[SGVLUG] shell script & nawk exposure

[Don Gibbs]

Bryan,

Turns out my data files have a time stamp embedded into the name: in 
this particular case the shell's wildcard expansion is effectively 
the same as 'ls -t'.  My initial solution had the markings of a 
ticking time-bomb!

Of course, your use of 'head' has negated my need for 'nawk'.  :)  At 
least for the moment.

Thanks for your useful reply!

   --Don

>Don Gibbs <donald.e.gibbs at jpl.nasa.gov> writes:
><snip>
>>    NAME=`echo ls -lrt *_data | nawk '{print $NF}'`
>>    echo $NAME
>
>Don,
>
>I don't think this is doing what you think it is.  The above does not
>run ls(1) at all.  It's just echoing those characters prior to all the
>filenames which are a result of the shell expanding the wildcard.  You
>might as well write it this way:
>
>   NAME=`echo *_data | nawk '{print $NF}'`
>   echo $NAME
>
>Given this, it should not be surprising that the echo separates the
>names with spaces rather than newlines, since that's what the shell is
>defined to do.  If you want to keep it all within the shell, you could
>do this:
>
>   NAME=`set *_data ; eval echo '$'{$#}`
>
>Please note: Each of the above will only do what you said you wanted
>assuming that the shell lists the files in a particular order.  To my
>knowledge, it lists them in the order they appear *in the directory*.
>This means that if you ever delete files and then add new ones, it may
>reuse slots and mess up your order assumptions.
>
>What I recommend doing instead, is:
>
>   NAME=`ls -t *_data | head -1`
>
>I also find this version clearer and there's a lot to be said for
>readability.  Both for myself looking at it again much later, and for
>anyone else looking at it.
>
>{Bryan}
>--
>Bryan D Howard
><bryan at alumni.caltech.edu>


-- 
========================================================================
Don Gibbs		| "Somehow it seems to fill my head with ideas
dgibbs at jpl.nasa.gov	| -- only I don't know exactly what they are!"
818 354-2990 - office	|
818 653-9531 - cell	|       Alice, after reading JABBERWOCKY

Sec 316, Flight S/W & Data Systems - Group B, GN&C and FSW Testing
Opinions expressed are not necessarily those of JPL/Caltech or NASA