Title here
Summary here
Command Line Arguments in Co-array Fortran
Command-line arguments are a common way to parameterize execution of programs. For example, ./program arg1 arg2 uses arg1 and arg2 as arguments to the program executable.
program main
use iso_fortran_env
implicit none
character(len=:), allocatable :: arg
integer :: i, num_args
! Get the number of command-line arguments
num_args = command_argument_count()
! Print all arguments, including the program name
do i = 0, num_args
call get_command_argument(i, length=len)
allocate(character(len) :: arg)
call get_command_argument(i, value=arg)
print *, arg
deallocate(arg)
end do
! Print arguments without the program name
do i = 1, num_args
call get_command_argument(i, length=len)
allocate(character(len) :: arg)
call get_command_argument(i, value=arg)
print *, arg
deallocate(arg)
end do
! Get individual argument (4th argument, if it exists)
if (num_args >= 3) then
call get_command_argument(3, length=len)
allocate(character(len) :: arg)
call get_command_argument(3, value=arg)
print *, arg
deallocate(arg)
else
print *, "Not enough arguments provided"
end if
end program mainIn Fortran, we use the command_argument_count() function to get the number of command-line arguments, and get_command_argument() to retrieve individual arguments. The first argument (index 0) is the program name.
To experiment with command-line arguments, compile the program first:
$ gfortran -o command_line_args command_line_args.f90
$ ./command_line_args a b c d
./command_line_args
a
b
c
d
a
b
c
d
cNote that Fortran arrays are typically 1-indexed, but command-line arguments are 0-indexed (with 0 being the program name). We’ve adjusted our code to account for this.
Next, we’ll look at more advanced command-line processing with additional Fortran features.