// Extracts all images (of types that iTextSharp knows how to decode) from a PDF file
public static List <Image> ExtractImages(Stream myBlob, string filename, TraceWriter log)
{
var images = new List <Image>();
try
{
using (var reader = new PdfReader(myBlob))
{
var parser = new PdfReaderContentParser(reader);
var listener = new ImageRenderListener(log);
for (var i = 1; i <= reader.NumberOfPages; i++)
{
parser.ProcessContent(i, listener);
if (listener.Images.Count > 0)
{
log.Verbose($"Found {listener.Images.Count} images on page {i}.");
images.AddRange(listener.Images);
listener.Images.Clear();
}
}
}
}
catch (Exception ex)
{
log.Error($"Error: {ex.Message}");
}
return(images);
}
private static Dictionary <string, System.Drawing.Image> ExtractImages(byte[] fileBytes)
{
var images = new Dictionary <string, System.Drawing.Image>();
using (var reader = new PdfReader(fileBytes))
{
var parser = new PdfReaderContentParser(reader);
for (var i = 1; i <= reader.NumberOfPages; i++)
{
ImageRenderListener listener = new ImageRenderListener();
parser.ProcessContent(i, listener);
var index = 1;
if (listener.Images.Count > 0)
{
foreach (var pair in listener.Images)
{
images.Add(string.Format("Page_{0}_Image_{1}{2}", i.ToString("D4"), index.ToString("D4"), pair.Value), pair.Key);
index++;
}
}
}
return(images);
}
}
public static List <PDFPage> GetPDFPages(Stream pdfStream, TraceWriter log, bool ocrImages = false)
{
var result = new List <PDFPage>();
pdfStream.Position = 0; // Ensure that we are at the start
// Note: PdfReader Dispose closes the stream...
using (PdfReader reader = new PdfReader(pdfStream))
{
var numberOfPages = reader.NumberOfPages;
var parser = new PdfReaderContentParser(reader);
ImageRenderListener listener = null; // = new ImageRenderListener(log);
for (var i = 1; i <= reader.NumberOfPages; i++)
{
var page = new PDFPage {
Number = i
};
try
{
parser.ProcessContent(i, (listener = new ImageRenderListener(log)));
}
catch (Exception ex)
{
log.Error(string.Format("Page {0} Image Processing Exception", i), ex);
}
if (listener.Images.Count > 0)
{
log.Info(string.Format("Found {0} images on page {1}.", listener.Images.Count, i));
page.ExtractedImages = listener.Images;
if (ocrImages)
{
if (listener.Images.Count < 10)
{
log.Info("Calling Vision API to OCR Page Images");
VisionAPIHelper.OCRPage(page, log);
}
else
{
log.Info("Too many Page Images for Vision API");
}
}
}
try
{
page.PageText = PdfTextExtractor.GetTextFromPage(reader, i, new SimpleTextExtractionStrategy());
}
catch (System.ArgumentException ex)
{
log.Error(string.Format("Page {0} Text Processing Exception", i), ex);
}
result.Add(page);
}
}
return(result);
}
public static IEnumerable <(Image image, string type)> GetImages(string filePath)
{
using (var reader = new PdfReader(filePath))
{
var parser = new PdfReaderContentParser(reader);
var renderListener = new ImageRenderListener();
for (int i = 1; i <= reader.NumberOfPages; i++)
{
parser.ProcessContent(i, renderListener);
}
return(renderListener.Buffer
.Select(info => (info.GetDrawingImage(), info.GetFileType())).ToArray());
}
}
public static PDFFileInfo GetPDFMetaData(Stream pdfStream, string filePath, bool deepinfo)
{
var fileInfo = new PDFFileInfo();
fileInfo.FilePath = filePath;
fileInfo.FileName = System.IO.Path.GetFileName(filePath);
if (pdfStream.Length == 0)
{
fileInfo.ErrorMessages.Add("Zero length file");
return(fileInfo);
}
try
{
pdfStream.Position = 0; // Ensure that we are at the start
}
catch (NotSupportedException) { }
try
{
// Note: PdfReader Dispose closes the stream...
using (PdfReader reader = new PdfReader(pdfStream))
{
fileInfo.PDFVersion = reader.PdfVersion;
fileInfo.PageCount = reader.NumberOfPages;
fileInfo.FileSize = reader.FileLength;
if (reader.Info != null)
{
try
{
if (reader.Info.ContainsKey("CreationDate"))
{
fileInfo.CreationDate = PdfDate.Decode(reader.Info["CreationDate"]);
}
if (reader.Info.ContainsKey("ModDate"))
{
fileInfo.ModDate = PdfDate.Decode(reader.Info["ModDate"]);
}
}
catch (Exception ex)
{
fileInfo.ErrorMessages.Add($"PdfDate Decode {ex.Message.Replace(',',' ')}");
}
if (reader.Info.ContainsKey("Creator"))
{
fileInfo.Creator = reader.Info["Creator"];
}
if (reader.Info.ContainsKey("Producer"))
{
fileInfo.Producer = reader.Info["Producer"];
}
}
if (deepinfo)
{
var parser = new PdfReaderContentParser(reader);
ImageRenderListener listener = null;
for (var i = 1; i <= reader.NumberOfPages; i++)
{
var page = new PDFPageInfo();
try
{
parser.ProcessContent(i, (listener = new ImageRenderListener(fileInfo.ErrorMessages)));
}
catch (Exception ex)
{
fileInfo.ErrorMessages.Add($"Page {i} Image Processing Exception: {ex.Message.Replace(',', ' ')}");
}
try
{
var pageText = PdfTextExtractor.GetTextFromPage(reader, i, new SimpleTextExtractionStrategy());
page.TextCharacters += pageText.Length;
}
catch (System.ArgumentException ex)
{
fileInfo.ErrorMessages.Add($"Page {i} Text Extraction Exception {ex.Message.Replace(',', ' ')}");
}
page.PageNum = i;
page.ImageCount = listener.Images.Count;
page.Images = listener.Images;
for (int j = 0; j < page.ImageCount; j++)
{
page.Images[j].ImageNum = j + 1;
page.ImageBytes += page.Images[j].ImageBytes;
}
fileInfo.Pages.Add(page);
fileInfo.ImageCount += page.ImageCount;
fileInfo.ImageBytes += page.ImageBytes;
fileInfo.TextCharacters += page.TextCharacters;
}
}
}
}
catch (Exception ex)
{
fileInfo.ErrorMessages.Add(ex.Message);
}
return(fileInfo);
}