Why is my PDF being generated as blank?

Question

I am using ItextSharp and c#, asp.net MVC to generate a PDF report. However, when I generate the report the PDF comes back as blank (apart from a header which is working fine). I would love your input.

The code that generates the report is as follows:

                using (var writer = PdfWriter.GetInstance(doc, ms))
                {
                    // This sorts out the Header and Footer parts.
                    var headerFooter = new ReportHeaderFooter(reportsAccessor);
                    writer.PageEvent = headerFooter;

                    var rootPath = ConfigurationManager.AppSettings["SaveFileRootPath"];
                    string html = File.ReadAllText(rootPath + "/reports/report.html");                      

                    // Perform the parsing to PDF
                    doc.Open();

                    // The html needs to be sorted before this call.
                    StringReader sr = new StringReader(html);
                    XMLWorkerHelper.GetInstance().ParseXHtml(writer, doc, sr);
                    doc.Close();
                    writer.Close();
                    res = ms.ToArray();
                }

I know there is a lot hidden, but for arguments sake an example of the HTML that this generates and puts in the StringReader is here:

    <table style="font-size:10pt">
        <tr>
            <td rowspan="4"> Address<br/> </td>
            <td>Phone: phone</td>
        </tr>
        <tr>
            <td>Fax: fax</td>
        </tr>
        <tr>
            <td>Email: [e-mail]</td>
        </tr>
        <tr>
            <td>Website: example.com</td>
        </tr>
    <table>

    <table style="font-size:10pt; width:100%">
        <tr style="width:50%">
            <td>Settlement: 30 days from invoice</td>
        </tr>
        <tr>
            <td>Delivery Charge Notes: N/A</td>
        </tr>
    </table>
    <p style="width:100%; font-size:10pt">
    I love notes</p>

    <table>
    <tr style="font-weight:bold">
        <td>Item</td>
        <td>Item Code</td>
        <td>Description</td>
        <td>Unit Size</td>
        <td>Units Per Outer</td>
        <td>Units Per Pallet</td>
        <td>Invoice Price Volume Breaks</td>
        <td>Branding</td>
        <td>Notes</td>
    </tr>

    </table>

However, this html generates a nice blank PDF file (not what I want). I can't see what might be wrong with this and would love some input in to two things:

[1] Why the report is blank [2] If this is an error in the parsing, why it doesn't come up with an error and instead a blank report (is there a setting or argument I can set that will throw errors which are much more useful than blank reports?)

UPDATE: I have added the code for the header/footer

public string HeaderTitle { get; set; } public IReportsAccessor ReportsAccessor { get; set; } public ReportHeaderFooter(IReportsAccessor reportsAccessor) { this.ReportsAccessor = reportsAccessor; }

    public override void OnStartPage(PdfWriter writer, Document document)
    {
        base.OnStartPage(writer, document);
        var rootPath = ConfigurationManager.AppSettings["SaveFileRootPath"];
        GetReportImageResult imgInfo = ReportsAccessor.GetImage(4);

        byte[] header_img = imgInfo.ReportImage;
        string logoFn = rootPath + "/tmp/logo.png";
        File.WriteAllBytes(logoFn, header_img);
        string html = File.ReadAllText(rootPath + "/reports/report_header.html");
        html = html.Replace("{{ title }}", HeaderTitle);
        html = html.Replace("{{ logo_img }}", logoFn);

        StringReader sr = new StringReader(html);
        XMLWorkerHelper.GetInstance().ParseXHtml(writer, document, sr);          
    }

Answer 1

When I run your HTML through my code, I do get an exception, because there's one sneaky error in your HTML. The first table doesn't have a closing tag: it says <table> instead of </table>. That's easily overlooked. You can find the corrected HTML here: table10.html

This is what the resulting PDF looks like: html_table_10.pdf

One major difference between your code and mine, is that I used iText / XML Worker 5.5.6 (Java) and you used iTextSharp / XML Worker (C#). While the Java iText and XML Worker are kept in sync with the C# iTextSharp and XML Worker (and should therefore have identical behavior), the underlying XML parser is different.

I know that we decided to behave like browsers (*), silently allowing some bad HTML constructs to fail, but my experience with the Java XML parsing is that is throws exceptions. Maybe the C# XML parsing doesn't.

(*) I'm the CEO of the iText Group, consisting of companies in Belgium, the US and Singapore.

As for the "css resolver crying" exception, we'll remove it in the next release: