The real world application of this exercise was to scripturally edit a load
of text configuration files in response to a change in specification.

Each config file consists of one record per line - not fixed length

The new specification didn't affect the first fifty-something characters of
each record, so they were to be retained intact.

Position fifty-something to end-of-line consisted of numerous 36 character
fields, plus a handful of trailing characters also not affected by the new
specification.

The new specification increased the 36 chars fields to 37 chars, requiring a
<space> to be appended as a filler.

The 'human' approach would be a nested 'for' loop - for each config file, do
for each record, do for each field . . .  etc

This 'human' is too lazy & impatient for such tomfoolery, when "there's got
to be a better way".

The environment is bash on Sco Unix.

gc

PS  yes - these config files were devised in the late 80's - long before
xml.



On 1 July 2011 15:24, Jim Cheetham <[email protected]> wrote:

> On Fri, Jul 1, 2011 at 2:28 PM, Kent Fredric <[email protected]>
> wrote:
> > Found an edge case:  abcd -> abcdZ
> >
> > https://gist.github.com/1057759
>
> Fantastic!
>
> Now I have to think about how to prevent it from happening (or
> actually, how to clean up afterwards), and whether preventing it is
> even necessary, given the original problem description.
>
> Hey Glenn, care to share some comments about what you were really
> doing with this? And what your limited platform is?
>
> -jim
>
> _______________________________________________
> Linux-users mailing list
> [email protected]
> http://lists.canterbury.ac.nz/mailman/listinfo/linux-users
>
_______________________________________________
Linux-users mailing list
[email protected]
http://lists.canterbury.ac.nz/mailman/listinfo/linux-users

Reply via email to