Html转XML

使用nekohtml进行转化

nekohtml下载地址：

源程序：

html2xml.java

import org.w3c.dom.Node;
import org.w3c.dom.DocumentFragment;
import org.w3c.dom.html.HTMLDocument;
import org.xml.sax.InputSource;
import org.apache.html.dom.HTMLDocumentImpl;
import org.cyberneko.html.parsers.DOMFragmentParser;
import javax.xml.transform.stream.StreamResult;
import javax.xml.transform.dom.DOMSource;
import javax.xml.transform.Transformer;
import javax.xml.transform.TransformerFactory;
import java.util.Properties;
import java.util.Calendar;
import java.io.File;
import java.io.InputStreamReader;
import java.io.InputStream;
import java.io.FileReader;
import java.net.HttpURLConnection;
import java.net.URL;

public class html2xml {
    public static void main(String args[]){
        if(args!=null&&args.length>=2){
            try {
                String path=args[0];
                String fromfile=args[1];
                String outputfile=getFileName();
                if(args.length>2){
                    outputfile=args[2];
                }
                boolean b=Boolean.valueOf(fromfile).booleanValue();
                html2xml h2x=new html2xml();
                DocumentFragment df=h2x.getSourceNode(path,b);
                File file=new File(outputfile);
                if(file.exists())
                    file.delete();
                h2x.genXmlFile(df,file);
                System.out.println("generate "+file.getCanonicalPath()+" successfully!");
            } catch (Exception e) {
                e.printStackTrace();  //To change body of catch statement use File | Settings | File Templates.
            }

        }else{
            System.out.println("usage:html2xml path fromfile [outputfile]");
            System.out.println("html2xml false D:/tempfile.xml");
            System.out.println("html2xml D:/htmlfile.htm true D:/tempfile.xml");
            System.out.println("--");
        }
    }

    public void genXmlFile(Node output,File file) throws Exception,Error{
            TransformerFactory tf=TransformerFactory.newInstance();
            Transformer transformer=tf.newTransformer();
            DOMSource source=new DOMSource(output);
            java.io.FileOutputStream fos=new java.io.FileOutputStream(file);
            StreamResult result=new StreamResult(fos);
            Properties props = new Properties();
            props.setProperty("encoding", "GB2312");
            props.setProperty("method", "xml");
            props.setProperty("omit-xml-declaration", "yes");

            transformer.setOutputProperties(props);

            transformer.transform(source,result);
            fos.close();

    }

    public DocumentFragment getSourceNode(String path,boolean fromfile) throws Exception,Error{
        DOMFragmentParser parser = new DOMFragmentParser();
        HTMLDocument document = new HTMLDocumentImpl();
        DocumentFragment fragment = document.createDocumentFragment();

            if(path!=null&&!path.trim().equals(""))
            {
                String tmp=path;

                if(fromfile){
                    File input = new File(path);
                    FileReader fr=new FileReader(input);
                    InputSource is=new InputSource(fr);
                    parser.parse(is,fragment);
                    fr.close();
                }else{

                    URL url = new URL(tmp);
                    HttpURLConnection con = (HttpURLConnection) url.openConnection();
                    InputStream inputs = con.getInputStream();
                    InputStreamReader isr=new InputStreamReader(inputs,"GBK");
                    InputSource source=new InputSource(isr);
                    parser.parse(source,fragment);
                }
                return fragment;
            }else{
                return null;
            }

    }

    public static String getFileName() throws Exception{
        Calendar c=Calendar.getInstance();
        String name="tmp"+c.get(Calendar.YEAR)+(c.get(Calendar.MONTH)<9?"0":"")+
                (c.get(Calendar.MONTH)+1)+(c.get(Calendar.DAY_OF_MONTH)<10?"0":"")+
                c.get(Calendar.DAY_OF_MONTH)+(c.get(Calendar.HOUR_OF_DAY)<10?"0":"")+
                c.get(Calendar.HOUR_OF_DAY)+(c.get(Calendar.MINUTE)<10?"0":"")+
                c.get(Calendar.MINUTE)+(c.get(Calendar.SECOND)<10?"0":"")+
                c.get(Calendar.SECOND)+(c.get(Calendar.MILLISECOND)<10?"0":"")+
                (c.get(Calendar.MILLISECOND)<100?"0":"")+c.get(Calendar.MILLISECOND);
        return name;
    }
}

目录结构：

html2xml
├─classes
├─lib
└─src

在目录html2xml下建立一个批处理文件run.bat，内容为：java -cp "./lib/nekohtml.jar;./lib/xercesImpl.jar;./lib/xml-apis.jar;./lib/commons-logging.jar;./lib/commons-discovery.jar;./lib/saaj.jar;./classes" html2xml %1 %2 %3

使用方法：

1.文件转化，从命令行输入：run D: est.html true D: est.xml，第一个参数是被转化的目标html文件，第二个参数是标志位，文件转化时使用“true”，第三个参数是输出的xml文件，第三个参数若省缺将会产生一个临时文件用于存储输出xml。

2.通过网页地址转换，从命令行输入：run false D:sina.xml，第一个参数是网页地址，第二个应该设置为false，第三个参数同上。

由于nekohtml具有较好的容错性，对于大多数情况都能够成功转化，需要注意的是目标网页的html标签存在多个时会发生错误，还有一个也是在使用中发现的，attribute='a'b'，对于这种属性无法解析，而一般是不会出现这种极烂的书写方式。

nekohtml可以从下载获得。