Fix autonomous write/read

At Davis, the robot went berserk during autonomous. The problem was due to 
unexpected digits at the beginning of lines in the recorded autonomous files. A 
partial workaround was implemented to skip those characters. We should 
implement a real fix now.

writeUTF is used to write each line of the file, with different separator 
characters between each pair of numbers/booleans.

Here's a short article explaining writeUTF and readUTF:
http://stackoverflow.com/questions/4009157/java-socket-writeutf-and-readutf

writeUTF encodes the string length in its output. That was the source of 
unexpected characters, since readUTF was not used to read the file.

In general, you should use matching methods with matching usage when 
writing/reading. If you use writeUTF to write, use readUTF to read. If you 
write one line at a time, read one line at a time. If you write a single string 
containing the whole file, do the same when reading.

We should also change to CSV format -- separate by commas and put a newline 
("\n") at the end. Unfortunately, our version of Java doesn't have 
String::split, and we can't use off-the-shelf CSV parsers, since they use other 
classes we don't have. However, you should be able to extract whole substrings 
using indexOf(char, startfrom) and substring(begin, end).

Original issue reported on code.google.com by briankgr...@gmail.com on 24 Mar 2013 at 10:09

proserven / team-100-wildhats

Fix autonomous write/read #96