用于整个文本文件的 OpenNLP 句子检测 API

Question

这是针对单个字符串的 OpenNLP Sentence Detector API 的代码：

package opennlp;

import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.InputStream;

import opennlp.tools.sentdetect.SentenceDetectorME;
import opennlp.tools.sentdetect.SentenceModel;

public class SentenceDetector {

    public static void main(String[] args) throws FileNotFoundException {
        InputStream modelIn = new FileInputStream("en-sent.zip");
        SentenceModel model = null;
        try {
           model = new SentenceModel(modelIn);  
        }
        catch (IOException e) {
          e.printStackTrace();
        }
        finally {
          if (modelIn != null) {
            try {
              modelIn.close();
            }
            catch (IOException e) {
            }
          }
        }
        SentenceDetectorME sentenceDetector = new SentenceDetectorME(model);
           String sentences[] = sentenceDetector.sentDetect(" First sentence. Second sentence.");

           for(String str : sentences)
               System.out.println(str);
    }
}

现在我的问题是如何传递整个文本文件并执行句子检测而不是单个字符串？

Answer 1

简单的方法：将整个文件作为字符串读取并以通常的方式传递。以下方法将文件内容读取为字符串：

public String readFileToString(String pathToFile) throws Exception{
    StringBuilder strFile = new StringBuilder();
    BufferedReader reader = new BufferedReader(new FileReader(pathToFile));
    char[] buffer = new char[512];
    int num = 0;
    while((num = reader.read(buffer)) != -1){
        String current = String.valueOf(buffer, 0, num);
        strFile.append(current);
        buffer = new char[512];
    }
    reader.close();
    return strFile.toString();
}