在字符串匹配后从下一行将值读入 Fortran 数组

Read values into a Fortran array from the next line after a string match

我正在尝试解析 Abaqus 输入文件并将所需的节点读入数组。我首先尝试比较一个字符串,如果它匹配,那么我想从下一行读入未知数量的整数值。

文本文件的示例如下所示。我想读取 nset=bf1:

下的值
** PARTS

*Part, name=Beam

*Nset, nset=bf1, instance=Beam-1

    1,    2,    5,   43,   45,   48,   49,   50,  105,  106, 1189, 1190, 1191, 1192, 1193, 1194

    5275, 5276, 5277, 5278

*Elset, elset=_Surf-3_SNEG, internal, instance=Beam-1

    1,    2,    3,    4,    5,    6,    7,    8,    9,   10,   11,   12,   13,   14,   15,   16

    17,   18,   19,   20,   21,   22,   23,   24,   25,   26,   27,   28,   29,   30,   31,   32

我现在正在做的是这个

program nset_parser
    implicit none

    integer, parameter :: rk = selected_real_kind(10,40)
    real(kind=rk), dimension(:), allocatable :: values
    integer ::  ios,l,val, n, nlines
    character(len=80) :: file_name = 'code_check.inp'
    character (len=*), parameter :: search_str = "bf1"
    character(len=5096) :: input_line, line
    character(len=10) :: keyword, get_keyword
    logical :: h_nset=.false.

     
    l = len(search_str)
     
    open(unit=10, file=file_name, status="old", action='read') ! open file
    do 
     read(10, '(a)',iostat=ios) input_line   ! read the lines
     if (ios /=0) exit
     input_line = adjustl(input_line)   
    
     if (input_line(1:1) == '*' .and. input_line(2:2) /= '*') then  ! check if it is a comment line or data line
      keyword=get_keyword(input_line)
      if (trim(keyword) == 'Nset') then
          read(input_line,'(a)') line
          val = index(line, search_str)
          if (val /= 0) then
              h_nset=.true.
              print*, line(val:val+l-1)
          end if
      end if
     end if
    end do

  close(10)
  end program nset_parser

但在这之后我有点难过。

我重写了我的答案以使其更笼统。此代码将处理带有“nset=bf1”的行与任何其他以“*”开头的行之间的多行整数。

因为整数行的长度可能不同,所以二维整数数组是不合适的。相反,我将所有整数存储在一维数组中,并存储每个新行的数组索引。

为避免读取所有内容两次,如有必要,我会即时增加数组的大小。

请注意,如果您的编译器不支持 MOVE_ALLOC (Fortran 2003),那么您只需将每个调用替换为 'deallocate, allocate, deallocate' 集。

program nset_parser
    implicit none

    INTEGER ::  ios,l,val, i, n, nlines, comma_idx, next_idx
    INTEGER, ALLOCATABLE :: my_ints(:), line_starts(:), work_ints(:)
    character(len=80) :: file_name = 'code_check.inp'
    character (len=*), parameter :: search_str = "bf1"
    character(len=5096) :: input_line
    character(len=10) :: keyword
    logical :: h_nset=.false.

    l = len(search_str)

    ALLOCATE(line_starts(10), my_ints(1000))
    next_idx = 1
    nlines = 0

    open(unit=10, file=file_name, status="old", action='read') ! open file
    DO 

      ! Read and adjust the line
      read(10, '(a)',iostat=ios) input_line
      if (ios /=0) exit
      input_line = adjustl(input_line)   

      ! Ignore blank lines
      IF(LEN(TRIM(input_line))==0) CYCLE

      ! This section picks up any line beginning *
      ! and either sets or unsets h_nset
      IF (input_line(1:1) == '*' .AND. input_line(2:2) /= '*') THEN  ! check if it is a comment line or data line

        ! Use of keyword below is a little unsafe if
        ! keywords can be longer than 10 chars (the length of 'keyword')
        ! Instead, better to check if input_line(2:5) == Nset
        ! I left it like this in case you want to be able to check multiple
        ! keywords or something.
        comma_idx = INDEX(input_line, ',')
        keyword = input_line(2:comma_idx-1)

        IF (TRIM(keyword) == 'Nset') THEN
          val = INDEX(input_line, search_str)
          IF (val /= 0) THEN
            h_nset=.TRUE.  ! Switch on when nset=bf1 line is found
            CYCLE
          END IF
        END IF
        h_nset = .FALSE.  ! Only reach this point if we are in a * line which isn't nset=bf1
      END IF

      IF(h_nset) THEN
        n = COUNT(TRANSFER(input_line, 'a', LEN(input_line)) == ",") + 1 !cast to chars and count occurrence of comma
        nlines = nlines + 1

        ! Double size of arrays on the fly if necessary
        IF(nlines > SIZE(line_starts)) THEN
          ALLOCATE(work_ints(nlines * 2))
          work_ints(1:SIZE(line_starts)) = line_starts
          CALL MOVE_ALLOC(work_ints, line_starts)
        END IF
        IF(next_idx+n > SIZE(my_ints)) THEN
          ALLOCATE(work_ints(2*SIZE(my_ints)))
          work_ints(1:SIZE(my_ints)) = my_ints
          CALL MOVE_ALLOC(work_ints, my_ints)
        END IF

        line_starts(nlines) = next_idx
        READ(input_line, *) my_ints(next_idx:next_idx+n-1)
        next_idx = next_idx + n
      END IF
    END DO
    close(10)

    ! This helps with the printing below
    line_starts(nlines+1) = line_starts(nlines) + n

    DO i=1, nlines
      PRINT *, my_ints(line_starts(i):line_starts(i+1)-1)
    END DO
  end program nset_parser