I'm failing to have a NSTextFieldCell created and drawn programmatically recreate the same visual appearance equivalently-configured NSTextFieldCells have when built in Interface Builder and drawn automatically by Cocoa.
Context is a custom NSView that displays content that user can select, delete or replace. Though this content is non-textual, I'd like the field to resemble an NSTextField, so I'm trying to press an NSTextFieldCell into service to draw a bezel, background fill, etc., that would match a corresponding NSTextField's.
I setup the cell as follows (during custom control init()):
textFieldCell = [[NSTextFieldCell alloc] init];
[textFieldCell setDrawsBackground: YES];
[textFieldCell setBackgroundColor: NSColor.blueColor];
[textFieldCell setBackgroundStyle: NSBackgroundStyleNormal];
[textFieldCell setBezeled: YES];
[textFieldCell setStringValue: #""];
and then in custom control's drawRect,
- (void) drawRect: (NSRect) rect
{
// Draw an NSTextField-like background
if( isSelectAndDeletable)
{
[textFieldCell drawWithFrame: [self bounds] inView: self];
}
... // more content drawing here
}
This produces on onscreen 1-pixel frame of the appropriate thickness and size, and fills it with blue. But the frame itself is gray, not blue, and the fill touches the frame on top (no white margin) and misses it by two pixels at the bottom (thick white margin). By contrast, when I let Cocoa draw my textFieldCells either as part of an NSTextField or as standalone cells created in Interface Builder, the frame is drawn WITH the cell's background color and the fill is inset with one pixel white margin all around.
This picture demonstrates the problem. It shows three NSTextFieldCells, the second of which is (defectively) drawn by my own call to drawWithFrame:inView:, the other two (correctly) directly by the runtime.
Any ideas what I'm doing wrong? I feel like I am passing the proper rect (the frame is in the correct location), and the single call produces both the correct stroke (frame) and incorrect fill (background). Is this some misconfigured bezel effect? I can hand draw the effect I'm after but would rather stick with NSTextFieldCell if I can fix my approach here.
Related
I've extended my window so that it has a button content border. The problem is that my NSSplitView covers the border. My thoughts were that I could make the split view transparent (but the controls on top of it opaque).
The following image shows what I am looking for, but with the missing splitter:
See how the bottom of the window has the button on it. This is the effect I am trying to achieve, but without any luck. The code that I am attempting to use is:
CALayer *transparentViewLayer = [CALayer layer];
[viewLayer setBackgroundColor:CGColorCreateGenericRGB(
256.0,
256.0,
256.0,
0)]; //RGB plus Alpha Channel
[splitView setWantsLayer:YES]; // view's backing store is using a Core Animation Layer
[splitView setLayer: transparentViewLayer];
If I do NOT set the transparentViewLayer, then I end up with the following:
Which shows the spitter, but hides the windows bottom bar. I have also tried subclassing NSSplitView and adding the following:
- (void) drawRect: (NSRect) dirtyRect
{
[[NSColor colorWithSRGBRed: 0.0 green: 255.0 blue: 0.0 alpha: 155.0] setFill];
NSRectFill(dirtyRect);
} // End of drawRect
Which left me with a green splitter and no alpha for the windows bottom bar. Any ideas on how I can achieve the effect I am looking for? (I want the bottom bar with my buttons and the splitter visible).
In your subclass, don't override -[NSSplitView drawRect:]. Instead, override -drawDividerInRect: and draw your divider ONLY in that rect. Always consult the documentation first when you want to "bend" a class to your will.
I have a NSView, subclassed, with custom background drawing, filling it with a gradient.
In IB, I've put a checkbox on it, somewhere in the middle.
This is the drawRect method.
-(void) drawRect:(NSRect)dirtyRect {
CGFloat sc = 0.9f;
CGFloat ec = 0.6f;
NSColor* startingColor = [NSColor colorWithDeviceRed:sc green:sc blue:sc alpha:1];
NSColor* endingColor = [NSColor colorWithDeviceRed:ec green:ec blue:ec alpha:1];
NSGradient *grad = [[NSGradient alloc] initWithStartingColor:startingColor endingColor:endingColor];
[grad drawInRect:dirtyRect angle:270];
}
What happens is, this same method gets called to draw the whole view area first and then for the part, where NSButton (checkbox) lies on top of it. OF course the checkbox background is drawn with a complete gradient and it is not right, since the portion is much smaller. The same happens with other controls I put on the said NSView.
