I'm trying to use
NSBitmapImageRep.bitmapImageRepByConvertingToColorSpace(NSColorSpace.genericRGBColorSpace(), renderingIntent: NSColorRenderingIntent.Perceptual);
to convert an NSImage to a format that can be handled by openGL, and it works (unlike bitmapImageRepByRetaggingWithColorSpace()), but I get an error:
<Error>: The function ‘CGContextClear’ is obsolete and will be removed in an upcoming update. Unfortunately, this application, or a library it uses, is using this obsolete function, and is thereby contributing to an overall degradation of system performance.
which is sort of useless, because it's Apple's own code, and I can't seem to find an alternative to bitmapImageRepByConvertingToColorSpace(), or any note that it too is deprecated.
EDIT:
var image=NSImage(size: frame);
image.lockFocus();
//println("NSImage: \(image.representations)");
string.drawAtPoint(NSMakePoint(0.0,0.0), withAttributes: attribs);
var bitmap2=NSBitmapImageRep(focusedViewRect: NSMakeRect(0.0,0.0,frame.width,frame.height))?;
image.unlockFocus();
bitmap=bitmap2!.bitmapImageRepByConvertingToColorSpace(NSColorSpace.genericRGBColorSpace(), renderingIntent: NSColorRenderingIntent.Perceptual);
For some reason this question interested me. It looks like the solution is to lock focus on your NSImage, then use NSBitmapImageRep(focusedViewRect:) to create the NSBitmapImageRep.
I tried to recreate your situation (as I understand it) and then create an NSBitmapImageRep with the following code:
var img:NSImage = NSImage(size: NSMakeSize(200,200))
img.lockFocus()
let testString:NSString = "test String"
testString.drawAtPoint(NSMakePoint(10,10), withAttributes: nil)
img.unlockFocus()
println("img Description = \(img.description)")
I got the following description:
img Description = NSImage 0x608000067ac0 Size={200, 200} Reps=(
"NSCGImageSnapshotRep:0x61000007cf40 cgImage=CGImage 0x6100001a2bc0")
I then extended this code to:
var img:NSImage = NSImage(size: NSMakeSize(200,200))
img.lockFocus()
let testString:NSString = "test String"
testString.drawAtPoint(NSMakePoint(10,10), withAttributes: nil)
img.unlockFocus()
println("img Description = \(img.description)")
img.lockFocus()
var bitmapRep:NSBitmapImageRep = NSBitmapImageRep(focusedViewRect: NSMakeRect(0.0, 0.0, img.size.width, img.size.height))!
img.unlockFocus()
println("bitmap data planes = \(bitmapRep.bitmapData)")
println("bitmap pixels wide = \(bitmapRep.size.width)")
println("bitmap pixels high = \(bitmapRep.size.height)")
println("bits per sample = \(bitmapRep.bitsPerSample)")
println("samples per pixel = \(bitmapRep.samplesPerPixel)")
println("has alpha = \(bitmapRep.alpha)")
println("is planar = \(bitmapRep.planar)")
println("color space name = \(bitmapRep.colorSpace)")
println("bitmap format = \(bitmapRep.bitmapFormat)")
println("bytes per row = \(bitmapRep.bytesPerRow)")
println("bits per pixel = \(bitmapRep.bitsPerPixel)")
The output for the NSBitmapImageRep parameters was:
bitmap data planes = 0x000000010b6ff100
bitmap pixels wide = 200.0
bitmap pixels high = 200.0
bits per sample = 8
samples per pixel = 4
has alpha = true
is planar = false
color space name = Color LCD colorspace
bitmap format = C.NSBitmapFormat
bytes per row = 800
bits per pixel = 32
I've posted some new code below. I've created two NSBitMapImageRep just like your code (except that I used the method bitmapImageRepByRetaggingWithColorSpace), and I exported both of them to a PNG file. I got the same image in both PNG files.
But I'm wondering a couple of things. First, the difference in your two tests was not only the OS X version, but the hardware produced different color spaces. You should check to be sure bitmap2 is not nil.
