手头有个C++项目用到了XML解析，我采用的是开源的tinyxml，项目运行过程中发现其对XML节点内容的解析过滤掉了“&”符号，于是乎开始读源码解决问题：

在文件tinyxmlparser.cpp中 Line 438,函数const char* TiXmlBase::GetEntity( const char* p,char* value,int* length,TiXmlEncoding encoding )实现代码中有：

<span style="font-size:18px;">// So it wasn't an entity,its unrecognized,or something like that.
        *value = *p;    // Don't put back the last one,since we return it!
        //*length = 1;    // Leave unrecognized entities - this doesn't really work.
                                        // Just writes strange XML.</span>

源码中将*length=1注释掉了，直接导致特殊字符“&”被过滤，若你也遇到同样的问题，

只需要将*length=1放开注释即可，如下：

<span style="font-size:18px;">// So it wasn't an entity,since we return it!
        *length = 1;    // Leave unrecognized entities - this doesn't really work.  no no no  by yangw
                                        // Just writes strange XML.</span>