public static void ConvertPdfToTextInThread()
{
string pdfs = @"..\..\..\..\..\";
string[] files = Directory.GetFiles(pdfs, "*.pdf");
List <Thread> threads = new List <Thread>();
for (int i = 0; i < files.Length; i++)
{
TArgument targ = new TArgument()
{
PdfFile = files[i],
PageNumber = 1
};
var t = new Thread((a) => ConvertToText(a));
t.Start(targ);
threads.Add(t);
}
foreach (var thread in threads)
{
thread.Join();
}
Console.WriteLine("Done.");
Console.ReadLine();
}
public static void ConvertToExcel(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string excelFile = Path.GetFileNameWithoutExtension(pdfFile) + ".xls";
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
// 'true' = Convert all data to spreadsheet (tabular and even textual).
// 'false' = Skip textual data and convert only tabular (tables) data.
f.ExcelOptions.ConvertNonTabularDataToSpreadsheet = true;
// 'true' = Preserve original page layout.
// 'false' = Place tables before text.
f.ExcelOptions.PreservePageLayout = true;
// The information includes the names for the culture, the writing system,
// the calendar used, the sort order of strings, and formatting for dates and numbers.
System.Globalization.CultureInfo ci = new System.Globalization.CultureInfo("en-US");
ci.NumberFormat.NumberDecimalSeparator = ",";
ci.NumberFormat.NumberGroupSeparator = ".";
f.ExcelOptions.CultureInfo = ci;
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToExcel(excelFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
System.Diagnostics.Process.Start(new System.Diagnostics.ProcessStartInfo(excelFile)
{
UseShellExecute = true
});
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertToExcel(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string excelFile = Path.ChangeExtension(pdfFile, ".xls");
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
// 'true' = Convert all data to spreadsheet (tabular and even textual).
// 'false' = Skip textual data and convert only tabular (tables) data.
f.ExcelOptions.ConvertNonTabularDataToSpreadsheet = true;
// 'true' = Preserve original page layout.
// 'false' = Place tables before text.
f.ExcelOptions.PreservePageLayout = true;
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToExcel(excelFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public Handler(ICollection <Task> tasks,
TArgument arg,
Action <Handler> doneCallback,
CancellableSignal overrideSignal)
{
this.tasks = new Task[tasks.Count];
int n = 0;
foreach (var task in tasks)
{
this.tasks[n++] = task;
}
this.arg = arg;
this.doneCallback = doneCallback;
this.signal = null != overrideSignal ? overrideSignal : new CancellableSignal();
}
public static void ConvertToWord(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string docxFile = Path.GetFileNameWithoutExtension(pdfFile) + ".docx";
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
f.WordOptions.Format = PdfFocus.CWordOptions.eWordDocument.Docx;
f.WordOptions.RenderMode = PdfFocus.CWordOptions.eRenderMode.Flowing;
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToWord(docxFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
System.Diagnostics.Process.Start(new System.Diagnostics.ProcessStartInfo(docxFile)
{
UseShellExecute = true
});
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertToPng(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string pngFile = Path.GetFileNameWithoutExtension(pdfFile) + ".png";
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
f.ImageOptions.ImageFormat = System.Drawing.Imaging.ImageFormat.Png;
f.ImageOptions.Dpi = 300;
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToImage(pngFile, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
System.Diagnostics.Process.Start(new System.Diagnostics.ProcessStartInfo(pngFile)
{
UseShellExecute = true
});
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertToXml(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string xmlFile = Path.GetFileNameWithoutExtension(pdfFile) + ".xml";
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
// Let's convert all data (textual and tabular) to XML.
f.XmlOptions.ConvertNonTabularDataToSpreadsheet = true;
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToXml(xmlFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
System.Diagnostics.Process.Start(new System.Diagnostics.ProcessStartInfo(xmlFile)
{
UseShellExecute = true
});
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertToHtml(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string htmlFile = targum.HtmlFile;
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
f.EmbeddedImagesFormat = PdfFocus.eImageFormat.Auto;
f.HtmlOptions.IncludeImageInHtml = false;
f.HtmlOptions.ImageSubFolder = String.Format("{0}_images", Path.GetFileNameWithoutExtension(pdfFile));
f.HtmlOptions.Title = String.Format("This document was produced from {0}.", Path.GetFileName(pdfFile));
f.HtmlOptions.ImageFileName = "picture";
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToHtml(htmlFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertToText(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string textFile = Path.GetFileNameWithoutExtension(pdfFile) + ".txt";
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
f.WordOptions.ShowInvisibleText = true;
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToText(textFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
System.Diagnostics.Process.Start(new System.Diagnostics.ProcessStartInfo(textFile)
{
UseShellExecute = true
});
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertToHtml(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string htmlFile = Path.ChangeExtension(pdfFile, ".html");
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
f.HtmlOptions.ImageType = PdfFocus.CHtmlOptions.eHtmlImageType.Png;
f.HtmlOptions.IncludeImageInHtml = false;
f.HtmlOptions.ImageSubFolder = String.Format("{0}_images", Path.GetFileNameWithoutExtension(pdfFile));
f.HtmlOptions.ImageFileName = "pict";
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToHtml(htmlFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertToText(object targ)
{
TArgument targum = (TArgument)targ;
string pdfFile = targum.PdfFile;
int page = targum.PageNumber;
string textFile = Path.ChangeExtension(pdfFile, ".txt");
SautinSoft.PdfFocus f = new SautinSoft.PdfFocus();
f.WordOptions.ShowInvisibleText = true;
f.OpenPdf(pdfFile);
bool done = false;
if (f.PageCount > 0)
{
if (page >= f.PageCount)
{
page = 1;
}
if (f.ToText(textFile, page, page) == 0)
{
done = true;
}
f.ClosePdf();
}
if (done)
{
Console.WriteLine("{0}\t - Done!", Path.GetFileName(pdfFile));
}
else
{
Console.WriteLine("{0}\t - Error!", Path.GetFileName(pdfFile));
}
}
public static void ConvertPdfToHtmlInThread()
{
string pdfDir = @"..\..\";
string[] pdfFiles = Directory.GetFiles(pdfDir, "*.pdf");
DirectoryInfo htmlDir = new DirectoryInfo("HTML results");
if (!htmlDir.Exists)
{
htmlDir.Create();
}
List <Thread> threads = new List <Thread>();
foreach (string pdfFile in pdfFiles)
{
TArgument targ = new TArgument()
{
PdfFile = pdfFile,
HtmlFile = Path.Combine(htmlDir.FullName, Path.GetFileNameWithoutExtension(pdfFile) + ".html"),
PageNumber = 1
};
var t = new Thread((a) => ConvertToHtml(a));
t.Start(targ);
threads.Add(t);
}
foreach (var thread in threads)
{
thread.Join();
}
Console.WriteLine("Done!");
// Open the result for demonstration purposes.
System.Diagnostics.Process.Start(new System.Diagnostics.ProcessStartInfo(htmlDir.FullName)
{
UseShellExecute = true
});
}
