In reply to @zithral
@zithral @DavidKButlerUoA I think the numbering normally goes in the order the matches happen, so \1 is the first group matched, \2 is the second, and so on
In reply to @christianp
@zithral @DavidKButlerUoA so if the input is "hello hell" and the regex is "(he)(l+)o \1\2", then \1 matches the other instance of "he" and \2 matches the other instance of "ll"