使用iText 7，导出Flate编码图像的正确方法是什么? [英] Using iText 7, what's the proper way to export a Flate encoded image?
问题描述
我正在尝试创建代码以使用iText版本7.19导出PDF中的图像.我在Flate编码的图像上遇到了一些问题.我以微软免费书中的所有Flate编码的图像为例(请参阅
I am trying to create code to export out the images within a PDF using iText Version 7.19. I'm having some issues with Flate encoded images. All the Flate encoded images from the Microsoft free book I'm using as an example (see Moving to Microsoft Visual Studio 2010) always coming out pink and depending upon how I try to copy the bytes they can come out distorted.
如果我尝试一次复制所有图像字节(请参见下面的代码中的SaveFlateEncodedImage2方法)，它们会像这样失真地出现:
If I attempt to copy all the image bytes at once (see the SaveFlateEncodedImage2 method in the code below), they come out distorted like this one:
如果我尝试逐行复制它们(请参见下面的代码中的SaveFlateEncodedImage方法)，则它们是粉红色的
If I attempt to copy them row by row (see the SaveFlateEncodedImage method in the code below), they are pink like this one
这是我用来导出它们的代码:
Here is the code that I'm using to export them:
using iText.Kernel;
using iText.Kernel.Pdf;
using iText.Kernel.Pdf.Filters;
using System;
using System.Drawing;
using System.Drawing.Imaging;
using System.IO;
using System.Runtime.InteropServices;
namespace ITextPdfStuff
{
public class MyPdfImageExtractor
{
private readonly string _pdfFileName;
public MyPdfImageExtractor(string pdfFileName)
{
_pdfFileName = pdfFileName;
}
public void ExtractToDirectory(string directoryName)
{
using (var reader = new PdfReader(_pdfFileName))
{
// Avoid iText.Kernel.Crypto.BadPasswordException: https://stackoverflow.com/a/48065052/97803
reader.SetUnethicalReading(true);
using (var pdfDoc = new PdfDocument(reader))
{
ExtractImagesOnAllPages(pdfDoc, directoryName);
}
}
}
private void ExtractImagesOnAllPages(PdfDocument pdfDoc, string directoryName)
{
Console.WriteLine($"Number of pdf {pdfDoc.GetNumberOfPdfObjects()} objects");
// Extract objects https://itextpdf.com/en/resources/examples/itext-7/extracting-objects-pdf
for (int objNumber = 1; objNumber <= pdfDoc.GetNumberOfPdfObjects(); objNumber++)
{
PdfObject currentObject = pdfDoc.GetPdfObject(objNumber);
if (currentObject != null && currentObject.IsStream())
{
try
{
ExtractImagesOneImage(currentObject as PdfStream, Path.Combine(directoryName, $"image{objNumber}.png"));
}
catch (Exception ex)
{
Console.WriteLine($"Object number {objNumber} is NOT an image! -- error: {ex.Message}");
}
}
}
}
private void ExtractImagesOneImage(PdfStream someStream, string fileName)
{
var pdfDict = (PdfDictionary)someStream;
string subType = pdfDict.Get(PdfName.Subtype)?.ToString() ?? string.Empty;
bool isImage = subType == "/Image";
if (isImage == false)
return;
bool decoded = false;
string filter = pdfDict.Get(PdfName.Filter).ToString();
if (filter == "/FlateDecode")
{
SaveFlateEncodedImage(fileName, pdfDict, someStream.GetBytes(false));
}
else
{
byte[] imgData;
try
{
imgData = someStream.GetBytes(decoded);
}
catch (PdfException ex)
{
imgData = someStream.GetBytes(!decoded);
}
SaveNormalImage(fileName, imgData);
}
}
private void SaveNormalImage(string fileName, byte[] imgData)
{
using (var memStream = new System.IO.MemoryStream(imgData))
using (var image = System.Drawing.Image.FromStream(memStream))
{
image.Save(fileName, ImageFormat.Png);
Console.WriteLine($"{Path.GetFileName(fileName)}");
}
}
private void SaveFlateEncodedImage(string fileName, PdfDictionary pdfDict, byte[] imgData)
{
int width = int.Parse(pdfDict.Get(PdfName.Width).ToString());
int height = int.Parse(pdfDict.Get(PdfName.Height).ToString());
int bpp = int.Parse(pdfDict.Get(PdfName.BitsPerComponent).ToString());
// Example that helped: https://stackoverflow.com/a/8517377/97803
