Chris Haas Chris Haas - 3 months ago 14
C# Question

How to convert HTML to PDF using iTextSharp

I want to convert the below HTML to PDF using iTextSharp but don't know where to start:

<style>
.headline{font-size:200%}
</style>
<p>
This <em>is </em>
<span class="headline" style="text-decoration: underline;">some</span>
<strong>sample<em> text</em></strong>
<span style="color: red;">!!!</span>
</p>

Answer

First, HTML and PDF are not related although they were created around the same time. HTML is intended to convey higher level information such as paragraphs and tables. Although there are methods to control it, it is ultimately up to the browser to draw these higher level concepts. PDF is intended to convey documents and the documents must "look" the same wherever they are rendered.

In an HTML document you might have a paragraph that's 100% wide and depending on the width of your monitor it might take 2 lines or 10 lines and when you print it it might be 7 lines and when you look at it on your phone it might take 20 lines. A PDF file, however, must be independent of the rendering device, so regardless of your screen size it must always render exactly the same.

Because of the musts above, PDF doesn't support abstract things like "tables" or "paragraphs". There are three basic things that PDF supports: text, lines/shapes and images. (There are other things like annotations and movies but I'm trying to keep it simple here.) In a PDF you don't say "here's a paragraph, browser do your thing!". Instead you say, "draw this text at this exact X,Y location using this exact font and don't worry, I've previously calculated the width of the text so I know it will all fit on this line". You also don't say "here's a table" but instead you say "draw this text at this exact location and then draw a rectangle at this other exact location that I've previously calculated so I know it will appear to be around the text".

Second, iText and iTextSharp parse HTML and CSS. That's it. ASP.Net, MVC, Razor, Struts, Spring, etc, are all HTML frameworks but iText/iTextSharp is 100% unaware of them. Same with DataGridViews, Repeaters, Templates, Views, etc. which are all framework-specific abstractions. It is your responsibility to get the HTML from your choice of framework, iText won't help you. If you get an exception saying The document has no pages or you think that "iText isn't parsing my HTML" it is almost definite that you don't actually have HTML, you only think you do.

Third, the built-in class that's been around for years is the HTMLWorker however this has been replaced with XMLWorker (Java / .Net). Zero work is being done on HTMLWorker which doesn't support CSS files and has only limited support for the most basic CSS properties and actually breaks on certain tags. If you do not see the HTML attribute or CSS property and value in this file then it probably isn't supported by HTMLWorker. XMLWorker can be more complicated sometimes but those complications also make it more extensible.

Below is C# code that shows how to parse HTML tags into iText abstractions that get automatically added to the document that you are working on. C# and Java are very similar so it should be relatively easy to convert this. Example #1 uses the built-in HTMLWorker to parse the HTML string. Since only inline styles are supported the class="headline" gets ignored but everything else should actually work. Example #2 is the same as the first except it uses XMLWorker instead. Example #3 also parses the simple CSS example.

//Create a byte array that will eventually hold our final PDF
Byte[] bytes;

//Boilerplate iTextSharp setup here
//Create a stream that we can write to, in this case a MemoryStream
using (var ms = new MemoryStream()) {

    //Create an iTextSharp Document which is an abstraction of a PDF but **NOT** a PDF
    using (var doc = new Document()) {

        //Create a writer that's bound to our PDF abstraction and our stream
        using (var writer = PdfWriter.GetInstance(doc, ms)) {

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

            //Our sample HTML and CSS
            var example_html = @"<p>This <em>is </em><span class=""headline"" style=""text-decoration: underline;"">some</span> <strong>sample <em> text</em></strong><span style=""color: red;"">!!!</span></p>";
            var example_css = @".headline{font-size:200%}";

            /**************************************************
             * Example #1                                     *
             *                                                *
             * Use the built-in HTMLWorker to parse the HTML. *
             * Only inline CSS is supported.                  *
             * ************************************************/

            //Create a new HTMLWorker bound to our document
            using (var htmlWorker = new iTextSharp.text.html.simpleparser.HTMLWorker(doc)) {

                //HTMLWorker doesn't read a string directly but instead needs a TextReader (which StringReader subclasses)
                using (var sr = new StringReader(example_html)) {

                    //Parse the HTML
                    htmlWorker.Parse(sr);
                }
            }

            /**************************************************
             * Example #2                                     *
             *                                                *
             * Use the XMLWorker to parse the HTML.           *
             * Only inline CSS and absolutely linked          *
             * CSS is supported                               *
             * ************************************************/

            //XMLWorker also reads from a TextReader and not directly from a string
            using (var srHtml = new StringReader(example_html)) {

                //Parse the HTML
                iTextSharp.tool.xml.XMLWorkerHelper.GetInstance().ParseXHtml(writer, doc, srHtml);
            }

            /**************************************************
             * Example #3                                     *
             *                                                *
             * Use the XMLWorker to parse HTML and CSS        *
             * ************************************************/

            //In order to read CSS as a string we need to switch to a different constructor
            //that takes Streams instead of TextReaders.
            //Below we convert the strings into UTF8 byte array and wrap those in MemoryStreams
            using (var msCss = new MemoryStream(System.Text.Encoding.UTF8.GetBytes(example_css))) {
                using (var msHtml = new MemoryStream(System.Text.Encoding.UTF8.GetBytes(example_html))) {

                    //Parse the HTML
                    iTextSharp.tool.xml.XMLWorkerHelper.GetInstance().ParseXHtml(writer, doc, msHtml, msCss);
                }
            }


            doc.Close();
        }
    }

    //After all of the PDF "stuff" above is done and closed but **before** we
    //close the MemoryStream, grab all of the active bytes from the stream
    bytes = ms.ToArray();
}

//Now we just need to do something with those bytes.
//Here I'm writing them to disk but if you were in ASP.Net you might Response.BinaryWrite() them.
//You could also write the bytes to a database in a varbinary() column (but please don't) or you
//could pass them to another function for further PDF processing.
var testFile = Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.Desktop), "test.pdf");
System.IO.File.WriteAllBytes(testFile, bytes);