Question

以下是单个字符串的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();
}

整个文本文件的OpenNLP句子检测API

1 个答案: