Why does the Integer.parseInt throw NumberFormatException on input that seems valid?

I'm doing a simple exercise from a book and I'm a little bit confused with how the java function parseInt works. I have read a line from an input file, used the StringTokenizer to split it and now I want to parse each part as an integer.

I have checked in the watch window that the input of the parseInt function is indeed a string which seems a valid integer (e.g. "35"). However, when I try to use the str.charAt function on my variable str holding the value "35", I get strange results :

str.charAt(0) == ""
str.charAt(1) == "3"
str.charAt(2) == ""
str.charAt(3) == "5"

This seems to be a problem probably somehow related to the encoding, so I have tried to fix it using this way of reading the file :

InputStreamReader reader 开发者_开发知识库= new InputStreamReader(new FileInputStream(inputfile), "UTF-8");

(I have explicitly saved the file using UTF-8 encoding in my editor), but this didn't help. Any ideas what could be the problem and how to fix it ?

EDIT : My sample

        InputStreamReader reader = new InputStreamReader(new FileInputStream(inputfile), "UTF-8");
        BufferedReader bfreader = new BufferedReader(reader);

        line = bfreader.readLine();
        while (line !=null)
        {
                String[] valueStrings = line.split(" ");
                String hole = valueStrings[0]; 

                int[] values = new int[4];
                for (int i = 0; i <values.length; i++){

                    String nr = valueStrings[i+1].trim(); 
                    values [i] = Integer.parseInt(nr);
                }

                // it breaks at the parseInt here, the rest is not even executed...

         }

My guess is that it's actually:

str.charAt(0) == '\0'
str.charAt(1) == '3'
str.charAt(2) == '\0'
str.charAt(3) == '5'

It sounds like it's probably actually saved in UTF-16 rather than UTF-8 - but if your text editor thought it was meant to save "null" characters, that would make sense. Try looking at the text file in a binary hex editor - I suspect you'll find that every other byte is 0.

If that doesn't help, please post a short but complete program which demonstrates the problem - so far we've only seen one line of your code.