PixelFormat pixelFormat;
switch (bpp)
{
case 1:
pixelFormat = PixelFormat.Format1bppIndexed;
break;
case 8:
pixelFormat = PixelFormat.Format8bppIndexed;
break;
case 24:
pixelFormat = PixelFormat.Format24bppRgb;
break;
default:
throw new Exception("Unknown pixel format " + bpp);
}
// .NET docs https://api.itextpdf.com/iText7/dotnet/7.1.9/classi_text_1_1_kernel_1_1_pdf_1_1_filters_1_1_flate_decode_strict_filter.html
// Java docs have more detail: https://api.itextpdf.com/iText7/java/7.1.7/com/itextpdf/kernel/pdf/filters/FlateDecodeFilter.html
imgData = FlateDecodeStrictFilter.FlateDecode(imgData, true);
// byte[] streamBytes = FlateDecodeStrictFilter.DecodePredictor(imgData, pdfDict);
// Copy the image one row at a time
using (var bmp = new Bitmap(width, height, pixelFormat))
{
BitmapData bmpData = bmp.LockBits(new Rectangle(0, 0, width, height), ImageLockMode.WriteOnly, pixelFormat);
int length = (int)Math.Ceiling(width * bpp / 8.0);
for (int i = 0; i < height; i++)
{
int offset = i * length;
int scanOffset = i * bmpData.Stride;
Marshal.Copy(imgData, offset, new IntPtr(bmpData.Scan0.ToInt64() + scanOffset), length);
}
bmp.UnlockBits(bmpData);
bmp.Save(fileName, ImageFormat.Png);
}
Console.WriteLine($"FlateDecode! {Path.GetFileName(fileName)}");
}
/// <summary>This method distorts the image badly</summary>
private void SaveFlateEncodedImage2(string fileName, PdfDictionary pdfDict, byte[] imgData)
{
int width = int.Parse(pdfDict.Get(PdfName.Width).ToString());
int height = int.Parse(pdfDict.Get(PdfName.Height).ToString());
int bpp = int.Parse(pdfDict.Get(PdfName.BitsPerComponent).ToString());
// Example that helped: https://stackoverflow.com/a/8517377/97803
PixelFormat pixelFormat;
switch (bpp)
{
case 1:
pixelFormat = PixelFormat.Format1bppIndexed;
break;
case 8:
pixelFormat = PixelFormat.Format8bppIndexed;
break;
case 24:
pixelFormat = PixelFormat.Format24bppRgb;
break;
default:
throw new Exception("Unknown pixel format " + bpp);
}
// .NET docs https://api.itextpdf.com/iText7/dotnet/7.1.9/classi_text_1_1_kernel_1_1_pdf_1_1_filters_1_1_flate_decode_strict_filter.html
// Java docs have more detail: https://api.itextpdf.com/iText7/java/7.1.7/com/itextpdf/kernel/pdf/filters/FlateDecodeFilter.html
imgData = FlateDecodeStrictFilter.FlateDecode(imgData, true);
// byte[] streamBytes = FlateDecodeStrictFilter.DecodePredictor(imgData, pdfDict);
// Copy the entire image in one go
using (var bmp = new Bitmap(width, height, pixelFormat))
{
BitmapData bmpData = bmp.LockBits(new Rectangle(0, 0, width, height), ImageLockMode.WriteOnly, pixelFormat);
Marshal.Copy(imgData, 0, bmpData.Scan0, imgData.Length);
bmp.UnlockBits(bmpData);
bmp.Save(fileName, ImageFormat.Png);
}
Console.WriteLine($"FlateDecode! {Path.GetFileName(fileName)}");
}
}
}
可以在.NET Core控制台应用程序中实例化并像这样调用代码:
The code can be instantiated and called like this from within a .NET Core console application:
string existingFileName = @"c:\temp\ReallyLongBook1.pdf";
var imageExtractor = new MyPdfImageExtractor(existingFileName);
imageExtractor.ExtractToDirectory(@"c:\temp\images");
我正在通过此代码运行以下免费的Microsoft书籍: 移动到Microsoft Visual Studio 2010
I'm running the following free Microsoft book through this code: Moving to Microsoft Visual Studio 2010
有问题的图片在第10页上，它是黑白的(不是粉红色的).
The image in question is on page 10 and it's black and white (not pink).
我不是PDF专家，而且我一直在研究此代码两天，现在挑选了一些示例以尝试将其组合在一起.任何帮助我摆脱粉红色图像的帮助，将不胜感激.
I'm no PDF expert and I've been banging on this code for a couple of days now picking apart a number of examples to try to piece this together. Any help that would get me past my pink images, would be greatly appreciated.