Second, I'm wondering if OpenGL cares. If bitmap2 is not nil and you pass OpenGL bitmap2.bitmapData, does it work?
My new code (deleting most of the println):
var img:NSImage = NSImage(size: NSMakeSize(200,200))
img.lockFocus()
let testString:NSString = "test String"
let font:NSFont = NSFont(name: "AppleCasual", size: 18.0)!
let textStyle = NSMutableParagraphStyle.defaultParagraphStyle().mutableCopy() as NSMutableParagraphStyle
textStyle.alignment = NSTextAlignment.LeftTextAlignment
let textColor:NSColor = NSColor(calibratedRed: 1.0, green: 0.0, blue: 1.0, alpha: 1.0)
let attribs:NSDictionary = [NSFontAttributeName: font,
NSForegroundColorAttributeName: textColor,
NSParagraphStyleAttributeName: textStyle]
testString.drawAtPoint(NSMakePoint(0.0, 0.0), withAttributes: attribs)
img.unlockFocus()
println("img Description = \(img.description)")
img.lockFocus()
var bitmapRep:NSBitmapImageRep = NSBitmapImageRep(focusedViewRect: NSMakeRect(0.0, 0.0, img.size.width, img.size.height))!
img.unlockFocus()
let pngPath:String = "TestStringBeforeChange.png"
let imageProps = [NSImageCompressionFactor: NSNumber(float: 1.0)]
let outputImageData = bitmapRep.representationUsingType(NSBitmapImageFileType.NSPNGFileType, properties: imageProps)
let fileSavePanel = NSSavePanel()
fileSavePanel.nameFieldStringValue = pngPath
var savePanelReturn = fileSavePanel.runModal()
if savePanelReturn == NSFileHandlingPanelOKButton
{
var theFileURL = fileSavePanel.URL
var saveStatus = outputImageData?.writeToURL(theFileURL!, atomically: true)
}
let bitmapRep2 = bitmapRep.bitmapImageRepByRetaggingWithColorSpace(NSColorSpace.genericRGBColorSpace())
let pngPath2:String = "TestStringAfterChange.png"
let outputImageData2 = bitmapRep2!.representationUsingType(NSBitmapImageFileType.NSPNGFileType, properties: imageProps)
let fileSavePanel2 = NSSavePanel()
fileSavePanel2.nameFieldStringValue = pngPath2
var savePanel2Return = fileSavePanel2.runModal()
if savePanel2Return == NSFileHandlingPanelOKButton
{
var theFileURL2 = fileSavePanel2.URL
var saveStatus2 = outputImageData2?.writeToURL(theFileURL2!, atomically: true)
}
Sorry for all the posts, but I don't want to let this go, and it's interesting.
I created 3 PNG files: the first using the original NSBitmapImageRep, a second using an NSBitmapImageRep created with bitmapImageRepByRetaggingWithColorSpace and a third using an NSBitmapImageRep created with bitmapImageRepByConvertingToColorSpace (I, of course, got the system warning when I used this function).
Looking at these three PNG files in Preview's Inspector, they were all RGB color models; only the colorSynch profile for the first file was different from that of the second and third file. So I thought maybe there's no difference in the bitmapData among these files.
I then wrote some code to compare the bitmapData:
var firstRepPointer:UnsafeMutablePointer<UInt8> = bitmapRep.bitmapData
var secondRepPointer:UnsafeMutablePointer<UInt8> = bitmapRep2!.bitmapData
var firstByte:UInt8 = 0
var secondByte:UInt8 = 0
for var i:Int = 0; i < 200; i++
{
for var j:Int = 0; j < 200; j++
{
for var k:Int = 0; k < 4; k++
{
memcpy(&firstByte, firstRepPointer, 1)
firstRepPointer += 1
memcpy(&secondByte, secondRepPointer, 1)
secondRepPointer += 1
if firstByte != secondByte {println("firstByte = \(firstByte) secondByte = \(secondByte)")}
}
}
}
As I expected, there were no differences when I compared the bitmapData from the original bitmap rep to the bitmapData from the bitmap rep created with bitmapImageRepByRetaggingWithColorSpace.
