Pdfizer, a dumb HTML to PDF converter, in C#

• Download source code - 519 Kb

Introduction

This article presents a basic HTML to PDF converter: with this library, you can transform simple HTML pages to nice and printable PDF files.

The HTML cleaning is done with NTidy (see [1]), a .NET wrapper for the HTML Tidy library (see [2]). The PDF generation is done with iTextSharp, a PDF generation library (see [3]).

Transformation Pipe

Transforming HTML documents to PDF is a fairly complex task. Hopefully, there exists powerful tools on the web that could help me accomplish this.

Parsing HTML

The first problem to handle was that HTML is usually "dirty": the structure is usually not XML conformant and trying to parse HTML pages with the XmlDocument will usually lead to a failure.

To overcome this problem, I had to write a .NET wrapper around HTML Tidy (see [2]). HTML Tidy is a very useful application that takes "dirty" HTML and returns it cleaned as much as possible. The .NET wrapper exposes a DOM-like class structure so that you can use it much like XmlDocument.

Hence, with NTidy, we can safely parse HTML document.

Creating PDF

The PDF creation is done by iTextSharp (see [3]), a .NET library hosted on SourceForge, that gives you the tool to create PDF easily. Hence, the PDF creation problem is solved.

Reading, Traversing

With NTidy and iTextSharp on my toolset, I could start to create the generator. The generator works like this: it first reads the input with NTidy, then traverses the DOM tree and generates the PDF fragments with iTextSharp.

Quick Example

The library usage is done through the HtmlToPdfConverter class. Creating a PDF file is done through the following steps, as illustrated in the example:

1. Create a converter,
2. Open a new PDF file using the Open method,
3. Add a chapter,
4. Feed HTML to the converter,
5. If you want another chapter, go to 3.
6. When finished, close the PDF file by calling Close.

// create converter

HtmlToPdfConverter html2pdf = new HtmlToPdfConverter();

// open new pdf file

html2pdf.Open(@"test");
// start a chapter

html2pdf.AddChapter(@"Dummy Chapter");
string html = ...;
// convert string

html2pdf.Run(html);
// add a new chapter

html2pdf.AddChapter(@"Boost page");
// read web page

html2pdf.Run(new Uri(@"http://www.boost.org/libs/libraries.htm"));
// close and finish pdf file.

html2pdf.Close();

What to expect and not expect

Don't expect too much from this tool, it will not work with complex HTML pages and will give fairly good results with simple HTML pages. Specially, tables are not yet supported.

Reference

1. NTidy, a .NET wrapper around Tidy.
2. HTML Tidy home page.
3. iTextSharp, PDF generation tool.