private void button1_Click(object sender, EventArgs e)
{
String input = @"..\..\..\..\..\..\Data\PDFTemplate-Az.pdf";
PdfDocument doc = new PdfDocument();
// Read a pdf file
doc.LoadFromFile(input);
// Get the first page
PdfPageBase page = doc.Pages[0];
// Extract text from page keeping white space
String text = page.ExtractText(true);
// Extract text from page without keeping white space
//String text = page.ExtractText(false);
String result = Path.GetFullPath("ExtractTextFromParticularPage_out.txt");
// Create a writer to put the extracted text
TextWriter tw = new StreamWriter(result);
// Write a line of text to the file
tw.WriteLine(text);
// Close the stream
tw.Close();
MessageBox.Show("\nText extracted successfully from particular pages of PDF Document.\nFile saved at " + result);
}
static void Main(string[] args)
{
//Load the PDF document
PdfLoadedDocument loadedDocument = new PdfLoadedDocument("../../../../../../Data/Invoice.pdf");
// Get the first page of the loaded PDF document
PdfPageBase page = loadedDocument.Pages[0];
TextLines lineCollection = new TextLines();
// Extract text from the first page with bounds
page.ExtractText(out lineCollection);
RectangleF textBounds = new RectangleF(474, 161, 50, 9);
string invoiceNumer = "";
//Get the text provided in the bounds
foreach (TextLine txtLine in lineCollection)
{
foreach (TextWord word in txtLine.WordCollection)
{
if (textBounds.IntersectsWith(word.Bounds))
{
invoiceNumer = word.Text;
break;
}
}
}
//Close the PDF document
loadedDocument.Close(true);
File.WriteAllText("data.txt", invoiceNumer);
}
private void button1_Click(object sender, EventArgs e)
{
string input = @"..\..\..\..\..\..\Data\ExtractTextFromSpecificArea.pdf";
//Load the PDF file
PdfDocument pdf = new PdfDocument();
pdf.LoadFromFile(input);
//Get the first page
PdfPageBase page = pdf.Pages[0];
//Extract text from a specific rectangular area within the page
string text = page.ExtractText(new RectangleF(80, 180, 500, 200));
//Save the text to a .txt file
StringBuilder sb = new StringBuilder();
sb.AppendLine(text);
string result = "ExtractText_result.txt";
File.WriteAllText(result, sb.ToString());
Viewer(result);
}
private void button1_Click(object sender, EventArgs e)
{
//Create a pdf document
PdfDocument doc = new PdfDocument();
//Load a pdf file
doc.LoadFromFile(@"..\..\..\..\..\..\Data\ExtractHighlightedText.pdf");
PdfPageBase page = doc.Pages[0];
PdfTextMarkupAnnotationWidget textMarkupAnnotation;
StringBuilder stringBuilder = new StringBuilder();
stringBuilder.AppendLine("Extracted hightlighted text:");
//Get PdfTextMarkupAnnotationWidget objects
for (int i = 0; i < page.AnnotationsWidget.Count; i++)
{
if (page.AnnotationsWidget[i] is PdfTextMarkupAnnotationWidget)
{
textMarkupAnnotation = page.AnnotationsWidget[i] as PdfTextMarkupAnnotationWidget;
//Get the highlighted text
stringBuilder.AppendLine(page.ExtractText(textMarkupAnnotation.Bounds));
//Get the highlighted color
Color color = textMarkupAnnotation.TextMarkupColor;
}
}
String result = "ExtractHighlightedText.txt";
File.WriteAllText(result, stringBuilder.ToString());
DocumentViewer(result);
}
static void Main(string[] args)
{
//Load the PDF document
PdfLoadedDocument loadedDocument = new PdfLoadedDocument("../../../../../../Data/Invoice.pdf");
//Get the first page of the loaded PDF document
PdfPageBase page = loadedDocument.Pages[0];
//Extract text with layout
string extractedText = page.ExtractText(true);
//Save text to file
File.WriteAllText("data.txt", extractedText);
//Close the PDF document
loadedDocument.Close(true);
}
private void getPDFMsg()
{
PdfDocument pdf = new PdfDocument();
pdf.LoadFromFile(pdfFileName[currentPageNum - 1]);
PdfPageBase page = pdf.Pages[0];
//从第一页的指定矩形区域内提取文本
string text = page.ExtractText(new RectangleF((int)(x.X - 70), (int)(x.Y - 30), (int)(y.X - 80), (int)(y.Y - 135)));
//string text = page.ExtractText(new RectangleF(50, 50, 100, 100));
StringBuilder sb = new StringBuilder();
sb.AppendLine(text);
File.WriteAllText("Extract.txt", sb.ToString().Replace("Evaluation Warning : The document was created with Spire.PDF for .NET.", ""));
Process.Start("Extract.txt");
}
public static async Task <string> Read(StorageFile openFile)
{
SaveFilePdf.openFile = openFile;
string extractedText = "";
if (openFile != null)
{
PdfLoadedDocument loadedDocument = new PdfLoadedDocument();
await loadedDocument.OpenAsync(openFile).ConfigureAwait(true);
PdfPageBase page = loadedDocument.Pages[0];
extractedText = page.ExtractText();
loadedDocument.Close(true);
loadedDocument.Dispose();
var mru = Windows.Storage.AccessCache.StorageApplicationPermissions.MostRecentlyUsedList;
string mruToken = mru.Add(openFile, "Pdf file");
}
return(extractedText);
}
static void Main(string[] args)
{
FileStream inputStream = new FileStream("../../../../../../../Data/Invoice.pdf", FileMode.Open);
//Load the PDF document
PdfLoadedDocument loadedDocument = new PdfLoadedDocument(inputStream);
// Get the first page of the loaded PDF document
PdfPageBase page = loadedDocument.Pages[0];
// Extract text from the first page with bounds
string extractedText = page.ExtractText();
//Close the document
loadedDocument.Close(true);
//Save the text to file
File.WriteAllText("data.txt", extractedText);
}
public string StripPDF(PDF filename)
{
//Load an existing PDF.
