Re: netpbm vs. pbmplus

[Craig Sanders <cas@muffin.pronet.com>]

On Thu, 25 Jan 1996, Dale Scheetz wrote:

> Because of the way/reasons the lines are created, the algorith is
> simple:
>
> If a line is followed by a line that is the same, except for an
> additional path element, that line is a path, otherwise the last
> element of the line is a file.

a passable first approximation, but not a final solution.

As I mentioned immediately after asking the question, this method will
identify an empty directory as a file.

e.g. if the list of directories & filenames to be processed contains
an empty directory called, for example, /usr/lib/whatsit then your
algorithm will incorrectly identify it as a file.


Dirk came up with a neat generic solution to the question using perl.

Craig

--
  cas@muffin.pronet.com                                cas@muffin.apana.org.au
   *       Unix Consulting:  Installation, Configuration, & Support.        *
   * --- Also, contact me if you need your Dos/Win/OS2 LAN connected to --- *
   * --- the Internet.                                                  --- *