public void PutRectAnno(float lx, float ly, float rx, float ry)
{
using (FileStream fs = new FileStream(highLightFile, FileMode.Create, FileAccess.Write, FileShare.None))
{
using (PdfStamper stamper = new PdfStamper(reader, fs))
{
iTextSharp.text.Rectangle rect = new iTextSharp.text.Rectangle(lx, ly, rx, ry);
float[] quad = { rect.Left, rect.Top, rect.Right, rect.Top, rect.Left, rect.Bottom, rect.Right, rect.Bottom };
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, null, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
highlight.Color = BaseColor.YELLOW;
if (rect.Width > 0 && rect.Height > 0)
{
stamper.AddAnnotation(highlight, 1);
}
stamper.Close();
}
fs.Close();
}
}
public void HighLightText(String[] searchText)
{
//Create a new file from our test file with highlighting
string highLightFile = Path.Combine(m_DirPath, "Highlighted.pdf");
MyTextExtractionStrategy[,] arr_t = null;
int Pages = 0;
//Parse page 1 of the document above
using (var r = new PdfReader(m_filename))
{
Pages = r.NumberOfPages;
//Create an array of our strategy
arr_t = new MyTextExtractionStrategy[r.NumberOfPages, searchText.Length];
for (int i = 0; i < r.NumberOfPages; i++)
{
for (int j = 0; j < searchText.Length; j++)
{
arr_t[i, j] = new MyTextExtractionStrategy(searchText[j], 1056f);
var ex = PdfTextExtractor.GetTextFromPage(r, i + 1, arr_t[i, j]);
}
}
}
//Bind a reader and stamper to our test PDF
PdfReader reader = new PdfReader(m_filename);
using (FileStream fs = new FileStream(highLightFile, FileMode.Create, FileAccess.Write, FileShare.None))
{
using (PdfStamper stamper = new PdfStamper(reader, fs))
{
for (int i = 0; i < Pages; i++)
{
for (int j = 0; j < searchText.Length; j++)
{
var t = arr_t[i, j];
foreach (var p in t.m_SearchResultsList)
{
Rectangle rect = p.rect;
//Create an array of quad points based on that rectangle. NOTE: The order below doesn't appear to match the actual spec but is what Acrobat produces
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
//Create our highlight
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, null, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
//Set the color
highlight.Color = BaseColor.YELLOW;
//Add the annotation
stamper.AddAnnotation(highlight, i + 1);
}
}
}
}
}
}
// private void highlightPDFAnnotation(string readerPath, string outputFile, int pageno, string[] highlightText)
private void highlightPDFAnnotation(string readerPath, string outputFile, string[] highlightText)
{
PdfReader reader = new PdfReader(readerPath);
PdfContentByte canvas;
using (FileStream fs = new FileStream(outputFile, FileMode.Create, FileAccess.Write, FileShare.None))
{
using (PdfStamper stamper = new PdfStamper(reader, fs))
{
int pageCount = reader.NumberOfPages;
for (int pageno = 1; pageno <= pageCount; pageno++)
{
var strategy = new HighLightTextLocation();
strategy.UndercontentHorizontalScaling = 100;
string currentText = PdfTextExtractor.GetTextFromPage(reader, pageno, strategy);
for (int i = 0; i < highlightText.Length; i++)
{
List <Rectangle> MatchesFound = strategy.GetTextLocations(highlightText[i].Trim(), StringComparison.CurrentCultureIgnoreCase);
foreach (Rectangle rect in MatchesFound)
{
float[] quad = { rect.Left - 3.0f, rect.Bottom, rect.Right, rect.Bottom, rect.Left - 3.0f, rect.Top + 1.0f, rect.Right, rect.Top + 1.0f };
//Create our hightlight
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, null, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
//Set the color
highlight.Color = BaseColor.YELLOW;
PdfAppearance appearance = PdfAppearance.CreateAppearance(stamper.Writer, rect.Width, rect.Height);
PdfGState state = new PdfGState();
