Andy, , has a dream - he wants to produce his very own dictionary. This is not an easy task for him, as the number of words that he knows is, well, not quite enough. Instead of thinking up all the words himself, he has a briliant idea. From his bookshelf he would pick one of his favourite story books, from which he would copy out all the distinct words. By arranging the words in alphabetical order, he is done! Of course, it is a really time-consuming job, and this is where a computer program is helpful.

You are asked to write a program that lists all the different words in the input text. In this problem, a word is defined as a consecutive sequence of alphabets, in upper and/or lower case. Words with only one letter are also to be considered. Furthermore, your program must be CaSe InSeNsItIvE. For example, words like "Apple", "apple" or "APPLE" must be considered the same.

The input file  lines. An input line has at most  characters. Input is terminated by EOF.

Your output should give a list of different words that appears .

 #include<iostream>
 #include<string>
 #include<set>
 #include<sstream>
 using namespace std;  

 set<string> dict;//set up a set called dict-short for dictionary,and it's based on string type;  

 int main(){
     string s,buf;
     while (cin>>s){
         ;i<s.length();i++)
           if (isalpha(s[i])) s[i]=tolower(s[i]);//if it's an letter,turn it lowercase.Others turn space and use stringstream to leave it out=ignore it
           else s[i]=' ';
         stringstream ss(s);
         while (ss>>buf) dict.insert(buf);//insert it into the set which is already in order,TYkon said it's a balanced tree inside
     }
     for (set<string>::iterator it=dict.begin();it!=dict.end();++it)//iterator just like a point,scan it from beginning to end and output
       cout<<*it<<endl;//NOTICE output by point
     ;
 }

 for (set<string>::iterator it=dict.begin();it!=dict.end();++it)//iterator just like a point,scan it from beginning to end and output
       cout<<*it<<endl;//NOTICE output by point