It got more interesting, however when I compared the bitmapData from the original bitmap rep to the data from the bitmap rep created with bitmapImageRepByConvertingToColorSpace.
There were lots of differences in the data. Small differences, but lots of them.
Even though there were no differences I could see in the PNG files, the underlying data had been changed by bitmapImageRepByConvertingToColorSpace.
All that said, I think you should be able to pass .bitmapData from any of these bitmap reps to an OpenGL texture. The data from the bitmap rep created with bitmapImageRepByConvertingToColorSpace might create a texture that looks different, but it should be valid data.
Related
The contents to the text layer are added from csv import. Some are short length and some are long, contain 2 words and take up 2 lines in the layer. What I need is after the content is added, the layer should be horizontally and vertically aligned to another layer. I want to do this alignment using a script.
var doc = app.activeDocument;
var grps = doc.layerSets;
var pnamegrp = grps.getByName('Group 1');
var childlyr = pnamegrp.layers.getByName('child');
childlyr.textItem.contents = pname; //come from a csv file
var parentlyr= pnamegrp.layers.getByName('ReferenceRectangle');
Align_HorizCenter_VerticalCenter_withreference( childlyr , parent);
function Align_HorizCenter_VerticalCenter_withreference( child, parent){
//need help to write this
}
I am using Photoshop cc 2015 and JavaScript jsx file scripting.
Just incase somebody is looking for a solution. Translate is the method to move layer. The number of pixels to be moved can be determined by the difference in the width between the target and reference layer.
var startRulerUnits = app.preferences.rulerUnits;
app.preferences.rulerUnits = Units.INCHES;
var doc = app.activeDocument;
var grps = doc.layerSets;
var pnamegrp = grps.getByName('Group 1');
var pnamelyr= pnamegrp.layers.getByName('pname'); //target
var pnameREF = pnamegrp.layers.getByName('Rectangle 1'); //reference var LB = pnameREF.bounds;
var RWidth = (LB[2].value) - (LB[0].value);
var RHeight = (LB[3].value) - (LB[1].value);
pnamelyr.textItem.contents = pnamearr[i];
LB = pnamelyr.bounds;
TWidth = (LB[2].value) - (LB[0].value);
THeight = (LB[3].value) - (LB[1].value);
var OffsetX = (RWidth - TWidth)/2;
var OffsetY = (RHeight - THeight)/2;
pnameTGT.translate(OffsetX,OffsetY); //move layer by offset pixels from the current position
I'm having an issue where I am getting a list of skills back from an api and I want them to stack one on top of the other in two different sections, a left column and a right column. It works well but if the skill is longer than the width of the label it drops to a new line with the same spacing as the rest of the labels. The skill Adobe Creative Suite looks like Adobe Creative as one and Suite as another. I would like Suite to be underneath Adobe Creative but much closer so you can tell it's all one skill.
My code is here:
lblLeft.text = ""
lblRight.text = ""
if let expertiseCount = helper.expertise {
for i in 0..<expertiseCount.count {
if i % 2 == 0 {
let paragraphStyle = NSMutableParagraphStyle()
paragraphStyle.lineSpacing = 10
let attrString = NSMutableAttributedString(string: lblLeft.text! + "\(expertiseCount[i].name ?? "")\n")
attrString.addAttribute(NSParagraphStyleAttributeName, value:paragraphStyle, range: NSMakeRange(0, attrString.length))
lblLeft.attributedText = attrString
} else {
let paragraphStyle = NSMutableParagraphStyle()
paragraphStyle.lineSpacing = 10
let attrString = NSMutableAttributedString(string: lblRight.text! + "\(expertiseCount[i].name ?? "")\n")
attrString.addAttribute(NSParagraphStyleAttributeName, value:paragraphStyle, range: NSMakeRange(0, attrString.length))
lblRight.attributedText = attrString
}
}
}
I've already tried line spacing and that just changes the size between all lines so the space between Adobe Creative and Suite takes on that change as well.