PdfLoadedDocument loadedDocument = new PdfLoadedDocument(filename.Path);
//Load the first page.
PdfPageBase page = loadedDocument.Pages[0];
//Extract text from first page.
string extractedText = page.ExtractText();
//Close the document
loadedDocument.Close(true);
return(extractedText);
}
public static void GetElements(string fileName)
{
try
{
PdfDocument doc = new PdfDocument();
doc.LoadFromFile(fileName);
PdfPageBase page = doc.Pages[0];
SimpleTextExtractionStrategy strategy = new SimpleTextExtractionStrategy();
string text = page.ExtractText(strategy);
FileStream fs = new FileStream(Path.GetDirectoryName(fileName) + "\\result_spire.txt", FileMode.Create);
StreamWriter sw = new StreamWriter(fs);
sw.Write(text);
sw.Flush();
sw.Close();
}
catch (Exception e)
{
Console.WriteLine(e.Message);
}
}
private void button2_Click(object sender, EventArgs e)
{
if (fileDialog.FileName.Length > 0)
{
PdfDocument doc = new PdfDocument();
doc.LoadFromFile(fileDialog.FileName);
PdfPageBase page = doc.Pages[2];
SimpleTextExtractionStrategy strategy = new SimpleTextExtractionStrategy();
string text = page.ExtractText(strategy);
FileStream fs = new FileStream("Result_PDF.txt", FileMode.Create);
StreamWriter sw = new StreamWriter(fs);
sw.Write(text);
sw.Flush();
sw.Close();
string textValue = System.IO.File.ReadAllText("Result_PDF.txt");
textBox1.Text = textValue;
}
else
{
MessageBox.Show("PDF 파일을 선택해 주세요.");
}
}
private void button1_Click(object sender, EventArgs e)
{
var fileContent = string.Empty;
var filePath = string.Empty;
using (OpenFileDialog openFileDialog = new OpenFileDialog())
{
this.statusBox.Text += Environment.NewLine + "Selecting file..." + Environment.NewLine;
openFileDialog.InitialDirectory = "c:\\";
openFileDialog.Filter = "pdf files (*.pdf)|*.pdf|All files (*.*)|*.*";
openFileDialog.FilterIndex = 2;
openFileDialog.RestoreDirectory = true;
if (openFileDialog.ShowDialog() == DialogResult.OK)
{
this.messageBox.Text = " ";
//Get the path of specified file
filePath = openFileDialog.FileName;
//Read the contents of the file into a stream
var fileStream = openFileDialog.OpenFile();
using (StreamReader reader = new StreamReader(fileStream))
{
fileContent = reader.ReadToEnd();
}
}
}
this.statusBox.Text += "Processing PDF file..." + Environment.NewLine;
this.fileName.Text = filePath;
PdfDocument doc = new PdfDocument();
doc.LoadFromFile(filePath);
PdfPageBase page = doc.Pages[0];
SimpleTextExtractionStrategy strategy = new SimpleTextExtractionStrategy();
string text = page.ExtractText(strategy);
//PdfTextFind[] results = null;
//results = page.FindText("MS61").Finds;
//foreach (PdfTextFind texts in results)
//{
// PointF p = texts.Position;
// float x = texts.Position.X;
// float y = texts.Position.Y;
// // this.displayText.Text += p + Environment.NewLine;
// string textArea = page.ExtractText(new RectangleF(x-30, y, 80, 180));
// // this.displayText.Text += Environment.NewLine + textArea + Environment.NewLine;
//}
//MessageBox.Show(text, " ", MessageBoxButtons.OK);
//Document pdfDocument = new Document(filePath);
//TextAbsorber textAbsorber = new TextAbsorber();
//pdfDocument.Pages.Accept(textAbsorber);
//string extractedText = textAbsorber.Text;
//Console.WriteLine(extractedText);
//MessageBox.Show(extractedText, "Upstate Gold Sheet", MessageBoxButtons.OK);
//PdfDocument PDF = PdfDocument.FromFile(filePath);
//string AllText = PDF.ExtractAllText();
string[] stringArray = text.Split(Environment.NewLine);
//string[] stringArray = text.Split("$");
this.statusBox.Text += "API Request/Response..." + Environment.NewLine;
foreach (string line in stringArray)
{
bool stringExists = line.Contains("$");
if (stringExists)
{
