J3/01-341r1

Date: 31-Aug-2001
To:   J3
From: Subgroup C
Subj: Command line and environment intrinsics

1) The first dummy argument, NAME, of the GET_ENVIRONMENT_VARIABLE
intrinsic subroutine is labeled in the Arguments description as
optional, but the notation in [297:1] has the argument NAME outside
the square brackets. A call to the subroutine without the NAME
argument would seem to serve little purpose. The NAME argument should
not be optional.

Edit: [297:5] Remove '(optional)'

2) The GET_COMMAND and GET_COMMAND_ARGUMENT intrinsics both have
intent(out) character variables which may be too small to hold the
intended data.  In that case, the data is truncated and the leading
characters are assigned to the corresponding actual argument.  This is
not flagged as an error. The user can detect this situation by testing
(len(arg) < length). The user should also check (status /= 0).  The
analogous POSIX pxf... routines return a non-zero status for the case
that the data length was too large for the provided actual argument.
Adopting this convention here would simplify usage of the routines
since only the status value needs to be checked. I would propose using
negative status values to indicate the actual argument length is too
small, and using positive values to indicate an error. This is similar
to the status values returned by I/O routines.

Edit: [295:41-43] Replace the last two sentences with "It is assigned
the value -1 if the length of the COMMAND
argument is less than the significant length of the command. It is
assigned a processor dependent positive value if the command retrieval
fails.  Otherwise it is assigned the value 0."

Edit: [296:33-35] Replace the last two sentences with "It is assigned
the value -1 if the length of the VALUE
argument is less than the significant length of the command argument
specified by NUMBER. It is assigned a processor-dependent positive
value if the argument retrieval fails.
Otherwise it is assigned the value 0."

3) The GET_ENVIRONMENT_VARIABLE intrinsic subroutine has an even more
peculiar error detection process. Using the LENGTH argument for either
a size or an error flag is a very different convention from that used
in the command line routines. In addition, this routine suffers from
the defect noted above of not flagging a short actual argument as an
error. A status argument should be added to this routine and the
argument meanings should be make similar to those for
GET_COMMAND_ARGUMENT.

Edits:

[297:1] Replace '<<LENGTH,>>' with '<<LENGTH, STATUS,>>'

[297:13-19] Replace the final four sentences with "If the specified
environment variable exists and has a value, LENGTH is set to the
length of that value. If the specified environment variable does not
exist, or if it exists and has no value, LENGTH is set to zero."

[297:19+] "STATUS (optional) shall be scalar and of type default
integer. It is an INTENT(OUT) argument. If the environment variable
exists and either has no value or its value is sucessfully assigned to
VALUE, STATUS is set to zero.  It is the value -1
if the length of the VALUE argument is less than the
length of the environment variable value. It is assigned the value
1 if the specified environment variable does not exist, and 2 if
the processor does not support environment variables.
Processor-dependent values greater than 2 may be returned for other
error conditions.