What is the suggested approach on such thing?
One option is to make controls height the same as the views' but this will result in problems in the future.
The answer is, always draw the WHOLE area of the view, not just the dirtyRect
[grad drawInRect:[self bounds] angle:270];
Hi in my application i have a custom view with CALayer (i use it for transition animation), in that view i add some subviews, one of them is a view with NSTableView. Each cell of NSTableView, draws it's content with shadow. Code like this:
NSShadow* shadow = [[NSShadow alloc] init];
[shadow setShadowOffset:NSMakeSize(2.0, -2.0)];
[shadow setShadowBlurRadius:2.0];
[shadow setShadowColor:[[NSColor blackColor] colorWithAlphaComponent:0.3]];
[shadow set];
[self.stringValue drawInRect:NSMakeRect(x, y, w - 60, h) withAttributes:attributes];
When my NSTableView has less than ~50 rows with CALayer turned on OR when i turn off CALayer in IB, i have cells drawn like that: http://i.stack.imgur.com/zMxmK.png
But when my NSTableView has more than ~50 row AND i turn on CALayer in IB, i have cells drawn like that: http://i.stack.imgur.com/IMyIP.png
How can it be? I break my had :(
Try changing -2.0 to +2.0 depending on [[NSGraphicsContext currentContext] isFlipped] property.
Try changing it depending on whether the view is layered and CALayer has isGeometryFlipped on.
If you have flipping transformation(s) applied, note they won't affect NSShadow:
Shadows are always drawn in the default user coordinate space, regardless of any transformations applied to that space. This means that rotations, translations and other transformations of the current transformation matrix (the CTM) do not affect the resulting shadow. Another way to think about this is that changes to the CTM do not move or change the apparent position of the shadow’s light source.
I've got a relatively simple Cocoa on Mac OS X 10.6 question to ask. I have a main NSView (ScreensaverView, actually) that is not layer backed and is otherwise unremarkable. It does some basic drawing in its drawRect via NSRectFill and NSBezierPath:stroke calls (dots and lines, basically).
I've also got a NSView-derived subclass that is acting as a child subview. I'm doing this with the goal to draw simple lines in the subview that draw on top of whatever is drawn in the main view, but then can be somehow "erased" revealing whatever the lines obscured. The code for this subview is quite simple:
- (id)initWithFrame:(NSRect)frame {
self = [super initWithFrame:frame];
if (self) {
// Initialization code here.
}
return self;
}
- (void)dealloc
{
[super dealloc];
}
- (void)drawRect:(NSRect)dirtyRect {
// Transparent background
[[NSColor clearColor] set];
NSRectFillUsingOperation(dirtyRect, NSCompositeCopy);
// If needed for this update, draw line
if (drawLine) {
// OMITTED: Code that sets opaque NSColor and draws a line using NSBezierPath:stroke
}
// If needed for this update, "erase" line
if (eraseLine) {
[[NSColor blackColor] set]; // clearColor?
// OMITTED: Code that draws the same line using NSBezierPath:stroke
}
}
With the code as shown above, any time the subview draws, the main view goes black and you only see the subview line. When the subview isn't being updated, the main view contents appear again.
Things I've tried, with varying results:
I tried experimenting with making the subview return YES from an overridden isOpaque (which I realize isn't really correct). When I do this, both views draw properly during a subview update, however the subview line overwrites anything it is drawn on (ok) and then when erased, also leaves a large black line where it was (what I was trying to avoid). Trying to "erase" the line using clearColor instead of blackColor results in the line remaining on screen.
I tried making both (and/or just the subview) layer-backed views via calling [self setWantsLayer:YES] in init, and this results in a completely black screen.
