Deserializing an object

Before we look at the

Serialize() method more in depth I will show you that de-serializing an object also works with two lines of code:

Listing 9.1: Using the Deserialize() method to read an XmlDocuement

void DeserializeACar()

{

  XmlSerializer ser = new XmlSerializer(typeof(Car));

  XmlTextReader reader = new XmlTextReader("car.xml");

  Car wifesCar = (Car)ser.Deserialize(reader);

  // wifesCar is a 1997 Ford Explorer again

  reader.Close();

}

This time we instantiate an XmlTextReader for the XML file we created in the previous example. Then we pass it to the Deserialize() method and the XmlSerializer determines that it 1) needs to instantiate a new Car object, 2) creates a new Car object and 3) sets the Car object’s fields to the values supplied by the XmlReader before it returns the new Car object to you. Note that the type cast of the Deserialize() method’s return value to Car is necessary because the Deserialize() is declared to return only an object and the assignment would not be valid without the cast.

Imagine you are developing an application which receives large XML documents with nested types and hundreds of nodes. How many lines of code would you have to write to parse these documents and assign their content to an object hierarchy for further processing? With XML data binding you define classes that map to the XML format of the documents you are receiving and let the XmlSerializer do the mapping. Instead or writing hundreds of lines of code you get by writing four.  Starting with section 9.2.2 we will learn how put this powerful functionality to work to bind objects to arbitrarily complex XML formats.

Serialize() and Deserialize() methods

Both methods, Serialize() and Deserialize() leverage the concepts of streams and abstract base classes we learned about in chapter 2. Several overloads shown in table 9.1 and table 9.2 are available for both methods to allow dealing with XML from a variety of sources.

Table 9.1: Available overloads for the Serialize() method. Each overloaded method allows different options to control encoding and formatting of the output. Some overloads also allow the declaration of XML namespaces used in the created XML document.

The XmlSerializer always operates on the current position of the input or output stream. Using raw streams you can even inject and retrieve XML from any kind of data stream, not only a well-formed XML document.

Table 9.2: The available overloads for the Deserialize() method allow to deserialize objects from a variety of sources: file-based, memory-based and over a network.