JUnit test suite involving file reading is performing differently on Windows 10 and Fedora 23

Suppose my entire Eclipse project is in a directory called project and in project, there is a regular non-empty, well-formatted file, data.txt, and a single directory src. In src, there are two directories: code and test for packages named exactly the same as the directories.

In code, I have the following file A.java

package code;

import java.util.*;
import java.io.*;

public class A {
  private static Map<Integer, String> m = makeMap("data.txt");

  private static Map<Integer, String> makeMap(String file) {
    Map<Integer, String> m = new HashMap<Integer, String>();
    try {
      int i = 0;
      Scanner s = new Scanner(new File(file));
      while (s.hasNextLine()) {
        m.put(i, s.nextLine());
      }
    } catch (FileNotFoundException e) {
      e.printStackTrace();
      System.exit(1);
    }
    return m;
  }

  // a bunch of public static methods that use the global static variable m
}

In test, I have ATest.java, which is a JUnit test file for A.java. It just contains several test cases that test the public static functions in A.java.

Also in test, I have another file TestSuite.java, which is the following

package test;

import org.junit.runners.Suite;
import org.junit.runners.Suite.SuiteClasses;
import org.junit.runner.RunWith;

@RunWith(Suite.class)
@SuiteClasses({ ATest.class })

public final class TestSuite { }

If I were to run ATest on any machine, it will pass all the tests. If I were to run TestSuite on a Windows 10 machine, the tests in ATest will fail, but if I were run TestSuite on a Fedora 23 Linux machine, TestSuite passes all of the tests.

In particular, on Windows, the program is able to successfully make it all the way through the try block without throwing an exception, but the problem is that s.hasNextLine() returns false so it never reads the contents of the file. This is contrary on a Fedora, where s.hasNextLine() returns true and proceeds to do be able to perform actions on the global map. The thing that confuses me is that the data.txt is the same on both platforms, so why would one platform perform differently than the other?

Does anybody know why this is happening, and what I can do to make my code be platform independent?

EDITS: data.txt contains Japanese characters, so it is encoded in UTF-8 if that helps

You need to specify the encoding of the file that you are reading. If it is not specified, it will use the platform default which is not always UTF-8

This is done using the Scanner(File, String charset) constructor.