Incrementing characters being read - c++

Question

I am trying to decode an input file that looks something like this:
abbaabbbbaababbaabababaabababaabbababaabababababababa...
and compare it to a makeshift mapping I have made using two arrays
int secretNumber[10];
string coding[10];
coding[0]="abb";
coding[1]="aabbbba";
coding[2]="abab";
...
I am not sure how I can start off by reading the first character which is 'a' then check if it's in the coding array. If it is print out the secretCoding and move the next character b. Else if it's not in the array then add the next character to the first in a string and check to see if "ab" is in the array and if that isn't either add the next character which makes "abb" and so on.
Something like this:
while (!(readFile.eof()) ){
for(int i=0; i<10; i++){
if(stringOfChars==coding[i]){
cout << secretNumber[i] <<endl;
//Now increment to next char
}
else{
//combine the current string with the next character
}
}
}
Question: How do I go about reading in a character if its a match move to next character if not combine current character and the next character until there's a match.

Answer 1

You sould use a design pattern called interpreter.
Here is a link to a c++ version.

Answer 2

If you want a solution that works for arbitrary input sizes, i.e. which doesn't store the entire input in memory, then you can use a queue (e.g. std::deque<char>) to read in a handful of characters at a time, pushing data in from the back. Then you check if the queue still has three, four or five characters left, and if so compare them to your patterns; if there's a match, you pop the corresponding characters off from the front of the queue.

Answer 3

I'm not sure but perhaps it seems like you are trying to implement the LZW compression algorithm. If that is the case, then you would have to change your approach a little. If you have decided that your secret code are integers, then you would have to assign a code to all the elements of the initial contents of the dictionary. The initial dictionary is basically all the strings in your source alphabet of size 1. In your case it would be "a to z", or only "a" and "b" if you are keeping it simple.
The other thing is that you need to look through your dictionary for any existing string which has been assigned a code. The best way to do that is to use STL map container which could map strings to integers in your case. Also, its a good idea to place a restriction on the size to which the dictionary could grow as new strings continue to be added to it.
Overall,
Use std::map< std::string, int > dictionary; as your dictionary for strings such as a, b, aa, ab, aab, etc... and the matching code for it.
The coding[0], coding[1] would not be required as they strings would serve as the key in this dictionary.
The secretNumber[0], secretNumber[1] also would not be needed as the value would for a key would give the secretNumber.
Here is what it may look like:
std::map< std::string, int > dictionary;
int iWordCount = 0;
/*
Initialize the dictionary with the code for strings of length 1 in your source alphabet.
*/
dictionary["a"] = 0;
dictionary["b"] = 1;
iWordCount = 2; // We would be incrementing this as we keep on adding more strings to the dictionary.
std::string newWord = "", existingWord = "";
while (!(readFile.eof()) ){
/*
I'm assuming the next character is read in the variable "ch".
*/
newWord += ch;
if ( dictionary.count(newWord) != 0 ) { // Existing word.
/*
Do something
*/
}
else { // We encountered this word for the first time.
/*
Do something else
*/
}
}