excel spreadsheet ... main workbook

with t(ord, str) as
(select 1, 'qwerty, asd' from dual union all select 2, 'hello, world' from dual)
, r(old_value, new_value) as
(select 'asd', 'xyz' from dual union all select 'hello', 'bye' from dual union all select 'world', 'war' from dual)
select str, result
from t
excel spreadsheet
  reference r on (select * from r) dimension by (old_value) measures (new_value)
main workbook
  partition by (ord) dimension by (0 dummy) measures (str, str result, str tmp)
  rules iterate (1e9) until (tmp[0] is null)
  (
    tmp[0] = regexp_substr(str[0], '\w+', 1, iteration_number + 1),
    result[0] = nvl2(r.new_value[tmp[0]], replace(result[0], tmp[0], r.new_value[tmp[0]]), result[0])
  )
/

STR          RESULT
------------ ------------
qwerty, asd  qwerty, xyz
hello, world bye, war