View PDF files in C# using the Xpdf and muPDF library, Print PostScript.

14:24 19 Jan '10

Hi, have you fixed the problem? It sound's like an issue with xpdf, or maybe I'm missing something, I will to take a look this weekend.
Sign In·View Thread·PermaLink

Page Width e Height

Member 2720133

4:52 2 Dec '09

In this project
ASP.NET PDF Viewer User Control Without Acrobat Reader Installed on Client or Server
your library is used to convert PDF to jpg on the fly.

This function


Public Shared Function GetOptimalDPI(ByVal filename As String, ByVal pageNumber As Integer, ByRef oSize As Drawing.Size) As Integer
    GetOptimalDPI = 0
    Dim pdfDoc As New PDFLibNet.PDFWrapper
    pdfDoc.LoadPDF(filename)
    If pdfDoc IsNot Nothing Then
      If pdfDoc.Pages(pageNumber).Width > 0 And pdfDoc.Pages(pageNumber).Height > 0 Then
        Dim picHeight As Integer = oSize.Height
        Dim picWidth As Integer = oSize.Width
        Dim docHeight As Integer = pdfDoc.Pages(pageNumber).Height
        Dim docWidth As Integer = pdfDoc.Pages(pageNumber).Width
        Dim HScale As Single = oSize.Width / docWidth
        Dim VScale As Single = oSize.Height / docHeight
        If VScale > HScale Then
          GetOptimalDPI = Math.Floor(253 * HScale)
        Else
          GetOptimalDPI = Math.Floor(253 * VScale)
        End If
      End If
      pdfDoc.Dispose()
    End If
  End Function

is used to calculate to optimal DPI.

I'm testing with this pdf http://www.sygest.it/upgrade/PdfViewASP/smi.pdf[^]

but I think that there is a problem with rotation.
pdfDoc.Pages(pageNumber).Width and pdfDoc.Pages(pageNumber).Height return right information but how can i know if the W/H are inverted (rotated)?
In the PDF page 1 i get : W:2099 H2970, for page 14 (rotated) i get : H4202 , W 2970.

Is there another way to get that info ?
Thank you

12:59 2 Dec '09

Yes, there is another way, I will to upload tonight the property exposed into PDFWrapper::Page
Sign In·View Thread·PermaLink

Member 2720133

3:00 3 Dec '09

oh thanks! can U please tell me whem can i download it ?
Sign In·View Thread·PermaLink

17:59 3 Dec '09

I have uploaded an update to the SVN. Check the property PDFWrapper::Page::Rotation, you can get the "real" page size from PDFWrapper:Page:Width/Height.
Sign In·View Thread·PermaLink

Getting started with source code

Member 2720133

21:15 3 Dec '09

thank you, but i haven't VC++ installed. Where can i get compiled binaries ?
Sign In·View Thread·PermaLink

bugal

3:51 1 Dec '09

Hello, Sorry for the trouble , but could you give me some guide lines on how to run version on my personal project C# .Net I struggle on understanding how to and what to download and came out with no results Thanks Best regards Eyal
Sign In·View Thread·PermaLink

Re: Getting started with source code

Extract Text and Fonts from Pdf [modified]

11:00 3 Dec '09

You need to add a reference to PDFLibNet.dll, see PDFViewer project for an example. PDFViewer has a control called PageViewer that can be used as the canvas. You need to trap PaintControl, NextPage and PreviousPage.