-------更新2020年2月4日------
-------Update Feb 4, 2020------
此处是MKL建议的更改之后的修订版本.他的变化所提取的图像比我的更多，并产生了我上面提到的书中出现的适当外观的图像:
Here is the revised version after MKL's suggested changes. His change extracted more images than mine and produced proper looking images that appear in the book I mentioned above:
using iText.Kernel.Pdf;
using iText.Kernel.Pdf.Canvas.Parser;
using iText.Kernel.Pdf.Canvas.Parser.Data;
using iText.Kernel.Pdf.Canvas.Parser.Listener;
using iText.Kernel.Pdf.Xobject;
using System;
using System.Collections.Generic;
using System.IO;
namespace ITextPdfStuff
{
public class MyPdfImageExtractor
{
private readonly string _pdfFileName;
public MyPdfImageExtractor(string pdfFileName)
{
_pdfFileName = pdfFileName;
}
public void ExtractToDirectory(string directoryName)
{
using (var reader = new PdfReader(_pdfFileName))
{
// Avoid iText.Kernel.Crypto.BadPasswordException: https://stackoverflow.com/a/48065052/97803
reader.SetUnethicalReading(true);
using (var pdfDoc = new PdfDocument(reader))
{
ExtractImagesOnAllPages(pdfDoc, directoryName);
}
}
}
private void ExtractImagesOnAllPages(PdfDocument pdfDoc, string directoryName)
{
Console.WriteLine($"Number of pdf {pdfDoc.GetNumberOfPdfObjects()} objects");
IEventListener strategy = new ImageRenderListener(Path.Combine(directoryName, @"image{0}.{1}"));
PdfCanvasProcessor parser = new PdfCanvasProcessor(strategy);
for (var i = 1; i <= pdfDoc.GetNumberOfPages(); i++)
{
parser.ProcessPageContent(pdfDoc.GetPage(i));
}
}
}
public class ImageRenderListener : IEventListener
{
public ImageRenderListener(string format)
{
this.format = format;
}
public void EventOccurred(IEventData data, EventType type)
{
if (data is ImageRenderInfo imageData)
{
try
{
PdfImageXObject imageObject = imageData.GetImage();
if (imageObject == null)
{
Console.WriteLine("Image could not be read.");
}
else
{
File.WriteAllBytes(string.Format(format, index++, imageObject.IdentifyImageFileExtension()), imageObject.GetImageBytes());
}
}
catch (Exception ex)
{
Console.WriteLine("Image could not be read: {0}.", ex.Message);
}
}
}
public ICollection<EventType> GetSupportedEvents()
{
return null;
}
string format;
int index = 0;
}
}
推荐答案
PDF内部支持非常灵活的位图图像格式，尤其是在涉及不同颜色空间时.
PDFs internally support a very flexible bitmap image format, in particular as far as different color spaces are concerned.
iText的解析API支持导出其子集，本质上是可以轻松导出为常规JPEG或PNG的图像子集.
iText in its parsing API supports export of a subset thereof, essentially the subset of images that easily can be exported as regular JPEGs or PNGs.
因此，首先尝试使用iText解析API进行导出是有意义的.您可以按照以下步骤进行操作:
Thus, it makes sense to try and export using the iText parsing API first. You can do that as follows:
Directory.CreateDirectory(@"extract\");
using (PdfReader reader = new PdfReader(@"Moving to Microsoft Visual Studio 2010 ebook.pdf"))
using (PdfDocument pdfDocument = new PdfDocument(reader))
{
IEventListener strategy = new ImageRenderListener(@"extract\Moving to Microsoft Visual Studio 2010 ebook-i7-{0}.{1}");
PdfCanvasProcessor parser = new PdfCanvasProcessor(strategy);
for (var i = 1; i <= pdfDocument.GetNumberOfPages(); i++)
{
parser.ProcessPageContent(pdfDocument.GetPage(i));
}
}
具有帮助器类ImageRenderListener:
public class ImageRenderListener : IEventListener
{
public ImageRenderListener(string format)
{
this.format = format;
}
public void EventOccurred(IEventData data, EventType type)
{
if (data is ImageRenderInfo imageData)
{
try
{
PdfImageXObject imageObject = imageData.GetImage();
if (imageObject == null)
{
Console.WriteLine("Image could not be read.");
}
else
{
File.WriteAllBytes(string.Format(format, index++, imageObject.IdentifyImageFileExtension()), imageObject.GetImageBytes());
}
}
catch (Exception ex)
{
Console.WriteLine("Image could not be read: {0}.", ex.Message);
}
}
}
public ICollection<EventType> GetSupportedEvents()
{
return null;
}
string format;
int index = 0;
}
如果您的示例文档成功导出了近400张图像，其中包括您上面的示例图像:
In case of your example document it exports nearly 400 images successfully, among them your example image above:
但是无法导出的图像也少于30张，在标准输出上，您会发现无法读取图像:不支持/DeviceN颜色空间."
But there also are less than 30 images it cannot export, on standard out you'll find "Image could not be read: The color space /DeviceN is not supported.."
这篇关于使用iText 7，导出Flate编码图像的正确方法是什么?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
