PlatformPixelBuffer(rfb::PixelFormat(32, 24, false, true,
255, 255, 255, 16, 8, 0),
width, height, NULL, width),
- bitmap(NULL)
+ image(NULL)
{
CGColorSpaceRef lut;
+ CGDataProviderRef provider;
data = new rdr::U8[width * height * format.bpp/8];
if (data == NULL)
throw rfb::Exception(_("Not enough memory for framebuffer"));
- lut = CGColorSpaceCreateDeviceRGB();
- if (!lut)
- throw rfb::Exception("CGColorSpaceCreateDeviceRGB");
-
- bitmap = CGBitmapContextCreate(data, width, height, 8, width*4, lut,
- kCGImageAlphaNoneSkipFirst | kCGBitmapByteOrder32Little);
+ lut = CGDisplayCopyColorSpace(kCGDirectMainDisplay);
+ if (!lut) {
+ vlog.error(_("Could not get screen color space. Using slower fallback."));
+ lut = CGColorSpaceCreateDeviceRGB();
+ if (!lut)
+ throw rfb::Exception("CGColorSpaceCreateDeviceRGB");
+ }
+
+ provider = CGDataProviderCreateWithData(NULL, data,
+ width * height * 4, NULL);
+ if (!provider)
+ throw rfb::Exception("CGDataProviderCreateWithData");
+
+ image = CGImageCreate(width, height, 8, 32, width * 4, lut,
+ kCGImageAlphaNoneSkipFirst | kCGBitmapByteOrder32Little,
+ provider, NULL, false, kCGRenderingIntentDefault);
CGColorSpaceRelease(lut);
- if (!bitmap)
- throw rfb::Exception("CGBitmapContextCreate");
+ CGDataProviderRelease(provider);
+ if (!image)
+ throw rfb::Exception("CGImageCreate");
}
OSXPixelBuffer::~OSXPixelBuffer()
{
- CFRelease((CGContextRef)bitmap);
+ CGImageRelease(image);
delete [] data;
}
void OSXPixelBuffer::draw(int src_x, int src_y, int x, int y, int w, int h)
{
- CGRect rect;
CGContextRef gc;
- CGAffineTransform at;
- CGImageRef image;
+ CGRect rect;
- gc = (CGContextRef)fl_gc;
+ gc = fl_gc;
CGContextSaveGState(gc);
- // We have to use clipping to partially display an image
+ // macOS Coordinates are from bottom left, not top left
+ src_y = height() - (src_y + h);
+ y = Fl_Window::current()->h() - (y + h);
+
+ // Reset the transformation matrix back to the default identity
+ // matrix as otherwise we get a massive performance hit
+ CGContextConcatCTM(gc, CGAffineTransformInvert(CGContextGetCTM(gc)));
- rect.origin.x = x - 0.5;
- rect.origin.y = y - 0.5;
+ // We have to use clipping to partially display an image
+ rect.origin.x = x;
+ rect.origin.y = y;
rect.size.width = w;
rect.size.height = h;
CGContextClipToRect(gc, rect);
-
- // Oh the hackiness that is OS X image handling...
- // The CGContextDrawImage() routine magically flips the images and offsets
- // them by 0.5,0.5 in order to compensate for OS X unfamiliar coordinate
- // system. But this breaks horribly when you set up the CTM to get the
- // more familiar top-down system (which FLTK does), meaning we have to
- // reset the CTM back to the identity matrix and calculate new origin
- // coordinates.
-
- at = CGContextGetCTM(gc);
- CGContextScaleCTM(gc, 1, -1);
- CGContextTranslateCTM(gc, -at.tx, -at.ty);
rect.origin.x = x - src_x;
- rect.origin.y = Fl_Window::current()->h() - (y - src_y);
+ rect.origin.y = y - src_y;
rect.size.width = width();
- rect.size.height = -height(); // Negative height does _not_ flip the image
+ rect.size.height = height();
- image = CGBitmapContextCreateImage((CGContextRef)bitmap);
CGContextDrawImage(gc, rect, image);
- CGImageRelease(image);
CGContextRestoreGState(gc);
}