string[] lineSplit = line.Split(' ');
// foreach ( string word in lineSplit)
// {
// this.displayText.Text += word + Environment.NewLine;
// }
// this.displayText.Text += line + Environment.NewLine;
// this.displayText.Text += " = " + lineSplit[3] + Environment.NewLine;
string requestString = "method=test&requestString=" + line;
this.displayText.Text += "--> " + lineSplit[0] + " " + lineSplit[1];
string responseString = APIRequest(requestString);
string[] responseArray = responseString.Split(':');
responseArray[1] = responseArray[1].Replace("}", string.Empty);
// MessageBox.Show(responseArray[1], "api response");
if (responseArray[1] == "0")
{
this.displayText.Text += ": Success" + Environment.NewLine;
}
else
{
// this.displayText.ForeColor
this.displayText.Text += ": Failure" + Environment.NewLine;
}
// this.displayText.Text += responseString + Environment.NewLine;
}
}
this.messageBox.Text = "Click the 'Exit' button to end the application.";
this.statusBox.Text += "Job End...";
// MessageBox.Show(fileContent, "Unconverted Content " + filePath, MessageBoxButtons.OK);
// MessageBox.Show(AllText, "Upstate Gold Sheet", MessageBoxButtons.OK);
// MessageBox.Show(fileContent, "File Content at path: " + filePath, MessageBoxButtons.OK);
// Show the dialog and get result.
//DialogResult result = openFileDialog1.ShowDialog();
//if (result == DialogResult.OK) // Test result.
//{
//}
// Console.WriteLine(result); // <-- For debugging use.
}
public ChartData GeneratePDFReport(PdfLoadedPageCollection pages)
{
try
{
StringBuilder extractedText = new StringBuilder();
if (pages != null)
{
//Parallel.ForEach<PdfPageBase>(pages.AsParallel(), page =>
//{
//});
foreach (PdfPageBase pageT in pages)
{
PdfPageBase page = pageT;
extractedText.Append(page.ExtractText());
}
}
var wordsCount = System.Text.RegularExpressions.Regex.Matches(extractedText.ToString(), "\\S+").Count;
var sentences = extractedText.ToString().Split(new string[] { ". ", "\r\n\\" }, StringSplitOptions.None);
var sentenceReport = (from sentence in sentences
where sentence != string.Empty
group sentence by sentence into tempBag
//let count = tempBag.Count()
//orderby count descending
select new { Value = tempBag.Key, Length = tempBag.Key.Length }
).ToList();
chartObj.NumberOfSentences = sentenceReport.Distinct().Count();
if (chartObj.NumberOfSentences > 0)
{
chartObj.AvgSetenceLength = sentenceReport.Sum(x => x.Length) / chartObj.NumberOfSentences;
}
string[] source = extractedText.ToString().Split(new char[] { '.', '?', '!', ' ', ';', ':', ',', '_' }, StringSplitOptions.RemoveEmptyEntries);
var matchQuery = from word in source
where word.ToLowerInvariant() == "\r\n".ToLowerInvariant()
select word;
chartObj.ParagraphCount = matchQuery.Count();
var wordReport = (from word in source
where word != string.Empty
group word by word into tempBag
let count = tempBag.Count()
orderby count descending
select new { Value = tempBag.Key, Count = count, Length = tempBag.Key.Length }
).ToList();
chartObj.TotalUniqueWords = wordReport.Count();
if (chartObj.TotalUniqueWords > 0)
{
chartObj.AverageWordLength = wordReport.Sum(x => x.Length) / chartObj.TotalUniqueWords;
}
var topTenOccuringWords = (from obj in wordReport.OrderByDescending(x => x.Count).ToList().Take(10)
where obj.Value != "\r\n"
select new ChartData
{
Word = obj.Value,
Count = obj.Count
}).ToList();
chartObj.Top10WordsFromPDFLoaded = new ObservableCollection <ChartData>(topTenOccuringWords);
chartObj.ListOfDetailsToPrint = GetOtherPDFReportData(chartObj);
}
catch (Exception ex)
{
throw new Exception("Error while generating PDF Reports in PdfViewer.ViewModels.PDFViewModel.GeneratePDFReport", ex);
}
return(chartObj);
}
