I've just created a custom media type for PDF document files. I just wanted to get more information about a document than just the extension and the filesize. Searching the net I found a project named iTextSharp. iTextSharp is a port of the iText open source java library for PDF generation written entirely in C# for the .NET platform.

The best solution to do this is to hook into the BeforeSave event for Media. When dealing with a umbracoFile that has the extension .pdf the iTextSharp library is invoced to get properties like the number of pages, title, pdf version, author and the table of contents.

My solution has three parts

1. Create a custom Media Type for a PDF Document with the extended properties
2. Change the database datatype for the "Label" Data Type to Ntext
3. Create an event handler for the BeforeSave event for Media

Create a Media Type like the image below. The name of the Media Type is "PdfDocument". The properties are:

• Upload File, umbracoFile, Upload
• Type, umbracoExtension, Label
• Size, umbracoBytes, Label
• Title, umbracoPdfTitle, Label
• Author, umbracoPdfAuthor, Label
• Version, umbracoPdfVersion, Label
• Number of pages, umbracoPdfNumberOfPages, Label
• Table of contents, umbracoPdfTableOfContents, Label

The following EventHandler is created for the BeforeSave event on Media

using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Text;
using System.Web;
using iTextSharp.text.pdf;
using umbraco.BusinessLogic;
using umbraco.cms.businesslogic;
using umbraco.cms.businesslogic.media;

namespace UmbracoWebApplication.CustomEvents
{
    public class PdfDocumentSaveEvent : ApplicationBase
    {
        public PdfDocumentSaveEvent()
        {
            Media.BeforeSave += Media_BeforeSave;
        }

        private void Media_BeforeSave(Media sender, SaveEventArgs e)
        {
            if (sender.ContentType.Alias == "PdfDocument")
            {
                //Check if we're dealing with a pdf document
                if (Path.GetExtension(sender.getProperty("umbracoFile").Value.ToString()) == ".pdf")
                {
                    //Get the mapped path for the pdf document
                    string pdfPath = sender.getProperty("umbracoFile").Value.ToString();
                    string mappedFilePath = HttpContext.Current.Server.MapPath(pdfPath);

                    var reader = new PdfReader(mappedFilePath);

                    //Fill the additional properties
                    sender.getProperty("umbracoPdfTitle").Value =
                        reader.Info.Where(p => p.Key == "Title").FirstOrDefault().Value;
                    sender.getProperty("umbracoPdfAuthor").Value =
                        reader.Info.Where(p => p.Key == "Author").FirstOrDefault().Value;
                    sender.getProperty("umbracoPdfVersion").Value = reader.PdfVersion;
                    sender.getProperty("umbracoPdfNumberOfPages").Value = reader.NumberOfPages;
                    sender.getProperty("umbracoPdfTableOfContent").Value = GenerateTextBookmark(reader);
                }
                else
                {
                    //If the extension is not pdf we clear the custom fields
                    //This is also a solution for removing the file
                    sender.getProperty("umbracoPdfTitle").Value = String.Empty;
                    sender.getProperty("umbracoPdfAuthor").Value = String.Empty;
                    sender.getProperty("umbracoPdfVersion").Value = String.Empty;
                    sender.getProperty("umbracoPdfNumberOfPages").Value = String.Empty;
                    sender.getProperty("umbracoPdfTableOfContent").Value = String.Empty;
                }
            }
        }

        private static string GenerateTextBookmark(PdfReader reader)
        {
            IEnumerable<Dictionary<string, object>> bookmarks = SimpleBookmark.GetBookmark(reader);

            var builder = new StringBuilder();

            return GenerateBookmarkSection(bookmarks, builder);
        }

        private static string GenerateBookmarkSection(IEnumerable<Dictionary<string, object>> section,
                                                      StringBuilder builder)
        {
            if (section == null) return String.Empty;

            builder.Append("<ul>");

            foreach (var bookmark in section)
            {
                object value;
                bookmark.TryGetValue("Title", out value);

                builder.Append("<li>");
                builder.Append(value.ToString());

                if (bookmark.Count == 5)
                {
                    GenerateBookmarkSection((IEnumerable<Dictionary<string, object>>) bookmark.Values.Last(), builder);
                }

                builder.Append("</li>");
            }

            builder.Append("</ul>");

            return builder.ToString();
        }
    }
}

More about iTextSharp can be found on http://itextpdf.com/ This is also a nice library to index PDF documents on the containing text.

The result of my custom event handler is shown in the following image

So have fun and see you in Gent (BE) on the level 2 training.

Greetings, Jan Kees Velthoven