Try:
lblLeft.numberOfLines = 0
lblLeft.lineBreakMode = .byWordWrapping
lblLeft.sizeToFit()
By setting number of lines to zero and turning word wrapping on, the label will grow to the required number of lines. sizeToFit() should size it properly.
Can somebody help me with the following code snippet to capture part of or the whole desktop on OSX ? I would like to specify the upper-left corner coordinates (x,y) and the width (w) and height (h) of the rectangle that defines the capture.
It's for a C# MonoMac application on OSX.
This is what I've done:
int windowNumber = 2;
System.Drawing.RectangleF bounds = new RectangleF(0,146,320,157);
CGImage screenImage = MonoMac.CoreGraphics.CGImage.ScreenImage(windowNumber,bounds);
MonoMac.Foundation.NSData bitmapData = screenImage.DataProvider.CopyData();
It looks like I have the bitmap data in 'bitmapData', but I'm not sure how I convert the NSData instance 'bitmapData' to an actual Bitmap; i.e. :
Bitmap screenCapture = ????
The documentation is really sparse and I've googled for examples without luck. So I'm hoping that there's a kind MonoMac expert out there who can point me in the right direction? - An example would be nice :o)
Thank you in advance!
This will give you the bytes of your capture in a .NET byte[], from where you can create a Bitmap or Image or whatever you want. Might not be exactly what you are looking for but should put you in the right direction.
int windowNumber = 2; System.Drawing.RectangleF bounds = new RectangleF(0,146,320,157);
CGImage screenImage = MonoMac.CoreGraphics.CGImage.ScreenImage(windowNumber,bounds);
using(NSBitmapImageRep imageRep = new NSBitmapImageRep(screenImage))
{
NSDictionary properties = NSDictionary.FromObjectAndKey(new NSNumber(1.0), new NSString("NSImageCompressionFactor"));
using(NSData tiffData = imageRep.RepresentationUsingTypeProperties(NSBitmapImageFileType.Png, properties))
{
byte[] imageBytes;
using(var ms = new MemoryStream())
{
tiffData.AsStream().CopyTo(ms);
imageBytes = ms.ToArray();
}
}
}
I'm trying to make a script that will resize the images in a google doc. What I have is:
var imgs = currentDoc.getImages();
for (var i = 1; i < imgs.length; i++)
{
cell = row.insertTableCell(1);
imgNew = imgs[i].setWidth(365);
cell.insertImage(1, imgNew.getBlob());
}
The image gets inserted correctly but the size does not change regardless of what I set the width to. Since the image is going into a cell (width=370), is it possible to just make the image take up 100% of the width and scale the height proportionally? If not I can deal with manually setting the number of pixels but that is not working either. Any ideas?
The problem is that the image size should be changed after it is inserted to a table. The following code works correctly
function test() {
var doc = DocumentApp.openById('here_is_doc_id');
var imgs = doc.getImages();
var table = doc.getTables()[0];
for (var i = 0; i < imgs.length; i++) {
var row = table.appendTableRow();
var cell = row.insertTableCell(0);
var imgNew = imgs[i].copy();
cell.insertImage(0, imgNew);
imgNew.setWidth(365);
}
}
Please mention, that array indexes, cells numbers, etc. start from 0 and not 1.
Just as an FYI, you don't need to call getBlob()... anything that has a getBlob() can be passed in directly wherever a Blob is needed.
Have you tried:
imgs[i].attr('width', '370');
Or try assigning a class that has width: 100%
I am in the process of moving a webforms app to MVC3. Ironically enough, everything is cool beans except one thing - images are served from a handler, specifically the Microsoft Generated Image Handler. It works really well - on average a 450kb photo gets output at roughly 20kb.
The actual photo on disk weighs in at 417kb, so i am getting a great reduction.
Moving over to MVC3 i would like to drop the handler and use a controller action. However i seem to be unable to achieve the same kind of file size reduction when rendering the image. I walked through the source and took an exact copy of their image transform code yet i am only achieving 230~kb, which is still a lot bigger than what the ms handler is outputting - 16kb.
