- CGShading
[permalink]
CGShading lets you do color ramps. Setting one up is a little complicated, juggling a couple of CG data structures and writing the ramping function.
float domain[2] = { 0.0, 1.0 }; // 1-in function
float range[8] = { 0.0, 1.0, // N-out, RGBA
0.0, 1.0,
0.0, 1.0,
0.0, 1.0 };
CGFunctionCallbacks callbacks = { 0, evaluate, NULL };
CGFunctionRef shaderFunction;
shaderFunction = CGFunctionCreate (self, // info / rock / context
1, // # of inputs for domain
domain, // domain
4, // # of inputs for range
range, // range
&callbacks);
CGColorSpaceRef deviceRGB;
deviceRGB = CGColorSpaceCreateDeviceRGB ();
CGShadingRef shader;
shader = CGShadingCreateAxial (deviceRGB, // colorspace
cgpoint(start), // start of axis
cgpoint(end), // end of axis
shaderFunction, // shader, 1-n, n-out
NO, // extend start
NO); // extend end
CGContextSaveGState (context); {
NSRect bounds = [self bounds];
CGContextClipToRect (context, cgrect(bounds));
CGContextDrawShading (context, shader);
} CGContextRestoreGState (context);
[self drawSpotAt: start size: 4];
[self drawSpotAt: end size: 4];
CGFunctionRelease (shaderFunction);
CGColorSpaceRelease (deviceRGB);
CGShadingRelease (shader);
And the evaluator function is given an array of inputs and outputs. Use the in value(s) (which run from your domain's start-to-finish values) to generate the out values (which should be in the range's start-to-finish values):
static void evaluate (void *info, const float *in, float *out)
{
float thing;
thing = in[0];
out[0] = thing;
out[1] = thing;
out[2] = thing;
out[3] = 1.0;
} // evaluate - Draw a string centered in a rectangle
[permalink]
NSString *blah = @"Placeholder Pane";
NSSize size = [blah sizeWithAttributes: nil];
NSPoint startPoint;
startPoint.x = bounds.origin.x + bounds.size.width / 2 - size.width / 2;
startPoint.y = bounds.origin.y + bounds.size.height / 2 - size.height / 2;
[blah drawAtPoint: startPoint
withAttributes: nil];
- Getting a CGContext from Cocoaland
[permalink]
You can use CG (CoreGraphics) functions inside of a Cocoa view - pass the graphics port of the current context to the CG functions, like this:
#define cgrect(nsrect) (*(CGRect *)&(nsrect))
- (void) drawRect: (NSRect) rect
{
NSRect bounds = [self bounds];
NSGraphicsContext *cocoaContext = [NSGraphicsContext currentContext];
CGContextRef context = (CGContextRef)[cocoaContext graphicsPort];
CGContextSetLineWidth (context, 5.0);
CGContextBeginPath(context); {
CGContextAddRect (context, cgrect(bounds));
CGContextSetRGBFillColor (context, 1.0, 0.9, 0.8, 1.0);
} CGContextFillPath(context);
} // drawRect - Grabbing the bits from the screen
[permalink]
(Many thanks to Paul Sobolik for giving us this an updated and more reliable version)
- (NSImage *) captureScreenImageWithFrame: (NSRect) frame
{
// Fetch a graphics port of the screen
CGrafPtr screenPort = CreateNewPort ();
Rect screenRect;
GetPortBounds (screenPort, &screenRect);
// Make a temporary window as a receptacle
NSWindow *grabWindow = [[NSWindow alloc] initWithContentRect: frame
styleMask: NSBorderlessWindowMask
backing: NSBackingStoreRetained
defer: NO
screen: nil];
CGrafPtr windowPort = GetWindowPort ([grabWindow windowRef]);
Rect windowRect;
GetPortBounds (windowPort, &windowRect);
SetPort (windowPort);
// Copy the screen to the temporary window
CopyBits (GetPortBitMapForCopyBits(screenPort),
GetPortBitMapForCopyBits(windowPort),
&screenRect,
&windowRect,
srcCopy,
NULL);
// Get the contents of the temporary window into an NSImage
NSView *grabContentView = [grabWindow contentView];
[grabContentView lockFocus];
NSBitmapImageRep *screenRep;
screenRep = [[NSBitmapImageRep alloc] initWithFocusedViewRect: frame];
[grabContentView unlockFocus];
NSImage *screenImage = [[NSImage alloc] initWithSize: frame.size];
[screenImage addRepresentation: screenRep];
// Clean up
[grabWindow close];
DisposePort(screenPort);
return (screenImage);
} // captureScreenImageWithFrame - Using CGLayers
[permalink]
CGLayer is a Tiger (and beyond) feature that will cache Quartz operations. Kind of like a PICT for CG. Make a new layer based on the context you're going to be drawing it in, then get a context from the layer, and draw into that. So something like this:
- (void) makeLayerInContext: (CGContextRef) enclosingContext
{
layer = CGLayerCreateWithContext (enclosingContext, CGSizeMake(100, 100),
NULL); // options - unused in Tiger
CGContextRef context;
context = CGLayerGetContext (layer);
// .. and do your drawing
} // makeLayer
And then to draw the layer at a particular point (like replicating it a bunch of different points as done here):
for (i = 0; i < pointCount; i++) {
CGContextDrawLayerAtPoint (context, locations[i], layer);
}
