/// <summary>
/// Makes a complete clone of the epub
/// </summary>
/// <returns>A full clone of the epub</returns>
public Epub Clone()
{
// copy the zip file to a memory stream and read it
Ionic.Zip.ZipFile newFile;
MemoryStream ms = new MemoryStream();
file.Save(ms);
ms.Position = 0;
newFile = Ionic.Zip.ZipFile.Read(ms);
Epub clone = new Epub(newFile)
{
Author = Author,
Title = Title,
};
clone.Entries = new Dictionary <string, HtmlEntry>();
clone.EntryOrder = new List <HtmlEntry>();
// copy all the entries
foreach (var e in EntryOrder)
{
var clonedEntry = new HtmlEntry()
{
Href = e.Href, Html = e.Html, MimeType = e.MimeType
};
clone.Entries.Add(e.Href, clonedEntry);
clone.EntryOrder.Add(clonedEntry);
}
return(clone);
}
/// <summary>
/// Fully read an epub file to memory and keep the text entries and some general info like Title and Author seperate
/// </summary>
/// <param name="path">The path of the epub file</param>
/// <returns>An epub object read from the given file</returns>
public static Epub FromFile(string path)
{
// read the entire file, and interpret it as a zip file
var epubBytes = System.IO.File.ReadAllBytes(path);
var file = Ionic.Zip.ZipFile.Read(epubBytes);
Epub epub = new Epub(file);
// read the metadata container xml info
XmlDocument doc = new XmlDocument();
using (MemoryStream ms = new MemoryStream())
{
file[@"META-INF\container.xml"].Extract(ms);
ms.Position = 0;
doc.Load(ms);
}
// determine the href of the content manifest, which is stored in the full-path attribute of the rootfile tag
var node = doc.ChildNodes.GetAllNodes().Where(n => n.Name == "rootfile").FirstOrDefault();
if (node != null)
{
string contentPath = node.Attributes["full-path"].Value;
// keep the relative path to the manifest file, because all entries in the manifest will be relative
string basePath = System.IO.Path.GetDirectoryName(contentPath);
using (MemoryStream ms = new MemoryStream())
{
file[contentPath].Extract(ms);
ms.Position = 0;
doc = new XmlDocument();
doc.LoadXml(XDocument.Load(ms).Root.StripNamespaces().ToString());
}
// read the title if present
var titleNode = doc.SelectSingleNode("package/metadata/title");
if (titleNode != null)
{
epub.Title = titleNode.InnerText;
}
// read the author if present
var authorNode = doc.SelectSingleNode("package/metadata/creator");
if (authorNode != null)
{
epub.Author = authorNode.InnerText;
}
// read all the entries in the manifest
var items = doc.SelectNodes("package/manifest/item");
Dictionary <string, HtmlEntry> entries = new Dictionary <string, HtmlEntry>(items.Count);
var entryOrder = new List <HtmlEntry>(items.Count);
foreach (var item in items.Cast <XmlNode>())
{
string href = System.IO.Path.Combine(basePath, item.Attributes["href"].Value);
string mimeType = item.Attributes["media-type"].Value;
// if the entry is a html file
if (mimeType == "application/xhtml+xml" || mimeType.Contains("html") || mimeType.Contains("xml"))
{
// extract the file to a a memory stream and read it to a string
using (MemoryStream ms = new MemoryStream())
{
file[Uri.UnescapeDataString(href)].Extract(ms);
ms.Position = 0;
StreamReader reader = new StreamReader(ms);
string html = reader.ReadToEnd();
// store the entry
var te = new HtmlEntry()
{
Href = href,
MimeType = mimeType,
Html = html
};
entries.Add(href, te);
entryOrder.Add(te);
}
}
}
epub.Entries = entries;
epub.EntryOrder = entryOrder;
}
else
{
throw new Exception("No content metadata");
}
return(epub);
}