state.BlendMode = new PdfName("Multiply");
appearance.SetGState(state);
appearance.Rectangle(0, 0, rect.Width, rect.Height);
appearance.SetColorFill(BaseColor.YELLOW);
appearance.Fill();
highlight.SetAppearance(PdfAnnotation.APPEARANCE_NORMAL, appearance);
//Add the annotation
stamper.AddAnnotation(highlight, pageno);
}
}
}
}
}
reader.Close();
}
private void Form1_Load(object sender, EventArgs e)
{
//Create a simple test file
string outputFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "Test.pdf");
using (FileStream fs = new FileStream(outputFile, FileMode.Create, FileAccess.Write, FileShare.None))
{
using (Document doc = new Document(PageSize.LETTER))
{
using (PdfWriter w = PdfWriter.GetInstance(doc, fs))
{
doc.Open();
doc.Add(new Paragraph("This is a test"));
doc.Close();
}
}
}
//Create a new file from our test file with highlighting
string highLightFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "Highlighted.pdf");
//Bind a reader and stamper to our test PDF
PdfReader reader = new PdfReader(outputFile);
using (FileStream fs = new FileStream(highLightFile, FileMode.Create, FileAccess.Write, FileShare.None))
{
using (PdfStamper stamper = new PdfStamper(reader, fs))
{
//Create a rectangle for the highlight. NOTE: Technically this isn't used but it helps with the quadpoint calculation
iTextSharp.text.Rectangle rect = new iTextSharp.text.Rectangle(60.6755f, 749.172f, 94.0195f, 735.3f);
//Create an array of quad points based on that rectangle. NOTE: The order below doesn't appear to match the actual spec but is what Acrobat produces
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
//Create our hightlight
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, null, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
//Set the color
highlight.Color = BaseColor.YELLOW;
//Add the annotation
stamper.AddAnnotation(highlight,1);
}
}
this.Close();
}
//--------------------------------------------------------------------------------------------------
void AddHeader(PdfStamper stamper, PdfReader reader, string stComment)
{
PdfContentByte canvas = stamper.GetOverContent(1);
canvas.BeginText();
canvas.SetFontAndSize(GetDefaultFont(), 12);
Rectangle rect = reader.GetPageSizeWithRotation(1);
canvas.ShowTextAligned(PdfContentByte.ALIGN_CENTER, stComment, (rect.Left + rect.Right) / 2, rect.Top - 12, 0);
canvas.EndText();
rect.Bottom = rect.Top - 18;
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, stComment, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
highlight.Color = BaseColor.YELLOW;
highlight.Title = "PDFCompare";
stamper.AddAnnotation(highlight, 1);
}
/// <summary>
/// Highlights the PDF with Yellow annotation
/// </summary>
/// <param name="inputFile">File to read from</param>
/// <param name="highLightFile">File to write to</param>
/// <param name="pageno">Which page no to read</param>
/// <param name="textToAnnotate">Texts to annotate</param>
private void HighlightPDFAnnotation(string inputFile, string highLightFile, int pageno, params string[] textToAnnotate)
{
PdfReader reader = new PdfReader(inputFile);
using (FileStream fs = new FileStream(highLightFile, FileMode.Create, FileAccess.Write, FileShare.None))
{
using (PdfStamper stamper = new PdfStamper(reader, fs))
{
MyLocationTextExtractionStrategy strategy = new MyLocationTextExtractionStrategy();
strategy.UndercontentHorizontalScaling = 100;
string currentText = PdfTextExtractor.GetTextFromPage(reader, pageno, strategy);
for (int i = 0; i < textToAnnotate.Length; i++)
{
if (string.IsNullOrWhiteSpace(textToAnnotate[i]))
{
continue;
}
