in reply to Re^3: read the filename column
in thread read the filename column

I did not notice, nor did I check. I ran the code with narrower fields of my own choosing in order to have more viewer-friendly output. But the question remains: Would not 'A*' be a better unpacker for the final (i.e., right-most) field?

Replies are listed 'Best First'.
Re^5: read the filename column
by Anonymous Monk on Jul 09, 2011 at 22:01 UTC

    But the question remains: Would not 'A*' be a better unpacker for the final (i.e., right-most) field?

    Better how? Yes, its more generic, but the record is fixed-width.

      ... the record is fixed-width.

      Yes, but fixed to what width?

      That's the problem with specification-by-example, often used in PM postings because it is simply too tedious to include a complete, formal specification. One can infer from the example data of the OP that the width of the final field is 8 or 9 characters and is always 'roadtest'. (GrandFather uses 9; I'm not sure how you settled on 10.) But if this field is always 'roadtest' (i.e., is a constant), why bother to parse it at all?

      That's when my PP (Programmer's Paranoia) kicked in and I thought it would be better to assume only that the final field would have no embedded whitespace (made originally for my variable field width solution) and so 'A*' would be safer, therefore better.

        That's the problem with specification-by-example ...That's when my PP (Programmer's Paranoia) kicked in and I thought it would be better to assume only...

        FWIW, I think its much ado about nothing :) the original poster learned about unpack, whether the example cut/paste works for OPs actual data doesn't matter -- the OP now knows enough to 1) use unpack 2) ask a follow up question