The Page

Quartz 2D uses the painter’s model for its imaging. In the painter’s model, each successive drawing operation applies a layer of “paint” to an output “canvas,” often called a page. The paint on the page can be modified by overlaying more paint through additional drawing operations. An object drawn on the page cannot be modified except by overlaying more paint. This model allows you to construct extremely sophisticated images from a small number of powerful primitives.

Figure 1-1 shows how the painter’s model works. To get the image in the top part of the figure, the shape on the left was drawn first followed by the solid shape. The solid shape overlays the first shape, obscuring all but the perimeter of the first shape. The shapes are drawn in the opposite order in the bottom of the figure, with the solid shape drawn first. As you can see, in the painter’s model the drawing order is important.

The page may be a real sheet of paper (if the output device is a printer); it may be a virtual sheet of paper (if the output device is a PDF file); it may even be a bitmap image. The exact nature of the page depends on the particular graphics context you use.

Drawing Destinations: The Graphics Context

A graphics context is an opaque data type (CGContextRef) that encapsulates the information Quartz uses to draw images to an output device, such as a PDF file, a bitmap, or a window on a display. The information inside a graphics context includes graphics drawing parameters and a device-specific representation of the paint on the page. All objects in Quartz are drawn to, or contained by, a graphics context.

You can think of a graphics context as a drawing destination, as shown in Figure 1-2. When you draw with Quartz, all device-specific characteristics are contained within the specific type of graphics context you use. In other words, you can draw the same image to a different device simply by providing a different graphics context to the same sequence of Quartz drawing routines. You do not need to perform any device-specific calculations; Quartz does it for you.

These graphics contexts are available to your application:

• A bitmap graphics context allows you to paint RGB colors, CMYK colors, or grayscale into a bitmap. A bitmap is a rectangular array (or raster) of pixels, each pixel representing a point in an image. Bitmap images are also called sampled images. See Creating a Bitmap Graphics Context.

• A PDF graphics context allows you to create a PDF file. In a PDF file, your drawing is preserved as a sequence of commands. There are some significant differences between PDF files and bitmaps:

  • PDF files, unlike bitmaps, may contain more than one page.

  • When you draw a page from a PDF file on a different device, the resulting image is optimized for the display characteristics of that device.

  • PDF files are resolution independent by nature—the size at which they are drawn can be increased or decreased infinitely without sacrificing image detail. The user-perceived quality of a bitmap image is tied to the resolution at which the bitmap is intended to be viewed.

  See Creating a PDF Graphics Context.

• A window graphics context is a graphics context that you can use to draw into a window. Note that because Quartz 2D is a graphics engine and not a window management system, you use one of the application frameworks to obtain a graphics context for a window. See Creating a Window Graphics Context in Mac OS X for details.

• A layer context (CGLayerRef) is an offscreen drawing destination associated with another graphics context. It is designed for optimal performance when drawing the layer to the graphics context that created it. A layer context can be a much better choice for offscreen drawing than a bitmap graphics context. See Core Graphics Layer Drawing.

• When you want to print in Mac OS X, you send your content to a PostScript graphics context that is managed by the printing framework. See Obtaining a Graphics Context for Printing for more information.

Quartz 2D Opaque Data Types

The Quartz 2D API defines a variety of opaque data types in addition to graphics contexts. Because the API is part of the Core Graphics framework, the data types and the routines that operate on them use the CG prefix.

Quartz 2D creates objects from opaque data types that your application operates on to achieve a particular drawing output. Figure 1-3 shows the sorts of results you can achieve when you apply drawing operations to three of the objects provided by Quartz 2D. For example:

• You can rotate and display a PDF page by creating a PDF page object, applying a rotation operation to the graphics context, and asking Quartz 2D to draw the page to a graphics context.

• You can draw a pattern by creating a pattern object, defining the shape that makes up the pattern, and setting up Quartz 2D to use the pattern as paint when it draws to a graphics context.

