java - How can I delete consecutive duplicates recursively without an int parameter? -

i need fix previous code. if input "aabbcdefgghijkllaa", output needs "abcdefghijkla". current code have changed outputs "alkjihgfedcba". (the input may not in alphabetical order, , output should in same order original input.) should create new method reverse or there way fix current methods?

here code:

package recursion;  public class recursion {      static string = "aabbcdefgghijkllaa";     static int b=a.length()-1;     static string result = "";      public static void main(string[] args){          system.out.println("input: " + a);         system.out.print("output: ");         removeduplicates(a);     }      public static string removeduplicates(string a){          if (b <= 0){             system.out.print(a.charat(0));             result=result+a.charat(0);         }         else if (a.charat(b) == a.charat(b-1)) {             b--;             removeduplicates(a);                       }         else {             system.out.print(a.charat(b));             result=result+a.charat(b);             b--;             removeduplicates(a);         }         return a;     } } 

changing b 0 (and - +) gives output of "a", not need. how can fix this?

it may better break specific goals instead:

• what want when string null or empty
• what want string of size < 2
• what want string of size >= 2

...and conditions each scenario.

note code follows untested; it's meant give rough idea not complete solution. works examples, definitely has shortcomings it.

---

let's think happens when example string "aabccdde". want keep track of first letter in string rest of string. wouldn't hurt have append either.

with that, come signature recursive method:

private static string deleteconsecutiveduplicates(string string,                                                   char lastseenletter,                                                   stringbuilder builder) 

this isn't expose anyone, cover method initial, "are empty/null?" checks.

static string deleteconsecutiveduplicates(string string) {      if(null == string || string.length() == 0) {         return string;     } else {         return deleteconsecutiveduplicates(string.substring(1),                                            string.charat(0),                                            new stringbuilder());     } } 

let's talk when see string.

round 1: 'a' | "abccdde" | ""

the character last saw same first character in string.

• add builder
• advance substring denoted index [2, n), , take "last seen letter" character @ position 1 of passed in string.

round 2: 'b' | "ccdde" | "a"

the 2 characters different.

• append b
• advance substring denoted index [1, n) , take "last seen letter" character @ position 0 of passed in string.

repeat these steps until left string "abcde".

---

the ideology here break problem smaller sub-problems easier digest , tackle. in general, you're ever adding 1 character @ time through last seen letter variable.

there other edge cases simple approach doesn't account (one letter, multiple repeated instances of it; i.e. "zzzzzzz" leave off one "z"), hope give somewhere start.

private static string deleteconsecutiveduplicates(string string, char lastseenletter, stringbuilder builder) {     if(string.length() == 0) {         builder.append(lastseenletter);     } else if(string.charat(0) == lastseenletter) {         builder.append(lastseenletter);         if(string.length() < 2 ) {             return builder.tostring();         }         return deleteconsecutiveduplicates(string.substring(2), string.charat(1), builder);     } else {         if(builder.charat(builder.length() - 1) != lastseenletter) {             builder.append(lastseenletter);         }         return deleteconsecutiveduplicates(string.substring(1), string.charat(0), builder);     }     return builder.tostring(); }