You can see an example of both the controller and the handler here
I have walked through the handler source code and cannot see anything that is compressing the image further. If you examine both images you can see a difference - the handler rendered image is less clear, more grainy looking, but still what i would consider satisfactory for my needs.
Can anyone give me any pointers here? is output compression somehow being used? or am i overlooking something very obvious?
The code below is used in my home controller to render the image, and is an exact copy of the FitImage method in the Image Transform class that the handler uses ...
public ActionResult MvcImage()
{
var file = Server.MapPath("~/Content/test.jpg");
var img = System.Drawing.Image.FromFile(file);
var sizedImg = MsScale(img);
var newFile = Server.MapPath("~/App_Data/test.jpg");
if (System.IO.File.Exists(newFile))
{
System.IO.File.Delete(newFile);
}
sizedImg.Save(newFile);
return File(newFile, "image/jpeg");
}
private Image MsScale(Image img)
{
var scaled_height = 267;
var scaled_width = 400;
int resizeWidth = 400;
int resizeHeight = 267;
if (img.Height == 0)
{
resizeWidth = img.Width;
resizeHeight = scaled_height;
}
else if (img.Width == 0)
{
resizeWidth = scaled_width;
resizeHeight = img.Height;
}
else
{
if (((float)img.Width / (float)img.Width < img.Height / (float)img.Height))
{
resizeWidth = img.Width;
resizeHeight = scaled_height;
}
else
{
resizeWidth = scaled_width;
resizeHeight = img.Height;
}
}
Bitmap newimage = new Bitmap(resizeWidth, resizeHeight);
Graphics gra = Graphics.FromImage(newimage);
SetupGraphics(gra);
gra.DrawImage(img, 0, 0, resizeWidth, resizeHeight);
return newimage;
}
private void SetupGraphics(Graphics graphics)
{
graphics.CompositingMode = CompositingMode.SourceCopy;
graphics.CompositingQuality = CompositingQuality.HighSpeed;
graphics.InterpolationMode = InterpolationMode.HighQualityBicubic;
graphics.SmoothingMode = SmoothingMode.HighSpeed;
}
If you don't set the quality on the encoder, it uses 100 by default. You'll never get a good size reduction by using 100 due to the way image formats like JPEG work. I've got a VB.net code example of how to set the quality parameter that you should be able to adapt.
80L here is the quality setting. 80 still gives you a fairly high quality image, but at DRASTIC size reduction over 100.
Dim graphic As System.Drawing.Graphics = System.Drawing.Graphics.FromImage(newImage)
graphic.InterpolationMode = Drawing.Drawing2D.InterpolationMode.HighQualityBicubic
graphic.SmoothingMode = Drawing.Drawing2D.SmoothingMode.HighQuality
graphic.PixelOffsetMode = Drawing.Drawing2D.PixelOffsetMode.HighQuality
graphic.CompositingQuality = Drawing.Drawing2D.CompositingQuality.HighQuality
graphic.DrawImage(sourceImage, 0, 0, width, height)
' now encode and send the new image
' This is the important part
Dim info() As Drawing.Imaging.ImageCodecInfo = Drawing.Imaging.ImageCodecInfo.GetImageEncoders()
Dim encoderParameters As New Drawing.Imaging.EncoderParameters(1)
encoderParameters.Param(0) = New Drawing.Imaging.EncoderParameter(Drawing.Imaging.Encoder.Quality, 80L)
ms = New System.IO.MemoryStream
newImage.Save(ms, info(1), encoderParameters)
When you save or otherwise write the image after setting the encoder parameters, it'll output it using the JPEG encoder (in this case) set to quality 80. That will get you the size savings you're looking for.
I believe it's defaulting to PNG format also, although Tridus' solution solves that also.
However, I highly suggest using this MVC-friendly library instead, as it avoids all the image resizing pitfalls and doesn't leak memory. It's very lightweight, free, and fully supported.