var lstMatches = strategy.GetTextLocations(textToAnnotate[i].Trim(), StringComparison.CurrentCultureIgnoreCase);
if (!this.chkAnnotation.Checked)
{
lstMatches = lstMatches.Take(1).ToList();
}
foreach (iTextSharp.text.Rectangle rectangle in lstMatches)
{
float[] quadPoints = { rectangle.Left - 3.0f,
rectangle.Bottom,
rectangle.Right,
rectangle.Bottom,
rectangle.Left - 3.0f,
rectangle.Top + 1.0f,
rectangle.Right,
rectangle.Top + 1.0f };
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer
, rectangle, null
, PdfAnnotation.MARKUP_HIGHLIGHT, quadPoints);
highlight.Color = BaseColor.YELLOW;
PdfGState state = new PdfGState();
state.BlendMode = new PdfName("Multiply");
PdfAppearance appearance = PdfAppearance.CreateAppearance(stamper.Writer, rectangle.Width, rectangle.Height);
appearance.SetGState(state);
appearance.Rectangle(0, 0, rectangle.Width, rectangle.Height);
appearance.SetColorFill(BaseColor.YELLOW);
appearance.Fill();
highlight.SetAppearance(PdfAnnotation.APPEARANCE_NORMAL, appearance);
//Add the annotation
stamper.AddAnnotation(highlight, pageno);
}
}
}
}
reader.Close();
}
public void ProcessPdf(StringComparison sc, string sourceFile, string destinationFile, string searchTerm, int excelRowNumber, List <KeyValuePair <int, string> > searchValues)
{
var sArr = searchTerm.Split(',');
myProgressBar.Maximum = searchValues.Count;
bool found = false;
Cursor = Cursors.WaitCursor;
if (File.Exists(sourceFile))
{
var pReader = new PdfReader(sourceFile);
myProgressBar.Value = 0;
PdfStamper stamper = null;
foreach (var item in searchValues)
{
var newStrings = item.Value.Split(',');
var foundText = string.Empty;
foreach (var search in newStrings)
{
for (var page = 1; page <= pReader.NumberOfPages; page++)
{
var t = new MyLocationTextExtractionStrategy(search, CompareOptions.Ordinal);
using (var r = new PdfReader(sourceFile))
{
var ex = PdfTextExtractor.GetTextFromPage(r, 1, t);
}
var matchesFound = t.MyPoints;
if (t.MyPoints.Count > 0)
{
found = true;
if (!string.IsNullOrEmpty(search))
{
foundText += "," + search;
}
if (!File.Exists(destinationFile))
{
stamper = new PdfStamper(pReader, new FileStream(destinationFile, FileMode.Create));
}
if (!_fileList.Contains(destinationFile))
{
_fileList.Add(destinationFile);
}
var cb = stamper.GetUnderContent(page);
cb.SetColorFill(BaseColor.BLACK);
foreach (var rect in matchesFound)
{
if (rect.Text == search)
{
cb.Rectangle(rect.Rect.Left, rect.Rect.Bottom, rect.Rect.Width, rect.Rect.Height);
float[] quad =
{
rect.Rect.Left,
rect.Rect.Bottom,
rect.Rect.Right,
rect.Rect.Bottom,
rect.Rect.Left,
rect.Rect.Top,
rect.Rect.Right,
rect.Rect.Top
};
var highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect.Rect,
Constants.vbNull.ToString(), PdfAnnotation.MARKUP_HIGHLIGHT, quad);
highlight.Color = BaseColor.YELLOW;
stamper.AddAnnotation(highlight, page);
}
}
}
}
}
if (found && !string.IsNullOrEmpty(foundText))
{
if (sourceFile == txtFirstPDF.Text)
{
_foundValuesInFirstPdf.Add(new KeyValuePair <int, string>(item.Key, foundText.Substring(1)));
}
else
{
_foundValuesInSecindPdf.Add(new KeyValuePair <int, string>(item.Key, foundText.Substring(1)));
}
}
myProgressBar.Value = myProgressBar.Value + 1;
}
if (stamper != null)
{
stamper.Close();
}
}
this.Cursor = Cursors.Default;
}
//--------------------------------------------------------------------------------------------------
void MarkText(List <TextItem> items, int iPosition, int iCount, PdfStamper stamper, BaseColor color, int iPageAdd = 0, string stComment = null)
{
const string stTittle = "PDFCompare";
if (iCount == 0)
{
float rWidth, rPosition;
if (iPosition >= items.Count)
{
iPosition = items.Count - 1;
