Need help figuring out what's going wrong

Question

With the explicit help/instructions of some users here I've managed to build some sort of a tokenizer.
There are a couple of mistakes in its rules which I am aware of, and probably it could be a lot more efficient, but I've found something funny. When copying a sentence from a WSJ article:
"Book publishers were locked in 11th-hour negotiations with Apple Inc. that could rewrite the industry's revenue model after the technology giant unveils its highly anticipated tablet device Wednesday."
I got:
Book
publishers
were
locked
in
11th-hour
negotiations
with
Apple
Incorporated
that
could
rewrin
exemplificationthe
industry
is
revenue
model
after
the
technology
giant
unveils
its
highly
anticipated
tablet
devin
exemplificationWednesday
./sd
Remarkably, it rewrites the "ite" from "rewrite" to "in exemplification", just as "ice" from "device".
Any thoughs as to why this mystery occurs?
Here's the code:
public static void main(String args[])
         InputStreamReader istream = new InputStreamReader(System.in) ;
         BufferedReader bufRead = new BufferedReader(istream) ;
         try {
              System.out.println("Enter input.");
              //read input
              String input = bufRead.readLine();
              //rewrite abbreviations
              input = input.replaceAll("dr.", "doctor");
              input = input.replaceAll("mr.", "mister");
              input = input.replaceAll("e.g.", "exempli gratia");
              input = input.replaceAll("vs.", "versus");
              input = input.replaceAll("i.e.", "in exemplification");
              input = input.replaceAll("Inc." , "Incorporated");
              //more abbreviations to come here
              //rewrite contractions
              input = input.replaceAll("'ll", " will");
              input = input.replaceAll("n't", " not"); // this one needs to be a bit smarter to cover cases like "won't"
              input = input.replaceAll("'s", " is"); // this one needs to be a lot smarter, since now it incorrectly rewrites "John's house" to "John is house"
              //perhaps rewrite some more stuff here
              String[] tokens = input.split("(?=\\p{Punct}(\\s|$))|\\s+");
              for(String t : tokens) {
                   //maybe rewrite back the stuff rewritten above?
                   //mark/tag sentence delimiters alreadys
                   if (t.equals(".") || t.equals("?") || t.equals("!") || t.equals(":") || t.equals(";")){
                        t = t + "/sd";
                   System.out.println(t);
         catch (IOException err) {
              System.out.println("Error reading line");
         }     

Answer 1

You're passing a regular expression into the replaceAll function (not just a normal String), so the "." you use in "i.e" really means "any character".
To mean literally a dot, you have to escape it: "i\\.e".
This problem will also happen in any of the other regular expressions with dots in them. So for example, "e.g." will replace "eggs", "edge", and "ergo" with "exempli gratia" as well.
Read all about it here !
PS- You might want to try using more meaningful post titles in the future...