C#探秘系列（七）：XML文件操作（一）

XML 是可扩展标记语言（Extensible Markup Language）的缩写，其中的 标记（markup）是关键部分。你可以创建内容，然后使用限定标记标记它，从而使每个单词、短语或块成为可识别、可分类的信息。因而在C#中，对xml文件的操作尤为重要。本博客着力于在C#中如何对xml文件实施诸如创建、搜索、序列化和反序列化等操作，默认你已有一定的xml基础。
xml的基础知识可参照下面博客：
http://www.ibm.com/developerworks/cn/xml/x-newxml/
下面介绍的操作均基于Customer对象：

<Customer>
  <FirstName>Orlando</FirstName>
  <LastName>Gee</LastName>
  <EmailAddress>orlando0@hotmail.com</EmailAddress>
</Customer>

一、利用C#创建xml文档

创建xml文件需要四步：
（一）新建xml文档: XmlDocument();
（二）创建根结点: CreateElement();
（三）添加子元素: AppendChild();
（四）添加元素属性: CreateElement(), AppendChild()。
首先，要使用C#中定义的操作方法，必须引入命名空间，即：

using System.Xml

1、新建xml文档

XmlDocument customerXml = new XmlDocument();

2、创建根结点并将其添加

XmlElement rootElem = customerXml.CreateElement("Customers");
customerXml.AppendChild(rootElem);

3、添加子元素

XmlElement customerElem = customerXml.CreateElement("Customer");

4、添加元素属性

XmlElement firstNameElem = customerXml.CreateElement("FirstName");
firstNameElem.InnerText  = customer.FirstName;
customerElem.AppendChild(firstNameElem);

5、实例代码及输出

//Create a list of customers
List<Customer> customers = CreateCustomerList();
//Create a XmlDocument
XmlDocument customerXml = new XmlDocument();
//Create a root element
XmlElement rootElem = customerXml.CreateElement("Customers");
customerXml.AppendChild(rootElem);

  foreach (Customer customer in customers)
  {
  //Create new element representing the customer object.
  XmlElement customerElem = customerXml.CreateElement("Customer");

  //Add element “FirstName” property to the customer element.
  XmlElement firstNameElem = customerXml.CreateElement("FirstName");
  firstNameElem.InnerText  = customer.FirstName;
  customerElem.AppendChild(firstNameElem);

  //Add element: LastName property  to the customer element.
  XmlElement lastNameElem = customerXml.CreateElement("LastName");
  lastNameElem.InnerText = customer.LastName;
  customerElem.AppendChild(lastNameElem);

  //Add element: EmailAddress property to the customer element.
  XmlElement emailAddress = customerXml.CreateElement("EmailAddress");
  emailAddress.InnerText = customer.EmailAddress;
  customerElem.AppendChild(emailAddress);

  //Finally add the customer element to the XML document
  rootElem.AppendChild(customerElem);
}
Console.WriteLine(customerXml.OuterXml);
Console.Read();

输出结果：

<Customers>
  <Customer>
    <FirstName>Orlando</FirstName>
    <LastName>Gee</LastName>
    <EmailAddress>orlando0@hotmail.com</EmailAddress>
  </Customer>
  <Customer>
    <FirstName>Keith</FirstName>
    <LastName>Harris</LastName>
    <EmailAddress>keith0@hotmail.com</EmailAddress>
  </Customer>
   <> …</>
</Customers>

如果想将customer的属性添加至元素表头，如：

<Customer FirstName="Orlando" LastName="Gee">
  <EmailAddress>…</EmailAddress>
</Customer>

可以使用CreateAttribute()方法:

// Add an attribute :FirstName property  to the customer element.
XmlAttribute firstNameAttr = customerXml.CreateAttribute("FirstName");
firstNameAttr.Value = customer.FirstName;
customerElem.Attributes.Append(firstNameAttr);

二、利用XPath在xml文件中搜索节点或属性

1.使用绝对路径搜索

XPath可以快速定位到Xml中的节点或者属性。XPath语法很简单，但是强大够用，有关其详细语法参照：
http://www.cnblogs.com/yukaizhao/archive/2011/07/25/xpath.html
搜索的结果包括两种：XmlNode与XmlEelment，其继承关系如下：
System.Object
System.Xml.XmlNode
System.Xml.XmlLinkedNode
System.Xml.XmlElement
在XmlNode类中主要用到了两个函数用来实现搜索：
SelectSingleNode(String xpath)与SelecNodes(String xpath)，差别无非是单复数的问题。
先来一段代码：

string xPath = "/Customers/Customer[@FirstName='Donna']";
XmlNode oneCustomer = customerXml.SelectSingleNode(xPath);