• You can fill an area with an axial or radial shading by creating a shading object, providing a function that determines the color at each point in the shading, and then asking Quartz 2D to use the shading as a fill color.

The opaque data types available in Quartz 2D include the following:

• CGPathRef, used for vector graphics to create paths that you fill or stroke. See Paths.

• CGImageRef, used to represent bitmap images and bitmap image masks based on sample data that you supply. See Bitmap Images and Image Masks.

• CGLayerRef, used to represent a drawing layer that can be used for repeated drawing (such as for backgrounds or patterns) and for offscreen drawing. See Core Graphics Layer Drawing

• CGPatternRef, used for repeated drawing. See Patterns.

• CGShadingRef and CGGradientRef, used to paint gradients. See Gradients.

• CGFunctionRef, used to define callback functions that take an arbitrary number of floating-point arguments. You use this data type when you create gradients for a shading. See Gradients.

• CGColorRef and CGColorSpaceRef, used to inform Quartz how to interpret color. See Color and Color Spaces.

• CGImageSourceRef and CGImageDestinationRef, which you use to move data into and out of Quartz. See Data Management in Quartz 2D and Image I/O Programming Guide.

• CGFontRef, used to draw text. See Text.

• CGPDFDictionaryRef, CGPDFObjectRef, CGPDFPageRef, CGPDFStream, CGPDFStringRef, and CGPDFArrayRef, which provide access to PDF metadata. See PDF Document Creation, Viewing, and Transforming.

• CGPDFScannerRef and CGPDFContentStreamRef, which parse PDF metadata. See PDF Document Parsing.

• CGPSConverterRef, used to convert PostScript to PDF. It is not available in iOS. See PostScript Conversion.

Graphics States

Quartz modifies the results of drawing operations according to the parameters in the current graphics state. The graphics state contains parameters that would otherwise be taken as arguments to drawing routines. Routines that draw to a graphics context consult the graphics state to determine how to render their results. For example, when you call a function to set the fill color, you are modifying a value stored in the current graphics state. Other commonly used elements of the current graphics state include the line width, the current position, and the text font size.

The graphics context contains a stack of graphics states. When Quartz creates a graphics context, the stack is empty. When you save the graphics state, Quartz pushes a copy of the current graphics state onto the stack. When you restore the graphics state, Quartz pops the graphics state off the top of the stack. The popped state becomes the current graphics state.

To save the current graphics state, use the function CGContextSaveGState to push a copy of the current graphics state onto the stack. To restore a previously saved graphics state, use the function CGContextRestoreGState to replace the current graphics state with the graphics state that’s on top of the stack.

Note that not all aspects of the current drawing environment are elements of the graphics state. For example, the current path is not considered part of the graphics state and is therefore not saved when you call the function CGContextSaveGState. The graphics state parameters that are saved when you call this function are listed in Table 1-1.

Quartz 2D Coordinate Systems

A coordinate system, shown in Figure 1-4, defines the range of locations used to express the location and sizes of objects to be drawn on the page. You specify the location and size of graphics in the user-space coordinate system, or, more simply, the user space. Coordinates are defined as floating-point values.

Because different devices have different underlying imaging capabilities, the locations and sizes of graphics must be defined in a device-independent manner. For example, a screen display device might be capable of displaying no more than 96 pixels per inch, while a printer might be capable of displaying 300 pixels per inch. If you define the coordinate system at the device level (in this example, either 96 pixels or 300 pixels), objects drawn in that space cannot be reproduced on other devices without visible distortion. They will appear too large or too small.

Quartz accomplishes device independence with a separate coordinate system—user space—mapping it to the coordinate system of the output device—device space—using the current transformation matrix, or CTM. A matrix is a mathematical construct used to efficiently describe a set of related equations. The current transformation matrix is a particular type of matrix called an affine transform, which maps points from one coordinate space to another by applying translation, rotation, and scaling operations (calculations that move, rotate, and resize a coordinate system).

