package cn.edu.bjut.textmining.chapter2;

import cn.edu.bjut.textmining.util.FileUtil;
import cn.edu.bjut.textmining.util.WordListUtil;

import java.nio.file.Path;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
import java.util.Set;

public class StopWordsDemo {
    public static void main(String[] args) {
        try {
            List<String> words = Arrays.asList("this", "is", "text", "mining", 
            		"in", "java", "book");
            Path stopwordFile = FileUtil.path(
                    "data", "chapter2", "basic-english-stopwords.txt");
            Set<String> stopWords = WordListUtil.readLowerCaseSet(stopwordFile);

            List<String> keywords = new ArrayList<String>();
            for (String word : words) {
                if (!stopWords.contains(word)) {
                    keywords.add(word);
                }
            }
            System.out.println(keywords);
    
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}