do_matching_including_no_literal_opts Subroutine

public subroutine do_matching_including_no_literal_opts(automaton, string, from, to)

Uses

    • forgex_utility_m

This procedure reads a text, performs regular expression matching using an automaton, and stores the string index in the argument if it contains a match.

Arguments

Type IntentOptional Attributes Name
type(automaton_t), intent(inout) :: automaton
character(len=*), intent(in) :: string
integer, intent(inout) :: from
integer, intent(inout) :: to

Source Code

   subroutine do_matching_including_no_literal_opts (automaton, string, from, to)
      use :: forgex_utility_m
      implicit none
      type(automaton_t), intent(inout) :: automaton
      character(*),      intent(in)    :: string
      integer,           intent(inout) :: from, to

      integer :: cur_i, dst_i ! current and destination index of DFA nodes
      integer :: ci           ! character index
      integer :: next_ci      ! next character index
      integer :: max_match    ! maximum value of match attempts
      integer :: start        ! starting character index
      integer :: i
      character(:), allocatable :: str

      str = string
      from = 0
      to = 0

      str = char(0)//string//char(0)

      cur_i = automaton%initial_index

      if (cur_i == DFA_NOT_INIT) then
         error stop "DFA have not been initialized."
      end if

      if (len(string) <= 1 .and. string == '') then
         if (automaton%dfa%nodes(cur_i)%accepted) then
            from = ACCEPTED_EMPTY
            to = ACCEPTED_EMPTY
         end if
         return
      end if

      loop_init: block
         i = 1
         start = i
      end block loop_init

      do while (start < len(str))
         max_match = 0
         ci = start
         cur_i = automaton%initial_index

         ! Traverse the DFA with the input string from the current starting position of ``cur_i`.
         do while (cur_i /= DFA_INVALID_INDEX)

            if (automaton%dfa%nodes(cur_i)%accepted .and. ci /= start) then
               max_match = ci
            end if

            if (ci > len(str)) exit

            next_ci = idxutf8(str, ci) + 1

            call automaton%construct(cur_i, dst_i, str(ci:next_ci-1))

            cur_i = dst_i
            ci = next_ci
         end do

         ! Update match position if a match is found.
         if (max_match > 0) then
            from = start-1
            if (from == 0) from = 1 ! handle leading NULL character.
            if (max_match >= len(str)) then
               to = len(string)
            else
               to = max_match-2
            end if
            return
         end if

         start = idxutf8(str, start) + 1 ! Bruteforce searching

      end do
   end subroutine do_matching_including_no_literal_opts