Note that the result of transforming the bounding box does not
necessarily result in a rectangular region in device space.
-
The
Image object determines what colors to render,
sampled according to the source to destination
coordinate mapping specified by the current Transform and the
optional image transform.
Default Rendering Attributes
The default values for the
Graphics2D rendering attributes are:
Paint
- The color of the
Component.
Font
- The
Font of the Component.
Stroke
- A square pen with a linewidth of 1, no dashing, miter segment joins
and square end caps.
Transform
- The
getDefaultTransform
for the GraphicsConfiguration of the Component.
Composite
- The
AlphaComposite.SRC_OVER rule.
Clip
- No rendering
Clip, the output is clipped to the
Component.
Rendering Compatibility Issues
The JDK(tm) 1.1 rendering model is based on a pixelization model
that specifies that coordinates
are infinitely thin, lying between the pixels. Drawing operations are
performed using a one-pixel wide pen that fills the
pixel below and to the right of the anchor point on the path.
The JDK 1.1 rendering model is consistent with the
capabilities of most of the existing class of platform
renderers that need to resolve integer coordinates to a
discrete pen that must fall completely on a specified number of pixels.
The Java 2D(tm) (Java(tm) 2 platform) API supports antialiasing renderers.
A pen with a width of one pixel does not need to fall
completely on pixel N as opposed to pixel N+1. The pen can fall
partially on both pixels. It is not necessary to choose a bias
direction for a wide pen since the blending that occurs along the
pen traversal edges makes the sub-pixel position of the pen
visible to the user. On the other hand, when antialiasing is
turned off by setting the
KEY_ANTIALIASING hint key
to the
VALUE_ANTIALIAS_OFF
hint value, the renderer might need
to apply a bias to determine which pixel to modify when the pen
is straddling a pixel boundary, such as when it is drawn
along an integer coordinate in device space. While the capabilities
of an antialiasing renderer make it no longer necessary for the
rendering model to specify a bias for the pen, it is desirable for the
antialiasing and non-antialiasing renderers to perform similarly for
the common cases of drawing one-pixel wide horizontal and vertical
lines on the screen. To ensure that turning on antialiasing by
setting the
KEY_ANTIALIASING hint
key to
VALUE_ANTIALIAS_ON
does not cause such lines to suddenly become twice as wide and half
as opaque, it is desirable to have the model specify a path for such
lines so that they completely cover a particular set of pixels to help
increase their crispness.
Java 2D API maintains compatibility with JDK 1.1 rendering
behavior, such that legacy operations and existing renderer
behavior is unchanged under Java 2D API. Legacy
methods that map onto general draw and
fill methods are defined, which clearly indicates
how Graphics2D extends Graphics based
on settings of Stroke and Transform
attributes and rendering hints. The definition
performs identically under default attribute settings.
For example, the default Stroke is a
BasicStroke with a width of 1 and no dashing and the
default Transform for screen drawing is an Identity transform.
The following two rules provide predictable rendering behavior whether
aliasing or antialiasing is being used.
- Device coordinates are defined to be between device pixels which
avoids any inconsistent results between aliased and antaliased
rendering. If coordinates were defined to be at a pixel's center, some
of the pixels covered by a shape, such as a rectangle, would only be
half covered.
With aliased rendering, the half covered pixels would either be
rendered inside the shape or outside the shape. With anti-aliased
rendering, the pixels on the entire edge of the shape would be half
covered. On the other hand, since coordinates are defined to be
between pixels, a shape like a rectangle would have no half covered
pixels, whether or not it is rendered using antialiasing.
- Lines and paths stroked using the
BasicStroke
object may be "normalized" to provide consistent rendering of the
outlines when positioned at various points on the drawable and
whether drawn with aliased or antialiased rendering. This
normalization process is controlled by the
KEY_STROKE_CONTROL hint.
The exact normalization algorithm is not specified, but the goals
of this normalization are to ensure that lines are rendered with
consistent visual appearance regardless of how they fall on the
pixel grid and to promote more solid horizontal and vertical
lines in antialiased mode so that they resemble their non-antialiased
counterparts more closely. A typical normalization step might
promote antialiased line endpoints to pixel centers to reduce the
amount of blending or adjust the subpixel positioning of
non-antialiased lines so that the floating point line widths
round to even or odd pixel counts with equal likelihood. This
process can move endpoints by up to half a pixel (usually towards
positive infinity along both axes) to promote these consistent
results.
The following definitions of general legacy methods
perform identically to previously specified behavior under default
attribute settings:
-
For
fill operations, including fillRect,
fillRoundRect, fillOval,
fillArc, fillPolygon, and
clearRect, fill can now be called
with the desired Shape. For example, when filling a
rectangle:
fill(new Rectangle(x, y, w, h));
is called.
-
Similarly, for draw operations, including
drawLine,
drawRect, drawRoundRect,
drawOval, drawArc, drawPolyline,
and drawPolygon, draw can now be
called with the desired Shape. For example, when drawing a
rectangle:
draw(new Rectangle(x, y, w, h));
is called.
-
The
draw3DRect and fill3DRect methods were
implemented in terms of the drawLine and
fillRect methods in the Graphics class which
would predicate their behavior upon the current Stroke
and Paint objects in a Graphics2D context.
This class overrides those implementations with versions that use
the current Color exclusively, overriding the current
Paint and which uses fillRect to describe
the exact same behavior as the preexisting methods regardless of the
setting of the current Stroke.
The
Graphics class defines only the
setColor
method to control the color to be painted. Since the Java 2D API extends
the
Color object to implement the new
Paint
interface, the existing
setColor method is now a convenience method for setting the
current
Paint attribute to a
Color object.
setColor(c) is equivalent to
setPaint(c).
The Graphics class defines two methods for controlling
how colors are applied to the destination.
-
The
setPaintMode method is implemented as a convenience
method to set the default Composite, equivalent to
setComposite(new AlphaComposite.SrcOver).
-
The
setXORMode(Color xorcolor) method is implemented
as a convenience method to set a special Composite object that
ignores the Alpha components of source colors and sets the
destination color to the value:
dstpixel = (PixelOf(srccolor) ^ PixelOf(xorcolor) ^ dstpixel);