Getting a number from (char)System.in.read

Question

I am very new to java so please forgive the seemingly basic question.
When I try to simply get a number for input into a program, and I use (char)Sysytem.in.read(),
it only gives me the acutual Keyboard number (ie.. 2 = 50, 1 = 49).
Even when I try to parse an Integer using :
public static void getNum( ) throws NumberFormatException
     char char1;
     String String1 = new String ("");
     int X;
     char1 = (char)System.in.read();
while ( char1 <= '0' && char1 <= '9' )      
     String1 = String1 + char1;
     char1 = (char)System.in.read();
     X = Integer.parseInt(String1);
System.in.read();
     return X;
     System.out.println(X);
This gives me a usable number but if I try to use a decimal point in the number and convert it to a double, it wont work because the decimal wont read into the (char)System.in.read();
Anyone have a suggestion ?

Answer 1

Hope this helps - although haven't tried out your code.
System.in.read() returns the next byte of information in the buffer, if the buffer is empty it'll prompt for input (passing control back to hardware/OS). Say the user types 10.56 and presses return - now the buffer has the information 10.56 plus a LF(line-feed) and and a CR(carriage return) character (this is how it is from DOS anyway) and control is passed back to your program which gets only the first byte of information from the buffer, which in this case would represent the character '1', only it doesnt get back the character '1', it gets an int type which represents the byte of information which represented the character '1' to your hardware/OS system (which I believe is decided by the machine local encoding system).
By casting this int to a char by the expression
(char) System.in.read()
its converted to the original character that the int represented from the buffer
If you were using System.in.read() only you would have to reconstruct the input as a String by concatenating the char's one by one only ignoring the last 2 bytes of the buffer (which captured the LF and CR).
Say you had the resulting String from the above reconstruction in the variable s1,
you could then attempt to convert the resulting String to a floating-point type (say a double) using
d = Double.parseDouble(s1)
where d was a double type (as mentioned by abnormal)
You'd have to put this in a try/catch statement because the user could type something invalid like 10.6a5
which would cause the parseDouble() method to throw an exception which unless you handle it yourself will cause your program to abend with a NumberFormatException.
This all seems to be a lot of hard work for input apparently so simple, which is why abnormal suggest using a BufferedReader to overcome the problem of 'manually' stripping the CR and LF from the input leaving you with just a String. I must admit I'm not sure of why abnormal's code wouldn't work for you. But I thought I'd post this anyway because you might find the background info helpful.