 PDFLibNet.PDFWrapper _pdfDoc=null;
        public frmPDFViewer()
        {
            InitializeComponent();
            pageViewer1.NextPage+=new PDFViewer.PageViewer.MovePageHandler(pageViewer1_NextPage);
            pageViewer1.PreviousPage+=new PDFViewer.PageViewer.MovePageHandler(pageViewer1_PreviousPage);
            pageViewer1.PaintControl+=new PDFViewer.PageViewer.PaintControlHandler(pageViewer1_PaintControl);
        }

        private void button1_Click(object sender, EventArgs e)
        {
            if (openFileDialog1.ShowDialog() == DialogResult.OK)
            {
                if (_pdfDoc != null)
                    _pdfDoc.Dispose();
                _pdfDoc = new PDFLibNet.PDFWrapper();
                //_pdfDoc.UseMuPDF = true;
                _pdfDoc.LoadPDF(openFileDialog1.FileName);
                _pdfDoc.FitToWidth(pageViewer1.Handle);
                _pdfDoc.RenderPage(pageViewer1.Handle);
                pageViewer1.PageSize = new Size(_pdfDoc.PageWidth, _pdfDoc.PageHeight);
            }
        }

        private void pageViewer1_PaintControl(object sender, Rectangle view, Point location, Graphics g)
        {
            if (_pdfDoc != null)
            {
                Size sF = new Size(view.Right, view.Bottom);
                Rectangle r = new Rectangle(location, sF);
                pageViewer1.PageSize = new Size(_pdfDoc.PageWidth, _pdfDoc.PageHeight);
                _pdfDoc.ClientBounds = r;
                _pdfDoc.CurrentX = view.X;
                _pdfDoc.CurrentY = view.Y;
                _pdfDoc.DrawPageHDC(g.GetHdc());
                g.ReleaseHdc();
            }
        }

        private bool pageViewer1_NextPage(object sender)
        {
            if (_pdfDoc != null && _pdfDoc.CurrentPage < _pdfDoc.PageCount)
            {
                _pdfDoc.NextPage();
                _pdfDoc.RenderPage(pageViewer1.Handle);
                return true;
            }
            return false; //No more pages
        }

        private bool pageViewer1_PreviousPage(object sender)
        {
            if (_pdfDoc != null && _pdfDoc.CurrentPage > 1)
            {
                _pdfDoc.PreviousPage();
                _pdfDoc.RenderPage(pageViewer1.Handle);
                return true;
            }
            return false; //No more pages
        }

        private void Form1_Load(object sender, EventArgs e)
        {
            pageViewer1.PageSize = new Size(pageViewer1.Width, (int)(pageViewer1.Width * 11 / 8.5));
            pageViewer1.Visible = true;
        }

that's all. You have a PDF viewer.
You can change between muPDF and xpdf setting _pdfDoc.UseMuPDF = true/false;

bharatmanwani

20:12 25 Nov '09

Is it possible to extract text and font information as well as the position on the pdf page using pdflibnet.dll? modified on Thursday, November 26, 2009 1:52 AM
Sign In·View Thread·PermaLink	5.00/5 (1 vote)

Re: Extract Text and Fonts from Pdf

build the C++ assembly with /clr:safe

5:34 26 Nov '09

Yes of course, it is on my To-Do List. For text extraction you can Use PDFWrapper::Pages[int]::Text or PDFWrapper::Pages[int]::ExtractText(Rectangle).
Sign In·View Thread·PermaLink

Unruled Boy

17:15 17 Nov '09

please, build the C++ assembly with /clr:safe, if possible, because:

it's ok like this:

Assembly a = Assembly.LoadFile(@"PDFLibNet.dll");

but exception like this:
byte[] b = File.ReadAllBytes(@"PDFLibNet.dll");
Assembly a = Assembly.Load(b);

throwing FileLoadException, telling me it could not verify the code or could not pass policy check. .net thinks that it is not a valid assembly.