rWidth = (items[iPosition].MaxY - items[iPosition].MinY) / 5;
rPosition = items[iPosition].MaxX;
}
else
{
rWidth = (items[iPosition].MaxY - items[iPosition].MinY) / 5;
rPosition = items[iPosition].MinX - rWidth;
}
Rectangle rect = new Rectangle(rPosition, items[iPosition].MaxY, rPosition + rWidth, items[iPosition].MinY);
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, stComment, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
highlight.Color = color;
highlight.Title = stTittle;
stamper.AddAnnotation(highlight, items[iPosition].Page + iPageAdd);
}
else
{
while (iCount > 0 && items[iPosition].IsImage)
{
TextItem item = items[iPosition];
Rectangle rect = new Rectangle(item.MinX, item.MaxY, item.MaxX, item.MinY);
PdfArray vertices = new PdfArray();
float rHalfWidth = 4;
vertices.Add(new PdfNumber(item.MinX - rHalfWidth));
vertices.Add(new PdfNumber(item.MinY - rHalfWidth));
vertices.Add(new PdfNumber(item.MaxX + rHalfWidth));
vertices.Add(new PdfNumber(item.MinY - rHalfWidth));
vertices.Add(new PdfNumber(item.MaxX + rHalfWidth));
vertices.Add(new PdfNumber(item.MaxY + rHalfWidth));
vertices.Add(new PdfNumber(item.MinX - rHalfWidth));
vertices.Add(new PdfNumber(item.MaxY + rHalfWidth));
vertices.Add(new PdfNumber(item.MinX - rHalfWidth));
vertices.Add(new PdfNumber(item.MinY - rHalfWidth));
PdfAnnotation highlight = PdfAnnotation.CreatePolygonPolyline(stamper.Writer, rect, stComment, true, vertices);
highlight.Color = color;
highlight.Title = stTittle;
highlight.Border = new PdfBorderArray(0, 0, rHalfWidth * 2);
stamper.AddAnnotation(highlight, item.Page + iPageAdd);
iPosition++;
iCount--;
}
if (iCount == 0)
{
return;
}
int i;
for (i = iPosition; i < iPosition + iCount; i++)
{
if (items[i].IsImage)
{
MarkText(items, iPosition, i - iPosition, stamper, color, iPageAdd, stComment);
MarkText(items, i, 1, stamper, color, iPageAdd, stComment);
if (i + 1 >= iPosition + iCount)
{
return;
}
iCount -= (i - iPosition) + 1;
iPosition = i + 1;
}
}
if (!items[iPosition].IsNewLine)
{
i = iPosition;
while (i < iPosition + iCount - 1 && !items[i + 1].IsNewLine)
{
i++;
}
Rectangle rect = new Rectangle(items[iPosition].MinX, items[iPosition].MaxY, items[i].MaxX, items[iPosition].MinY);
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, stComment, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
highlight.Color = color;
highlight.Title = stTittle;
stamper.AddAnnotation(highlight, items[iPosition].Page + iPageAdd);
iCount -= (i - iPosition) + 1;
iPosition = i + 1;
}
if (iCount == 0)
{
return;
}
if (items.Count > iPosition + iCount && !items[iPosition + iCount].IsNewLine)
{
i = iPosition + iCount - 1;
while (!items[i].IsNewLine)
{
i--;
}
Rectangle rect = new Rectangle(items[i].MinX, items[i].MaxY, items[i = iPosition + iCount - 1].MaxX, items[i].MinY);
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, stComment, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
highlight.Color = color;
highlight.Title = stTittle;
stamper.AddAnnotation(highlight, items[iPosition].Page + iPageAdd);
iCount -= iPosition + iCount - i;
}
if (iCount == 0)
{
return;
}
i = iPosition;
float rMinX = items[iPosition].MinX;
float rMaxX = items[iPosition].MaxX;
while (i < iPosition + iCount)
{
i++;
if (i == iPosition + iCount || items[iPosition].Page != items[i].Page)
{
Rectangle rect = new Rectangle(rMinX, items[iPosition].MaxY, rMaxX, items[i - 1].MinY);
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, stComment, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
highlight.Color = color;
highlight.Title = stTittle;
stamper.AddAnnotation(highlight, items[iPosition].Page + iPageAdd);
