Very Computer

for example: if I had a CSH script:

#! /bin/csh -f

foreach file ($argv[1-])
    if(-e $file) then
        cat $file |first_pp|second_pp|third_pp|fourth_pp >$file.dat
        if($status) then
            echo compilation failed
            exit 1
        endif
    endif
end

The problem is that the $status only reflects the exit status of
"fourth_pp".
I need to exit with errors if "first_pp", or "second_pp", or "third_pp" also
exited abnormally.

I could do each process separately and make intermediate files, but the file
sizes are too large in-between, and that seems like a kludge. Is there an
elegant way ?

My only other way is to have each preprocessor create an error file, and
detect its existence at the end. That also seems like a kludge.

Is there a way to do this? another Shell type that support this.

#! /bin/csh -ef  does not do this either.

HELP!

please email:

 Bitter> I have a script which will build some source code by
 Bitter> filtering it first through a set of preprocessors first.

 Bitter> for example: if I had a CSH script:

that's your first mistake. Don't ever write scripts in csh.

The problem is solveable in Bourne/POSIX type shells, though the
contortions required are somewhat painful. Using bash it is trivial
thanks to the PIPESTATUS array variable.

In a POSIX type shell one solution is along these lines:

results=$( ( ( prog_1 <$infile || echo "prog_1 failed: $?" >&3)
             | (prog_2 || echo "prog_2 failed: $?" >&3)
             | (prog_3 || echo "prog_3 failed: $?" >&3) >$outfile
           ) 3>&1 )
if [ -n "$results" ]
then
    # something failed
    echo "processing failed:"
    echo "$results"
    exit 1
fi

--
Andrew.

comp.unix.programmer FAQ: see <URL: http://www.erlenstar.demon.co.uk/unix/>
                           or <URL: http://www.whitefang.com/unix/>

Quote:>The problem is solveable in Bourne/POSIX type shells, though the
>contortions required are somewhat painful. Using bash it is trivial
>thanks to the PIPESTATUS array variable.

>In a POSIX type shell one solution is along these lines:

>results=$( ( ( prog_1 <$infile || echo "prog_1 failed: $?" >&3)
>             | (prog_2 || echo "prog_2 failed: $?" >&3)
>             | (prog_3 || echo "prog_3 failed: $?" >&3) >$outfile
>           ) 3>&1 )
>if [ -n "$results" ]
>then
>    # something failed
>    echo "processing failed:"
>    echo "$results"
>    exit 1
>fi

  STAT1=/tmp/stat.$$.1
  STAT2=/tmp/stat.$$.2
  STAT3=/tmp/stat.$$.3

  rm -f $STAT1 $STAT2 $STAT3
  trap "rm -f $STAT1 $STAT2 $STAT3" 0 1 2 3 15

  (prog1; echo $? > $STAT1) < infile |
  (prog2; echo $? > $STAT2) |
  (prog3; echo $? > $STAT3) > $outfile

  # each status is now available in the corresponding file

A similar trick might work in csh, but programming is a lot easier
in virtually any other shell. :-)
--
In-Real-Life: Chris Torek, Wind River Systems (BSD engineering)

The solution (although awkward) is possible in sh but impossible in csh.
The solution is to do this, given A|B, where you want the exit status of A.

((A 2>&1 1>&3 3>&- 4>&- ; echo $? >&4)| B 1>&2 3>&- 4>&- ) 4>&1

Cheers,
Dave.