Regards,
unruledboy_at_gmail_dot_com
http://www.xnlab.com

Re: build the C++ assembly with /clr:safe

simple way to export to jpg? [modified]

5:33 26 Nov '09

There is no way to compile it with /clr:safe because we are using unsafe pointers, Sorry
Sign In·View Thread·PermaLink

Unruled Boy

17:26 16 Nov '09

I want a simple way to export pdf to image files, png will be the best, and I do NOT want a dummy form to hold as a handle. here is my code:

            if(_pdfDoc.LoadPDF(filename))
            {
            for (int i = 0; i < _pdfDoc.PageCount; i++)
            {
                _pdfDoc.CurrentPage = i + 1;
                Bitmap b = new Bitmap(600, 800);
                using (System.Drawing.Graphics g = Graphics.FromImage(b))
                {
                    //method 1: some pages are correctly rendered, some are duplicated. multithreaded problem?
                    //_pdfDoc.RenderPage(FooForm.Handle, true);
                    //_pdfDoc.ExportJpg(target + "page" + (i + 1) + ".jpg", 90);

                    //method 2: all empty pages
                    _pdfDoc.RenderPage(g.GetHdc(), true);
                    b.Save(target + "page" + (i + 1) + ".jpg", System.Drawing.Imaging.ImageFormat.Png);
                }
            }
            }

Regards,
unruledboy_at_gmail_dot_com
http://www.xnlab.com

modified on Tuesday, November 17, 2009 3:12 AM

Re: simple way to export to jpg?

Re: simple way to export to jpg? [modified]

5:05 17 Nov '09

Yes, it is a multithread problem.
RenderPage and ExportJpg uses the same instance for render the page.
You must not call RenderPage before of ExportJpg.
if you are using the viewer at the same time you need check if the thread is Busy (_pdfDoc.IsBusy) before calling to ExportJpg.
_pdfDoc.RenderPage(hdc,false) the last param disable the pre-render of the next page.
In words of performance the first method is faster, but you need to remove "RenderPage".

ExportJpg can export a range of pages:

if(_pdfDoc != null && _pdfDoc.IsJpgBusy)
    if(MessageBox.Show("It is another pending export process, it will cancelled. Continue?")==No)
        return;

_pdfDoc.ExportJpgProgress += ....;
_pdfDoc.ExportJpgFinished += ....;
if(_pdfDoc.LoadPDF(filename))
    _pdfDoc.ExportJpg(target + "page%d.jpg",1,_pdfDoc.PageCount,150,90);

That's all for jpg export.

Another method is using _pdfDoc.Pages[0].GetBitmap(....).

Regards.

Unruled Boy

15:11 17 Nov '09

hi, it crashed after exported the first 3 pages with using() or explicit _pdfDoc.Dispose()

without using() or _pdfDoc.Dispose(), it will be fine.

btw, how to make it synchronized, I want it to finish before doing next task.

and, the following method is really slow, a lot slow than the ExportJpg way, how to make it faster?

                    Bitmap b = _pdfDoc.Pages[i + 1].GetBitmap(150, true);
                    b.Save(target + "page" + (i + 1) + ".png", System.Drawing.Imaging.ImageFormat.Png);

Regards,
unruledboy_at_gmail_dot_com
http://www.xnlab.com

modified on Tuesday, November 17, 2009 8:24 PM

Re: simple way to export to jpg? [modified]

5:28 26 Nov '09

Hi I has been outside of the country. I will to check the memory leak. About the syncronized I will to add a "WaitTime" parameter to ExportJpg. I tell you how is going this weekend.
The exportJpg method is faster because it converts the DIB directly to JPEG using libjpeg.
GetBitmap Converts the DIB to a Windows Bitmap and then you convert it to PNG using GDI+ that is slower than any other library. I will to implement libpng later.

modified on Thursday, November 26, 2009 10:38 AM

Now another issue

11:03 8 Nov '09

The logic of what I'm doing with the .dll isn't entirely clear, but your short code example seemed to be enough to get me going:
PictureBox pic =new PictureBox();
pic.Width=800;
pic.Height=1024;
_pdfDoc.FitToWidth(pic.Handle);
pic.Height = _pdfDoc.PageHeight;
_pdfDoc.RenderPage(pic.Handle);

Bitmap _backbuffer = new Bitmap(_pdfDoc.PageWidth, _pdfDoc.PageHeight);
using (Graphics g = Graphics.FromImage(_backbuffer))
{
_pdfDoc.RenderHDC(g.GetHdc);
g.ReleaseHdc();
}
pic.Image = _backbuffer;

But in the latest version there is no .RenderHDC. I've tried .DrawPageHDC but nothing happens.

