1.html page content:
<div class="mnr-c _yE">
<div class="_kk _wI">In the news</div>
<li class="card-section _df g _mZd">
<div class="_K2 _SYd">
<div style="overflow:hidden;width:134px;height:100px" class="thumb">
<a href="http://www.bbc.co.uk/news/uk-30172110" onmousedown="return rwt(this,'','','','2','AFQjCNG3I0r8D75WjgjZODuobF8ne7wCNw','','0CCwQpwIwAQ','','',event)">
<img height="100" id="uid_0" src="data:image/gif;base64,R0lGODlhAQABAIAAAP///////yH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==" width="134" border="0">
</a>
</div>
</div>
<div class="_I2">
<a class="_Dk" href="http://www.bbc.co.uk/news/uk-30172110" onmousedown="return rwt(this,'','','','2','AFQjCNG3I0r8D75WjgjZODuobF8ne7wCNw','','0CC0QqQIwAQ','','',event)">
Google case over online abuse settled</a>
<div class="_Ck kv">
<cite>BBC News</cite>
<span class="f"> - </span>
<span class="f" style="white-space:nowrap">21 hours ago
</span>
</div>
</div>
<span class="_dwd st s std" style="margin-left:144px">
A UK businessman who took <em>Google</em> to court over malicious web postings about him ...</span>
</li>
<div>
<li class="g _Nn _wbb card-section">
<a class="_Dk" href="http://www.pcworld.com/article/2851812/google-to-apps-users-take-more-responsibility-for-protecting-your-accounts.html" onmousedown="return rwt(this,'','','','3','AFQjCNH0fmBCNMjPanXErfX6GQmDNsZK7Q','','0CC8QqQIwAg','','',event)">
New Google Apps dashboard helps users protect accounts</a>
<div class="_Ck kv">
<cite>PCWorld</cite><span class="f"> - </span>
<span class="f" style="white-space:nowrap">5 hours ago</span>
</div>
</li>
<li class="g _Nn _Abb card-section">
<a class="_Dk" href="http://www.forbes.com/sites/georgeanders/2014/11/24/google-and-facebook-rewire-the-internet-as-fcc-dithers/" onmousedown="return rwt(this,'','','','4','AFQjCNGcPEbPFsUfSxeCneg_aFYBX65fNQ','','0CDEQqQIwAw','','',event)">
Google And Facebook Rewire The Internet As FCC Dithers</a>
<div class="_Ck kv">
<cite>Forbes</cite><span class="f"> - </span>
<span class="f" style="white-space:nowrap">8 hours ago</span>
</div>
</li>
</div>
2.获取标签：
//获取文档中所有的class="g"或者包含"g"的<li>标签
var allLiNodes = htmlDoc.DocumentNode.SelectNodes(@"//li[@class='g' or contains(@class,'g')]");
//获取当前节点及其所有子节点中的具有先辈的<img>的单个<a>标签
var imageNode = aImageTagNode.SelectSingleNode(@".//img[./ancestor::a/@href]");
3.w3cshcool 实例：
http://www.w3school.com.cn/xpath/xpath_axes.asp
第一次接触，主要是对爬虫的结果进行解析，然后存储测试与发布，准确率还挺高的。
另：若是浏览器，直接获取xpath的方法：
F12 开发真工具，找到对应的元素，在标签上右键，就可以看到一个copy xpath 直接复制即可。