Java之查找html文件当中的所有标签

本文主要是介绍从html文件中获取到所有标签：

1. 
   
2. 主要还是字符串的操作
   
3. 使用了Map
   
4. 使用了Formatter类对输出格式进行控制，可以固定输出的长度以及设置对齐方式
   

import java.util.*;
import java.io.*;


public class ShowTags {


    public static Map<String , Integer> SelectTags(String filePath){
        Map<String , Integer> map = new HashMap<String, Integer>();
        //读取文件
        try{
            String encoding = "UTF-8";
            File file = new File(filePath);
            if(file.exists() && file.isFile()){
                InputStreamReader read = new InputStreamReader(new FileInputStream(file),encoding);
                BufferedReader bufferedReader = new BufferedReader(read);
                String lineTxt = null;
                while((lineTxt = bufferedReader.readLine())!=null){
                    //System.out.println("lineTxt: "+ lineTxt);

                    for(int i = 1 ; i < lineTxt.length() ; i++){
                        String temp1 = lineTxt.substring(i-1,i);

//                      if(i == 1){
//                          System.out.println("temp1: " + temp1);
//                      }
                        //判断是否符合标签的第一个要求，以<开头
                        if(temp1.equals("<")){

                            for(int j = i+1 ; j < lineTxt.length() ; j++){


                                String temp2 = lineTxt.substring(j-1,j);

//                              if(j==2){
//                                  System.out.println("temp2: " + temp2);
//                              }
                                //具体的一些判断
                                boolean ano_con = false;
                                if(temp2.equals("!") || temp2.equals("/") || temp2.equals("！") || temp2.equals("'") || temp2.equals(" ")){
                                    ano_con = true;
                                }
                                if(j == i+1 && ano_con){
                                    i = j - 1;
                                    break;
                                }else if(temp2.equals("'")){
                                    i = j - 1;
                                    break;
                                }else if(temp2.equals(" ")||temp2.equals(">") || temp2.equals("/")){
                                    boolean have = false;
                                    String temp3 = lineTxt.substring(i,j-1);
                                    //System.out.println("temp3: " + temp3);
                                    for(String key:map.keySet()){
                                        if(key.equals(temp3)){
                                            map.put(key, map.get(key)+1);
                                            have = true;
                                            i = j-1;
                                            break;
                                        }
                                    }
                                    boolean blank = true;
                                    if(temp3.equals("\r") || temp3.equals("//") || temp3.equals("\\") || temp3.equals("\"")){
                                        blank = false;
                                    }
                                    if(!have && blank){
                                        map.put(temp3, 1);
                                        i = j -1;
                                    }
                                    break;
                                }
                            }

                        }

                    }
                }
                read.close();
            }else{
                System.out.println("cant find the file!");
            }
        }catch(Exception e){
            System.out.println("read file error!");
            e.printStackTrace();
        }
        return map;
    }
    public static void main(String[] args){
        String path = "D:\\eclipse\\WebJava2\\vacation.htm";
        Map<String , Integer> answerMap = new HashMap<String,Integer>();

        answerMap = SelectTags(path);

        Formatter f = new Formatter(System.out);
        //key和value分别输出20和15个字符长度，-号表示左边对齐
        f.format("%-20s %-15s\n", "key", "value");
        for(String key:answerMap.keySet()){

            f.format("%-20s %-15s\n", key, answerMap.get(key));
        }
    }

}