This Delphi sample code shows you how to extract text and images from one PDF and insert it into a new PDF at the same locations.

----------------------------

var

  FH: Integer;

  PR: Integer;

  SL: TStringList;

  Data: string;

  Font: string;

  Color: string;

  Size: string;

  X1, Y1, X2, Y2, X3, Y3, X4, Y4: string;

  Text: string;

  X: Integer;

  IL: Integer;

  TextBlockLeft: Double;

  TextBlockTop: Double;

  PageNum: Integer;

  ImageData: string;

  ImageLeft, ImageTop, ImageWidth, ImageHeight: Double;

...

// Open the file in direct access mode and store the file handle

FH := QP.DAOpenFile('Xpod1228090001.pdf', '');

// Loop through all the pages

for PageNum := 1 to QP.DAGetPageCount(FH) do

begin

  // Start a new document

  QP.NewDocument;

  // Specify that images should be compressed

  QP.CompressImages(1);

  // Get a page reference to the current page

  PR := QP.DAFindPage(FH, PageNum);

  // Create a string list to hold the text data

  SL := TStringList.Create;

  try

    // Extract the text from the current page

    SL.Text := QP.DAExtractPageText(FH, PR, 4);

    // Add each block of text to the new documen

    for X := 0 to SL.Count - 1 do

    begin

      Data := SL[X];

      Font := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(Font) + 1);

      Color := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(Color) + 1);

      Size := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(Size) + 1);

      X1 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(X1) + 1);

      Y1 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(Y1) + 1);

      X2 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(X2) + 1);

      Y2 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(Y2) + 1);

      X3 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(X3) + 1);

      Y3 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(Y3) + 1);

      X4 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(X4) + 1);

      Y4 := Copy(Data, 1, Pos(',', Data) - 1);

      Delete(Data, 1, Length(Y4) + 1);

      Text := Copy(Data, 2, Length(Data) - 2);

      // Replace the utf-8 encoded TM symbol with the

      // PDF WinAnsi character code

      if Pos(#226#132#162, Text) > 0 then

        Text := StringReplace(Text, #226#132#162, #153,

          [rfReplaceAll]);

      // Set the text size

      QP.SetTextSize(StrToFloat(Size));

      // Draw the text, shift up by the font's "descent" value

      QP.DrawText(StrToFloat(X1),

        StrToFloat(Y1) - QP.GetTextDescent,

        Text);

    end;

  finally

    SL.Free;

  end;

  // Find all the images on the page

  IL := QP.DAGetPageImageList(FH, PR);

  // Loop through all the images

  for X := 1 to QP.DAGetImageListCount(FH, IL) do

  begin

    // Read the image data

    ImageData := QP.DAGetImageDataToString(FH, IL, X);

    // Add the image data to the new document

    QP.AddImageFromString(ImageData, 0);

    // Determine the location and size of the image on the page

    ImageLeft := QP.DAGetImageDblProperty(FH, IL, X, 501);

    ImageTop := QP.DAGetImageDblProperty(FH, IL, X, 502);

    ImageWidth := QP.DAGetImageDblProperty(FH, IL, X, 503) -

      QP.DAGetImageDblProperty(FH, IL, X, 501);

    ImageHeight := QP.DAGetImageDblProperty(FH, IL, X, 502) -

      QP.DAGetImageDblProperty(FH, IL, X, 508);

    // Draw the image onto the new document's page

    QP.DrawImage(ImageLeft, ImageTop, ImageWidth, ImageHeight);

  end;  // End image loop

  // Compress the page description commands

  QP.CompressContent;

  // Save the file

  QP.SaveToFile('XPod-' + IntToStr(PageNum) + '.pdf');

  // Remove the document

  QP.RemoveDocument(QP.SelectedDocument);

end;  // End page loop

I would like to be able to select an existing font in the document so I can then do some calculations on character widths of the strings returned from the DAExtractPageText call.

 

I am not there in my testing yet, but I would assume that this is NOT the proper way to copy pages from one file to another, but it does show text and image extraction quite well.

 

Howard