应用ICTCLAS的中文分词小程序

最新推荐文章于 2020-10-30 00:56:33 发布

monkey123apple

最新推荐文章于 2020-10-30 00:56:33 发布

阅读量430

点赞数

分类专栏：文本分类（svm）文章标签：程序

本文链接：https://blog.csdn.net/monkey123apple/article/details/8515180

版权

文本分类（svm）专栏收录该内容

4 篇文章 0 订阅

订阅专栏

package org.liwei.dm;

import java.io.BufferedReader;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.PrintStream;

import org.ictclas4j.bean.SegResult;
import org.ictclas4j.segment.SegTag;

public class Divid {

public static void main(String[] args) throws Exception{
  File file = new File ("c:" + File.separator + "dm" + File.separator + "dividsource");
  String path[] = file.list();
  System.out.println(path.length);
  PrintStream out = System.out;
  for (int i = 0; i<path.length; i++){

   System.out.println(path[i]);
   BufferedReader buf = null;
   File f = new File ("c:" + File.separator + "dm" + File.separator + "dividsource"+ File.separator + path[i]);
   System.setIn(new FileInputStream(f));
   InputStream input = System.in;
   buf = new BufferedReader(new InputStreamReader(input));
   char b[]=new char[100000];
   int a=0,len=0;
   while((a=buf.read(b,0,100000))!=-1){
    len = a;
   }
   SegTag st = new SegTag(1);
   SegResult sr = st.split(new String(b,0,len));
   File s = new File ("c:" + File.separator + "dm" + File.separator + "dividresult"+ File.separator + i+".txt");
   PrintStream ps=new PrintStream(new FileOutputStream(s));
   System.setOut(ps);
   System.out.println(sr.getFinalResult());
   System.setOut(out);
   buf.close();
   buf=null;
  }