要求：把程序迁移到web平台，通过用户上传TXT的方式接收文件。建议(但不强制要求)保留并维护Console版本，有利于测试。
在页面上设置上传的控件,然后在servlet中接受，得到的是一个字节流，然后转化为字符型在原有代码中进行统计。
jsp页面的代码如下
<%@ page language="java" contentType="text/html; charset=utf-8"
pageEncoding="utf-8"%>
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<title>Insert title here</title>
</head>
<body>
<table>
<tr>
<td>
<form action="server/CountWordServlet" method="post" enctype="multipart/form-data">
请上传要统计的文件<input type="file" name="sourceFile"/>
<input type="submit" value="上传">
</form>
</td>
</tr>
</table>
</body>
</html>
展示结果的页面如下
<%@page import="com.server.servlet.Word"%>
<%@page import="java.util.ArrayList"%>
<%@ page language="java" contentType="text/html; charset=utf-8"
pageEncoding="utf-8"%>
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<%ArrayList<Word> list=(ArrayList<Word>)request.getAttribute("list"); %>
<title>Insert title here</title>
</head>
<body>
<table> <%
if(list!=null&&list.size()!=0){
%>
<tr> <td>单词</td><td>数量</td> </tr>
<%
for(int i=0;i<list.size();i++){
String word=((Word)list.get(i)).getWord();
int num=((Word)list.get(i)).getNum();
%><tr>
<td><%=word%></td>
<td><%=num%></td>
</tr>
<%
}
}else{ %>
<td>此文件没有单词或者文件不存在</td>
<% }
%> </table>
</body>
</html>
servle中的代码如下
public class CountWordServlet extends HttpServlet {
private static final long serialVersionUID = 1L;
protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
try {
request.setCharacterEncoding("utf-8");
ArrayList<Word> list=new ArrayList<>();
DiskFileItemFactory factory=new DiskFileItemFactory();
ServletFileUpload upload=new ServletFileUpload(factory);
FileItemIterator iterator=upload.getItemIterator(request);
while(iterator.hasNext()){
InputStream input=iterator.next().openStream();
WordCountFreq wcf=new WordCountFreq();
list=(ArrayList<Word>) wcf.sortAndOutput(input);
request.setAttribute("list", list);
}
} catch (FileUploadException e) {
e.printStackTrace();
}
System.out.println("成功了！");
response.setContentType("text/html;charset=utf-8");
request.getRequestDispatcher("/show.jsp").forward(request, response);
}
}
然后将统计过程的关键方法sortAndOutput（）展示如下
public List<Word> sortAndOutput(InputStream input) throws IOException {
BufferedInputStream bis=new BufferedInputStream(input);
byte [] buf = new byte[1024];
int len = -1;
String temp = "";
String lastWord = "";
while((len = bis.read(buf)) != -1) {
//将读取到的字节数据转化为字符串打印出来
String str = new String(buf,0,len);
temp = "";
temp += lastWord;
for (int i = 0; i < str.length(); i++) {
temp += str.charAt(i);
}
lastWord = "";
if (Character.isLetter(str.charAt(str.length()-1))) {
int j, t;
for (j = str.length() - 1, t = 0; Character.isLetter(str.charAt(j)); j--, t++);
temp = temp.substring(0, temp.length() - t);
for (int k = j + 1; k < str.length(); k++) {
lastWord += str.charAt(k);
}
}
root = generateCharTree(temp);
}
示例如下
在没做web版本之前，只是传入文件的路径进行处理。改为web版本之后将遇见的一点小困难是要将字节流转化为字符进行处理，经过查询也很快就解决了。
ssh:git@git.coding.net:muziliquan/GUIVersion.git
git:git://git.coding.net/muziliquan/GUIVersion.git