Re: bash Scripting

[Steve Scherbinski <sski@xxxxxxxxxxxxxxxx>]

I've seen scripts that you would run a file through and it would convert
any number
of consecutive spaces into a single colon (:), you could then use cut
and change the
field delimiter to the colon and get the fields you want no matter what
the size.  I
believe the O'Reillys Unix Power Tools book probably has it, if not I'm
sure it's 
around on the web somewhere.  Let me see if I can find the version I
have at work
tomorrow.  I think it's on one of my backup cd's I have their.

Steve


Jason Hackney wrote:
> 
> Greetings all,
> 
> I have been trying to create a feature in my backup script that will
> scan my backup medium (Zip disk) and retrieve filesize and filename,
> sorted by the oldest files first. The command `ls -rst1` seems to
> provide just the information I seek. However, I can't seem to get the
> info. into a workable variable(s). I've tried using the `cut` command,
> but that is too inconsistent. I would like to be able to put the
> filesize into one array and the filename into another array ala
> `filesize[5] = 12391`, and `filename[5] = backup020401.bz2`. Then I
> would do calculations based on the filesize to determine how many files
> to delete to free up space for a recently created backup.
> 
> I'm wearing out my O'Reilly bash book (which by the way is an excellent
> reference, but perhaps a bit too concise for a rank beginner like
> myself), and I must surely be overlooking something in the man pages ...
> anyway does anyone have a better suggestion on how to make this work
> with variable file sizes? I simply want to capture the filesizes and
> filenames keeping them mutually dependant, but separate so calculations
> can be performed on the filesizes. `cut` does work wonderful if the
> filesize is consistent in that it takes up the same number of columns
> for an entire directory listing. If I have a 1 KB file and a 1 MB file,
> then `cut` goes out the door.  Thank you in advance,
> 
> Jason