Arabic in pdf using iTextSharp in c#

NOBLE M.O. picture NOBLE M.O. · Dec 30, 2015 · Viewed 11.4k times · Source

I want to create a PDF file with Arabic text content in C#. I'm using iTextSharp to create this. I followed the instruction in http://geekswithblogs.net/JaydPage/archive/2011/11/02/using-itextsharp-to-correctly-display-hebrew--arabic-text-right.aspx. I want to insert the following Arabic sentence in pdf.

تم إبرام هذا العقد في هذا اليوم [●] م الموافق [●] من قبل وبين .

The [●] need to be replaced by dynamic English words. I tried to implement this by using ARIALUNI.TTF [This tutorial link suggested it]. The code is given below.

public void WriteDocument()
{
    //Declare a itextSharp document 
    Document document = new Document(PageSize.A4);

    //Create our file stream and bind the writer to the document and the stream 
    PdfWriter writer = PdfWriter.GetInstance(document, new FileStream(@"D:\Test.Pdf", FileMode.Create));

    //Open the document for writing 
    document.Open();

    //Add a new page 
    document.NewPage();

    //Reference a Unicode font to be sure that the symbols are present. 
    BaseFont bfArialUniCode = BaseFont.CreateFont(@"D:\ARIALUNI.TTF", BaseFont.IDENTITY_H, BaseFont.EMBEDDED);
    //Create a font from the base font 
    Font font = new Font(bfArialUniCode, 12);

    //Use a table so that we can set the text direction 
    PdfPTable table = new PdfPTable(1);
    //Ensure that wrapping is on, otherwise Right to Left text will not display 
    table.DefaultCell.NoWrap = false;

    //Create a regex expression to detect hebrew or arabic code points 
    const string regex_match_arabic_hebrew = @"[\u0600-\u06FF,\u0590-\u05FF]+";
    if (Regex.IsMatch("م الموافق", regex_match_arabic_hebrew, RegexOptions.IgnoreCase))
    {
        table.RunDirection = PdfWriter.RUN_DIRECTION_RTL;
    }

    //Create a cell and add text to it 
    PdfPCell text = new PdfPCell(new Phrase(" : "+"من قبل وبين" + " 2007 " + "م الموافق" + " dsdsdsdsds " + "تم إبرام هذا العقد في هذا اليوم ", font));
    //Ensure that wrapping is on, otherwise Right to Left text will not display 
    text.NoWrap = false;

    //Add the cell to the table 
    table.AddCell(text);

    //Add the table to the document 
    document.Add(table);

    //Close the document 
    document.Close();

    //Launch the document if you have a file association set for PDF's 
    Process AcrobatReader = new Process();
    AcrobatReader.StartInfo.FileName = @"D:\Test.Pdf";
    AcrobatReader.Start();
}

While calling this function, I got a PDF with some Unicode as given below.

اذه يف دقعلا اذه ماربإ مت dsdsdsdsds قفاوملا م 2007 نيبو لبق نم مويلا

It is not matching with our hard coded Arabic sentence. Is this a issue of font? Please help me or suggest me any other method to implement the same.

Answer

blagae picture blagae · Dec 30, 2015

@csharpcoder has the right idea, but his execution is off. He doesn't add the cell to a table, and the table doesn't end up in the document.

void Go()
{
    Document doc = new Document(PageSize.LETTER);
    string yourPath = "foo/bar/baz.pdf";
    using (FileStream os = new FileStream(yourPath, FileMode.Create))
    {
        PdfWriter.GetInstance(doc, os); // you don't need the return value

        doc.Open();

        string fontLoc = @"c:\windows\fonts\arialuni.ttf"; // make sure to have the correct path to the font file
        BaseFont bf = BaseFont.CreateFont(fontLoc, BaseFont.IDENTITY_H, BaseFont.EMBEDDED);
        Font f = new Font(bf, 12);

        PdfPTable table = new PdfPTable(1); // a table with 1 cell
        Phrase text = new Phrase("العقد", f);
        PdfPCell cell = new PdfPCell(text);
        table.RunDirection = PdfWriter.RUN_DIRECTION_RTL; // can also be set on the cell
        table.AddCell(cell);
        doc.Add(table);
        doc.Close();
    }
}

You will probably want to get rid of the cell borders etc, but that information can be found elsewhere on SO or the iText website. iText should be able to handle text that contains both RTL and LTR characters.

EDIT

I think the source problem is actually with how the Arabic text is rendered in Visual Studio and in Firefox (my browser), or alternatively with how the Strings are concatenated. I'm not very familiar with Arabic text editors, but the text seems to come out correctly if we do this:

Arabic text in Visual Studio

FYI I had to take a screenshot, because copy-pasting into the browser from VS (and vice versa) messes up the order of the parts of the text.