iCount -= i - iPosition;
iPosition = i;
if (iCount == 0)
{
return;
}
rMinX = items[iPosition].MinX;
rMaxX = items[iPosition].MaxX;
}
if (items[i].MinX < rMinX)
{
rMinX = items[i].MinX;
}
if (items[i].MaxX > rMaxX)
{
rMaxX = items[i].MaxX;
}
}
/*
* for(i = iPosition; i < iPosition + iCount; i++)
* {
* TextItem item = items[i];
* Rectangle rect = new Rectangle(item.MinX, item.MaxY, item.MaxX, item.MinY);
* float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
* PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, stComment, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
* highlight.Color = color;
* highlight.Title = stTittle;
* stamper.AddAnnotation(highlight, item.Page + iPageAdd);
* }
*/
}
}
public Dictionary <int, string> ExecuteHighlight(BackgroundWorker backgroundWorker)
{
OutputMg.OutputContent(backgroundWorker, "Starting highlight file " + pdfFilePath);
List <string> topicTerms = ReadTargetTopicTerms.ParseTopicTerms(this.topicTermPath, this.targetTopicName);
AddUserSearchTerms(topicTerms);
string origiFile = pdfFilePath;
//Create a new file from our test file with highlighting
string highLightFile = highlightedPDFPath;
int pdfNum = 0;
PdfReader reader = new PdfReader(origiFile);
using (FileStream fs = new FileStream(highLightFile, FileMode.Create, FileAccess.Write, FileShare.None))
{
using (PdfStamper stamper = new PdfStamper(reader, fs))
{
using (var r = new PdfReader(origiFile))
{
pdfNum = r.NumberOfPages;
string ex = "";
ITextExtractionStrategy strategy;
for (int i = 1; i <= pdfNum; i++)
{
OutputMg.OutputContent(backgroundWorker, "Parsing page: " + i);
Rectangle pageRect = r.GetPageSize(i);
Document doc = new Document(pageRect);
float leftMargin = doc.LeftMargin;
float rightMargin = doc.RightMargin;
float lineWidth = pageRect.Width;
var textPos = new FutherLocationTextExtractionStrategy(topicTerms);
//Create an instance of our strategy
ex = PdfTextExtractor.GetTextFromPage(r, i, textPos); //store the text and the position info in textPos
List <iTextSharp.text.Rectangle> quadList = new List <iTextSharp.text.Rectangle>();
foreach (var p in textPos.myPoints)
{
string p_text = p.Text;
iTextSharp.text.Rectangle rect = p.Rect;
quadList.Add(rect);//collect the coordination of keywords
}
List <string> pageContent = new List <string>();
if (quadList.Count > 0)
{
List <iTextSharp.text.Rectangle> orderedRect = orderRectByBottom(quadList);
//merge and adjust the rectangle, highlight the adjusted rect
List <iTextSharp.text.Rectangle> adjustedRect = adjustRect(orderedRect, lineWidth, leftMargin);
foreach (Rectangle rect in adjustedRect)
{
//Create an array of quad points based on that rectangle. NOTE: The order below doesn't appear to match the actual spec but is what Acrobat produces
//the co-ordination of four points
float[] quad = { rect.Left, rect.Bottom, rect.Right, rect.Bottom, rect.Left, rect.Top, rect.Right, rect.Top };
////Create our hightlight
PdfAnnotation highlight = PdfAnnotation.CreateMarkup(stamper.Writer, rect, null, PdfAnnotation.MARKUP_HIGHLIGHT, quad);
////Set the color
highlight.Color = BaseColor.YELLOW;
stamper.AddAnnotation(highlight, i); // i is the page
//get the text of highlighting
RenderFilter[] filter = { new RegionTextRenderFilter(rect) };
strategy = new MyFilteredTextRenderListener(new LocationTextExtractionStrategy(), filter);
string text = PdfTextExtractor.GetTextFromPage(reader, i, strategy).Trim();
if (!pageContent.Contains(text))
{
pageContent.Add(text);
}
}
StringBuilder sb = new StringBuilder();
foreach (string tmp in pageContent)
{
sb.AppendLine(tmp);
}
pageContents.Add(i, sb.ToString());
}
}
}
}
}
return(pageContents);
}