PL SQL中如何去掉字符串中重复的字符

CREATE OR REPLACE FUNCTION remove_rame_string(oldstr VARCHAR2,
                                              p_sign VARCHAR2) RETURN VARCHAR2 IS
  /**************************************************** 
  ** Oracle去掉重复字符串 
  ** 函数名称：RemoveSameStr 
  ** 参    数：【名称】         【类型 】      【说明】 
  **           oldStr           varchar2       要处理的字符串 
  **           sign             varchar2       字符串分隔符 
  ** 返 回 值：l_RESULT         varchar2       不包含重复子串的记录 
  ****************************************************/
  str          VARCHAR2(2000);
  currentindex NUMBER;
  startindex   NUMBER;
  endindex     NUMBER;
  TYPE str_type IS TABLE OF VARCHAR2(30) INDEX BY BINARY_INTEGER;
  arr      str_type;
  l_result VARCHAR2(1000);
BEGIN
  -- 空字符串 
  IF oldstr IS NULL THEN
    RETURN('');
  END IF;
  --字符串太长 
  IF length(oldstr) > 2000 THEN
    RETURN(oldstr);
  END IF;
  str          := oldstr;
  currentindex := 0;
  startindex   := 0;
  LOOP
    currentindex := currentindex + 1;
    endindex     := instr(str, p_sign, 1, currentindex);
    IF (endindex <= 0) THEN
      EXIT;
    END IF;
  
    arr(currentindex) := TRIM(substr(str, startindex + 1, endindex - startindex - 1));
    startindex := endindex;
  END LOOP;
  --取最后一个字符串: 
  arr(currentindex) := substr(str, startindex + 1, length(str));
  --去掉重复出现的字符串: 
  FOR i IN 1 .. currentindex - 1 LOOP
    FOR j IN i + 1 .. currentindex LOOP
      IF arr(i) = arr(j) AND arr(i) = p_sign THEN
        arr(j) := '';
      END IF;
    END LOOP;
  END LOOP;
  str := '';
  FOR i IN 1 .. currentindex LOOP
    IF arr(i) IS NOT NULL THEN
      str := str || p_sign || arr(i);
      --数组置空: 
      arr(i) := '';
    END IF;
  END LOOP;
  --去掉前面的标识符: 
  l_result := substr(str, 2, length(str));
  RETURN(l_result);
END remove_rame_string;