其中第一个/表示从根节点开始搜索，第二个/就表示根节点之后的子节点了。[ ]表示搜索条件，@表示对应的属性。这样就从xml文件中搜索到了一个满足条件的xml节点。
代码示例：

string xPath = "/Customers/Customer[@FirstName='Donna']";
 XmlNode oneCustomer = customerXml.SelectSingleNode(xPath);
 Console.WriteLine("
SelectSingleNode("{0}")...", xPath);
 if (oneCustomer != null)
    Console.WriteLine(oneCustomer.OuterXml);
else
    Console.WriteLine("Not found");

xPath = "/Customers/Customer[@FirstName='Donna']";
XmlElement customerElem = customerXml.SelectSingleNode(xPath) as XmlElement;
Console.WriteLine("
SelectSingleNode("{0}")...", xPath);
if (customerElem != null)
{
 Console.WriteLine(customerElem.OuterXml);
 Console.WriteLine("customerElem.HasAttributes = {0}", customerElem.HasAttributes);
}
else
  Console.WriteLine("Not found");

输出结果：

SelectSingleNode("/Customers/Customer[@FirstName='Donna']")...
<Customer FirstName="Donna" LastName="Carreras">
   <EmailAddress>donna0@adventure-works.com</EmailAddress>
</Customer>

SelectSingleNode("/Customers/Customer[@FirstName='Donna']")...
<Customer FirstName="Donna" LastName="Carreras">
   <EmailAddress>donna0@adventure-works.com</EmailAddress>
</Customer>
customerElem.HasAttributes = True

2、使用相对路径：Axes坐标轴

如果我并不了解xml文件的构造方式，那就无法用绝对路径来搜索了，相对路径就派上用场了：

xPath = "descendant::Customer[@FirstName='Donna']";
oneCustomer = customerXml.SelectSingleNode(xPath);

其中descendant表示搜索所在位置后的所有符合条件的后代，即子结点。
示例代码：

xPath = "descendant::Customer[@FirstName='Donna']";
oneCustomer = customerXml.SelectSingleNode(xPath);
Console.WriteLine("
SelectSingleNode("{0}")...", xPath);
if (oneCustomer != null)
   Console.WriteLine(oneCustomer.OuterXml);
else
   Console.WriteLine("Not found");

输出结果：

SelectSingleNode("descendant::Customer[@FirstName='Donna']")...
<Customer FirstName="Donna" LastName="Carreras">
    <EmailAddress>donna0@adventure-works.com</EmailAddress>
</Customer>

3、使用XPath函数确定条件

xPath = "descendant::Customer[starts-with(@LastName, 'G') " +
    "and contains(EmailAddress, 'adventure-works.com')]";
customers = customerXml.SelectNodes(xPath);
Console.WriteLine("
SelectNodes("{0}")...", xPath);
if (customers != null)
{
   foreach (XmlNode customer in customers)
        Console.WriteLine(customer.OuterXml);
}
else
   Console.WriteLine("Not found");

常用的XPath包括：
start-with()表示传入的属性以某个字段开头//Customer[starts_with(@LastName, ‘G’)]表示取lasName以’G’打头的customer节点；
contains()表示包含第二个参数描述的属性的节点//Customer[contains(EmailAddress, ‘adventure-works.com’)]表示取EmailAddress为adventure-works.com的customer
position() 表示节点的序号例如 //cat[position() = 2] 表示取序号为2的dog节点；
last() 表示取最后一个节点 //cat[last()] ；
name() 表示当前节点名字 /pets/*[name() != ‘pig’] 表示/pets下名字不是pig的子节点。
故输出结果为：

SelectNodes("descendant::Customer[starts-with(@LastName, 'G') and contains(EmailAddress, 'adventure-works.com')]")...
<Customer FirstName="Orlando" LastName="Gee">
   <EmailAddress>orlando0@adventure-works.com</EmailAddress>
</Customer>
<Customer FirstName="Janet" LastName="Gates">
   <EmailAddress>janet1@adventure-works.com</EmailAddress>
</Customer>

暂驻笔于此，洒家且歇歇~