CODEWITHSUNDEEP
javastringfileopennlp
Saprativa Bhattacharjee
Saprativa Bhattacharjee

Reputation: 184

OpenNLP Sentence Detection API for entire text file

Here is the code for OpenNLP Sentence Detector API for a single String:

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);
    }
}

Now my question is how do I pass an entire text file and perform sentence detection instead of a single string?

Upvotes: 2

Views: 2445

Answers (1)

mishadoff
mishadoff

Reputation: 10789

Simple way: to read whole file as string and pas it in usual way. Following method reads file content as string:

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();
}

Upvotes: 2

Related Questions