The current transformation matrix has a secondary purpose: It allows you to transform how objects are drawn. For example, to draw a box rotated by 45 degrees, you rotate the coordinate system of the page (the CTM) before you draw the box. Quartz draws to the output device using the rotated coordinate system.

A point in user space is represented by a coordinate pair (x,y), where x represents the location along the horizontal axis (left and right) and y represents the vertical axis (up and down). The origin of the user coordinate space is the point (0,0). The origin is located at the lower-left corner of the page, as shown in Figure 1-4. In the default coordinate system for Quartz, the x-axis increases as it moves from the left toward the right of the page. The y-axis increases in value as it moves from the bottom toward the top of the page.

Some technologies set up their graphics contexts using a different default coordinate system than the one used by Quartz. Relative to Quartz, such a coordinate system is a modified coordinate system and must be compensated for when performing some Quartz drawing operations. The most common modified coordinate system places the origin in the upper-left corner of the context and changes the y-axis to point towards the bottom of the page. A few places where you might see this specific coordinate system used are the following:

• In Mac OS X, a subclass of NSView that overrides its isFlipped method to return YES.

• In iOS, a drawing context returned by an UIView.

• In iOS, a drawing context created by calling the UIGraphicsBeginImageContextWithOptions function.

The reason UIKit returns Quartz drawing contexts with modified coordinate systems is that UIKit uses a different default coordinate convention; it applies the transform to Quartz contexts it creates so that they match its conventions. If your application wants to use the same drawing routines to draw to both a UIView object and a PDF graphics context (which is created by Quartz and uses the default coordinate system), you need to apply a transform so that the PDF graphics context receives the same modified coordinate system. To do this, apply a transform that translates the origin to the upper-left corner of the PDF context and scales the y-coordinate by -1.

Using a scaling transform to negate the y-coordinate alters some conventions in Quartz drawing. For example, if you call CGContextDrawImage to draw an image into the context, the image is modified by the transform when it is drawn into the destination. Similarly, path drawing routines accept parameters that specify whether an arc is drawn in a clockwise or counterclockwise direction in the default coordinate system. If a coordinate system is modified, the result is also modified, as if the image were reflected in a mirror. In Figure 1-5, passing the same parameters into Quartz results in a clockwise arc in the default coordinate system and a counterclockwise arc after the y-coordinate is negated by the transform.

It is up to your application to adjust any Quartz calls it makes to a context that has a transform applied to it. For example, if you want an image or PDF to draw correctly into a graphics context, your application may need to temporarily adjust the CTM of the graphics context. In iOS, if you use a UIImage object to wrap a CGImage object you create, you do not need to modify the CTM. The UIImage object automatically compensates for the modified coordinate system applied by UIKit.

Memory Management: Object Ownership

Quartz uses the Core Foundation memory management model, in which objects are reference counted. When created, Core Foundation objects start out with a reference count of 1. You can increment the reference count by calling a function to retain the object, and decrement the reference count by calling a function to release the object. When the reference count is decremented to 0, the object is freed. This model allows objects to safely share references to other objects.

There are a few simple rules to keep in mind:

• If you create or copy an object, you own it, and therefore you must release it. That is, in general, if you obtain an object from a function with the words “Create” or “Copy” in its name, you must release the object when you’re done with it. Otherwise, a memory leak results.

• If you obtain an object from a function that does not contain the words “Create” or “Copy” in its name, you do not own a reference to the object, and you must not release it. The object will be released by its owner at some point in the future.

• If you do not own an object and you need to keep it around, you must retain it and release it when you’re done with it. You use the Quartz 2D functions specific to an object to retain and release that object. For example, if you receive a reference to a CGColorspace object, you use the functions CGColorSpaceRetain and CGColorSpaceRelease to retain and release the object as needed. You can also use the Core Foundation functions CFRetain and CFRelease, but you must be careful not to pass NULL to these functions.