htmlfilter.java

一个朴素贝叶斯文本分类

/**
 * <p>Title: </p>
 *
 * <p>Description: </p>
 *
 * <p>Copyright: Copyright (c) 2005</p>
 *
 * <p>Company: </p>
 *
 * @author not attributable
 * @version 1.0
 */
public class HtmlFilter {
  public HtmlFilter() {
  }
  static boolean bTag=false;
  public static String TagRemove(String stringLine){
    char []des=new char[1000];
    char []src=stringLine.toCharArray();
    int i=0,j=0;
    while(i<src.length){
      if(bTag){
        if(src[i]=='>'||src[i]=='}'){
          bTag = false;
          des[j++]=' ';
        }
        i++;
      }
      else{
        if(src[i]=='<'||src[i]=='{')
          bTag=true;
        else{
          des[j]=src[i];
          j++;
        }
        i++;
      }
    }
    String str;
    if(j>0)str=String.copyValueOf(des,0,j);
    else str="";
    return str;
  }
}