Where do I go to find more information on how to use your .dll to make sure I'm doing the right thing?

Sorry I'm such a bother!

Steven

Re: Now another issue

4:54 9 Nov '09

Sorry, I will to update the article to the new example. The library was updated to avoid flickering on render.

PictureBox pic =new PictureBox();
pic.Width=800;
pic.Height=1024;
_pdfDoc.FitToWidth(pic.Handle);
pic.Height = _pdfDoc.PageHeight;
_pdfDoc.RenderPage(pic.Handle);
/*Added since 1.0.6.2*/
_pdfDoc.CurrentX = 0;
_pdfDoc.CurrentY = 0;
_pdfDoc.ClientBounds = new Rectangle(0, 0, _pdfDoc.PageWidth, _pdfDoc.PageHeight);

Bitmap _backbuffer = new Bitmap(_pdfDoc.PageWidth, _pdfDoc.PageHeight);
using (Graphics g = Graphics.FromImage(_backbuffer))
{
/*New thread safe method*/
_pdfDoc.DrawPageHDC(g.GetHdc());
g.ReleaseHdc();
}
pic.Image = _backbuffer;

The new version also has a new method:
Bitmap bmp = _pdfDoc.Pages[0].GetBitmap(int width, int height, bool antialias);
Bitmap bmp = _pdfDoc.Pages[0].GetBitmap(double renderDPI);

But the only thread safe is RenderPage and DrawPageHDC.

Re: Now another issue

5:47 9 Nov '09

Brilliant! I now have a fully-working PDF viewer and I look forward to playing with it properly. Thanks for your help in sorting out my problems. Steven
Sign In·View Thread·PermaLink

I'm confused/stupid

8:55 7 Nov '09

Having looked around some more and visited both the XPDF and MuPDF sites, I'm even more confused.

What, if anything, do I need to have in order for PDFLibNet.dll to work?

I saw a whole bunch of stuff in one of Ron Schuler's articles with the same error that I'm currently getting "Could not load PDFLibNet.dll or one of its assemblies" but was not clear what the solution was. Should I (as Ron's article implies) have a whole bunch of other .dlls or are you simplifying the whole thing?

Do I have to have 5MB (unzipped) of MuPDF.exe to be shipped with my programs if I'm using the 2MB PDFLibNet.dll or are you somehow encapsulating all the magic in one .dll? Whichever parts of MuPDF or XPDF I put into the folder with PDFViewer.exe, I still get the same error above.

I think you're assuming that everyone is as clever/knowledgeable as you are, but some of use are simple folk who would love a simple .dll which allows us to read pdfs in our C# programs. Your article seems to suggest that, but it's unclear.

Thanks

Steven

Re: I'm confused/stupid

6:13 8 Nov '09

Hi Steven. I think that you have reason I have suposed that every one has worked with VC++.
The library is builded around C++\CLI, so you need the release version of the CRT dll's:

For the version compiled under VS2005:
msvcr80.dll
msvcm80.dll
For the version compiled under VS2008:
msvcr90.dll
msvcm90.dll

If you does not have VC++ installed you need to download the redistributable package:
VC++ 2005 Redistributable[^]
VC++ 2008 Redistributable package[^]

I have uploaded two versions to google, one compiled under VC++2005 and other compiled under VC++ 2008.
You dont need any other library to make it work, everything is encapsulated in PDFLibNet.dll.
Just xPDF needs some Language support packages[^], I'm working on embbed those files into one dll.

The size of the library is 3MB in VC++2005 and 3.6MB in VC++2008.
MuPDF already support chinnese, japanese and korean.

Another problem is caused because the PDFLibNet is not registered in the GAC (confirmed in Windows Server 2003 and ASP.NET).

Re: I'm confused/stupid