汉语词频统计工具java,java处理文本计算词频工具类

主要代码：

1、工具方法类

public class WordFreqUtil {

public static List> wordFreqCount(String filePath,int sortType) throws IOException {

Map map = new TreeMap();  //存储单词计数信息，key值为单词，value为单词数

BufferedReader reader=new BufferedReader(new FileReader(filePath));

String readLine = null;

while((readLine = reader.readLine()) != null){

String [] wordArray = filterWord(readLine);

for(String word : wordArray) {

word = word.trim();

if (word != null && !word.equalsIgnoreCase("")) {

if(map.get(word) != null){

map.put(word, map.get(word) + 1);

}else{

map.put(word ,1);

}

}

}

}

return sortWordFreq(map,sortType);

}

private static List> sortWordFreq(Map map, final int sortType) {

List> list = new ArrayList>(map.entrySet());

Collections.sort(list, new Comparator>() {

@Override

public int compare(Map.Entry o1, Map.Entry o2) {

if(sortType==1){

return o1.getValue() - o2.getValue();  //升序

}else{

return o2.getValue() - o1.getValue();  //降序

}

}

});

return list;

}

public static String[] filterWord(String line){

line = line.toLowerCase();

return line.split("[^a-zA-Z]");

}

}

2、单词统计类

public class WordCount {     private int count;     public WordCount(int count){         this.count=count;     } }