public class TextFileExample {
static String[] readColumns(BufferedReader in) {
   LinkedList<String> columns = new LinkedList<String>();
   String s;
   while (!(s = in.readLine()).equals("~~~~~~")) {
       columns.add(s);   
   }
   return columns.toArray(new String[0]);
}
static String[] readRecord(BufferedReader in, String[] columns) {
   String[] record = new String[columns.length];
   for (int i=0;i<columns.length;i++) {
       record[i] = in.readLine();
       record[i] = record[i].substring(record[i].indexOf('=')+1);
   }
   return record;
}
public static void main(String[] args) throws Exception{
   BufferedReader in = new BufferedReader(new FileReader("test.txt"));
   String[] header = readColumns(in);
   String s;
   do {
       String[] record = readRecord(in, header);
       for (String key : record)
           System.out.println(key);
       System.out.println("------");
   } while ((s = in.readLine()) != null);
}
}

This is not very good code (it assumes that at least one record exists),
but it should give you a starting point.

You might want to play with making your returned object an array of
Chars and not Strings, John. (I.E. char[])
as well.  It's faster.

It is not clear exactly what your file contains .  are those ~~
actually in the file or did you just add tem to the post to logically
break up the groups?

In any case you read a line at a time.  See
http://mindprod.com/applets/fileio.html for sample code to do that.
You want buffered characters.

Then you look at the first char to decide if it is a ~~ or data line.

Then you use indexOf to find the =.

Then you use substring to graph the field name and value.  Then you
test the field value, to figure out what sort of a value you have.

--
Roedy Green Canadian Mind Products
The Java Glossary
http://mindprod.com