I feel like I'm missing something really basic, but for whatever reason, I can't seem to figure it out. Any and all suggestions are greatly appreciated.
I think you are fundamentally misunderstanding how view drawing works. Basically, every time drawRect: is called on your view, the drawing commands in drawRect: are executed. This might happen automatically or when you issue the view a setNeedsDisplay: message.
Normally, when drawRect: is called on the view, the view is erased and drawing begins anew. Nothing remains in the view from the previous call to drawRect:. You do not need to fill the view with [NSColor clearColor] in order for it to be transparent.
Note that this is only the case if your view returns NO from isOpaque, which is the default. If your view is returning YES from isOpaque then you will need to ensure you erase the view before drawing, however in your particular case you should not return YES from isOpaque because you want your view to be transparent.
You do not need to "erase" the line you've drawn. It will be erased for you. Instead, you need to simply not draw it when you don't want it drawn.
Basically, your view should store some flag (such as your BOOL ivar named drawLine or something similar). Then, in your drawing code you should simply check if this value is set. If it is, you should draw the line. If not, then just do nothing. Your code should be reduced to this:
- (void)drawRect:(NSRect)rect
{
// If needed for this update, draw line
if (drawLine) {
// OMITTED: Code that sets opaque NSColor and draws a line using NSBezierPath:stroke
}
}
If you want to change the state and redraw the view, all you need to do is change the value of the drawLine ivar and ask the view to redraw using [yourView setNeedsDisplay:YES].
You can easily do this with a timer:
- (void)awakeFromNib
{
[NSTimer scheduledTimerWithTimeInterval:1.0 target:self selector:#selector(update:) userInfo:nil repeats:YES];
}
- (void)update:(NSTimer*)timer
{
drawLine = !drawLine;
[self setNeedsDisplay:YES];
}
I am doing manual layouting for my Cocoa application and at some point I need to figure out what the inner size of a NSView subclass is. (E.g. What is the height available for my child view inside of a NSBox?)
One of the reasons is that I am using a coordinate system with origin at the top-left and need to perform coordinate transformations.
I could not figure out a way to get this size so far and would be glad if somebody can give me a hint.
Another very interesting property I would like to know is the minimum size of a view.
-bounds is the one you're looking for in most views. NSBox is a bit of a special case, however, since you want to look at the bounds of the box's content view, not the bounds of the box view itself (the box view includes the title, edges, etc.). Also, the bounds rect is always the real size of the box, while the frame rect can be modified relative to the bounds to apply transformations to the view's contents (such as squashing a 200x200 image into a 200x100 frame).
So, for most views you just use [parentView bounds], and for NSBox you'll use [[theBox contentView] bounds], and you'll use [[theBox contentView] addSubview: myView] rather than [parentView addSubview: myView] to add your content.
Unfortunately, there is no standard way to do this for all NSView subclasses. In your specific example, the position and size of a child view within an NSBox can be computed as follows:
NSRect availableRect = [someNSBox bounds];
NSSize boxMargins = [someBox contentViewMargins];
availableRect = NSInsetRect(availableRect, boxMargins.width, boxMargins.height);
If you find yourself using this often, you could create a category on NSBox as follows:
// MyNSBoxCategories.h
#interface NSBox (MyCategories)
- (NSRect)contentFrame;
#end
// MyNSBoxCategories.m
#implementation NSBox (MyCategories)
- (NSRect)contentFrame
{
NSRect frameRect = [self bounds];
NSSize margins = [self contentViewMargins];
return NSInsetRect(frameRect, margins.width, margins.height);
}
#end
And you would use it like so:
#import "MyNSBoxCategories.h"
//...
NSRect frameRect = [someNSBox contentFrame];
[myContentView setFrame:frameRect];
[someNSBox addSubview:myContentView];
The bounds property of NSView returns an NSRect with the origin (usually (0,0)) and the size of an NSView. See this Apple Developer documentation page.
I'm not sure (I never had to go too deep in that stuff), but isn't it [NSView bounds]?
http://www.cocoadev.com/index.pl?DifferenceBetweenFrameAndBounds