java文档常用英文,Java统计文档中英文单纯词个数

Java统计文档中英文单词个数

public class CountWords {

public static void main(String[] args) {

BufferedReader br = null;

try {

br = new BufferedReader(new FileReader("english.txt"));

} catch (FileNotFoundException e) {

e.printStackTrace();

}

StringBuffer sb = new StringBuffer();

String line = null;

try {

while((line = br.readLine()) != null) {

sb = sb.append(line);

}

} catch (IOException e) {

e.printStackTrace();

}

try {

br.close();

} catch (IOException e1) {

e1.printStackTrace();

}

Pattern pattern = Pattern.compile("[a-zA-Z']+");

Matcher matcher = pattern.matcher(sb);

Map map = new HashMap();

String word = "";

Integer num = null;

int total = 0;

while(matcher.find()) {

word = matcher.group();

total ++;

if(map.containsKey(word)) {

num = map.get(word);

num += 1;

} else {

num = 1;

}

map.put(word, num);

}

PrintWriter pw = null;

try {

pw = new PrintWriter(new FileWriter("result.txt"), true);

} catch (IOException e) {

e.printStackTrace();

}

Iterator iterator = map.keySet().iterator();

while(iterator.hasNext()) {

String tmp = iterator.next();

pw.println(tmp + " : " + map.get(tmp));

}

pw.println("total words : " + total);

pw.println("different words : " + map.size());

pw.close();

}

}