Re: wierd problem with go scripts

[Brent A McDaniel <gte328r@xxxxxxxxxxxxxxxxxxxxx>]

Ferret Users,

I described this wierd bug of trying to pass scripts to ferret via command
line.  Thanks to Ansley Manke, a solution has been found.  If you use the
-server option, the script will execute correctly.  The downside is that
you will not see the script run, only see a series of yes?'s.

This may only be a linux bug (a problem with readline?) but I wanted to
post it so that the solution will be in the mail archives for others.  You
still need to include the -batch option as well.  So to run a script, for
instance airlpreg.jnl, you would execute the following:

ferret -batch -server < airlpreg.jnl

Cheers,

Brent



Brent A McDaniel wrote:

> Hi Ferreters,
>
> I'm experiencing a wierd bug on my system that I wasn't able to find in
> the mail archives.  The basics of it are: if I try to call certain go
> scripts (.jnl) from the unix command line (ferret -batch < scriptname.jnl
> ), it completely fails, but if I enter the exact same commands manually that the
> script would have run, it works fine.  If I try to call the same script
> from within ferret (yes? go scriptname.jnl), it works fine.  This is a
> problem as I'm trying to use shell scripts to automate some data analysis.
> I've called .jnl files from the command line using ferret -batch <
> scriptname.jnl before and it's worked ok but the scripts were shorter in
> length.
>
> Here's the script that I'm calling:
> [gte328r@ao qdata2]$ more aanomreg.jnl
>
> set memory/size=100
> use aoindex.cdf
> use aanom.1958.nc
> use aanom.1959.nc
> use aanom.1960.nc
> let
> testa2=airanom[d=2,z=1000,t=1-jan-1958:28-feb-1958]*index[d=1,z=1000,gt=`airanom[d=2],return=GRID`@asn]
> let
> testb2=airanom[d=2,z=1000,t=1-dec-1958:31-dec-1958]*index[d=1,z=1000,gt=`airanom[d=2],return=GRID`@asn]
> let temp2=testa2[t=1-jan-1958:28-feb-1958@sum] +
> testb2[t=1-dec-1958:31-dec-1958@sum]
> let
> testa3=airanom[d=3,z=1000,t=1-jan-1959:28-feb-1959]*index[d=1,z=1000,gt=`airanom[d=3],return=GRID`@asn]
> let
> testb3=airanom[d=3,z=1000,t=1-dec-1959:31-dec-1959]*index[d=1,z=1000,gt=`airanom[d=3],return=GRID`@asn]
> let temp3=testa3[t=1-jan-1959:28-feb-1959@sum] +
> testb3[t=1-dec-1959:31-dec-1959@sum]
> let
> testa4=airanom[d=4,z=1000,t=1-jan-1960:28-feb-1960]*index[d=1,z=1000,gt=`airanom[d=4],return=GRID`@asn]
> let
> testb4=airanom[d=4,z=1000,t=1-dec-1960:31-dec-1960]*index[d=1,z=1000,gt=`airanom[d=4],return=GRID`@asn]
> let temp4=testa4[t=1-jan-1960:28-feb-1960@sum] +
> testb4[t=1-dec-1960:31-dec-1960@sum]
> let reg1 = temp2 + temp3 + temp4
> save/file=aanomtemp.nc reg1
> quit
>
> It's hard for me to debug what's going on as the error messages are
> garbled to say the least.  Here's what I get when I try to call the script
> from the command line:
>
> [gte328r@ao qdata2]$ ferret -batch < aanomreg.jnl
>         NOAA/PMEL TMAP
>         Program FERRET
>         Version 5.22 - 07/27/00
>         09-Apr-01 17:34
>
> yes? emp3 + temp4
>  **ERROR: unknown command: emp3
> yes? save/file=aanomtemp.nc reg1
>  **ERROR: variable unknown or not in data set: REG1
> yes? quit
>
> I'm running ferret v5.22 on a linux machine.  Has anyone else experienced
> this?  Is it a linux problem only?
>
> Thanks for any help.
>
> Brent
>
> --
> Brent A. McDaniel
>
> Dept of Earth and Atmospheric Sciences
> Georgia Institute of Technology
> Atlanta, Ga.  USA

--
Ansley Manke  Pacific Marine Environmental Laboratory  Seattle WA  (206)526-6246