Thanks very much....jay

What is lrecl?  It isn't an awk (or unix) term I've heard of.

If you can tell us what you want done in words we understand, then
perhaps we can tell you how to do it with awk.

Showing what things would like before and after is very helpful
in explaining what wou want to do.

Chuck Demas
Needham, Mass.

--
  Eat Healthy    |   _ _   | Nothing would be done at all,

  Die Anyway     |    v    | That no one could find fault with it.

Maybe inapt, but LRECL is an IBM mainframe term meaning logical record
length. You've lived a blessed life if LRECL, JCL, DSN and DDN aren't
seared into your memory.

To the original poster: are you doing this on a mainframe? If so, would
you be using awk?? If so, then you should be able to read each record
from the file with the extended record length into awk's $0 and write
sequential substrings of specified maximum length to the new file with
the shorter record length. Presumably you'd want to leave words and
numbers intact, so you should split into substrings after the rightmost
whitespace character to the left of your output record length.

BEGIN { newlen = 72 }
{ s = $0
  do {
    ss = substr(s, 1, newlen)
    sub(/[^ \t]*$/, "", ss)
    printf("%-*s", newlen, ss)
    s = substr(s, length(ss))
  } while (length(s))

If you have a mainframe awk, you may have a mainframe version of unix's
fold command, which would be better suited to this task. Then, since
this isn't exactly an uncommon operation on mainframes, isn't there a
system command or standard REXX or CLIST script to do this?

If you're not on a mainframe, then odds are very good you have fold. Use
it to do this instead of awk.

Sent via Deja.com http://www.deja.com/
Before you buy.

     s = substr(s, length(ss) + 1)

Sent via Deja.com http://www.deja.com/
Before you buy.

--
Jim Monty

Tempe, Arizona USA

Sorry but I have no idea what you just said........jay

Thanks for the information.

As to application, I am writing reports from survey verbatum input files on
a PC, using SAS.

Thanks again.....jay

Why? Just a hunch, based on the morsel of information you posted
about your problem. dd is a utility that may or may not be readily
available to you, and may or may not be applicable to your problem.
There's no way for me to know.

See the dd(1) man page at URL <http://www.freebsd.org/cgi/man.cgi?
query=dd&sektion=1&apropos=0&manpath=FreeBSD+4.0-RELEASE%2fPorts>.

--
Jim Monty

Tempe, Arizona USA

Since the OP mentioned in another branch of this thread that this he's
trying to use awk in conjunction with SAS, I'm guessing his unix
background is slight to nonexistent. So "Use dd(1)" would have been as
helpful as unicode Mandarin rendered as ASCII. There are a few awks that
don't come with a full set of unix-like tools, so it's not appropriate
to take unix familiarity for granted.

For the OP: dd(1) refers to unix's/POSIX's dd command (the '(1)' means
it's in the manual's first section). This command allows users to run
low-level device I/O tasks. One of the things it can do is handle
different input and output block sizes. However, if the input file in
question is plain text, then dd can do stupid things when words or
numbers span output block boundaries. For variable length plain text
records, the unix/POSIX fold command will usually do a better job.

No doubt I'll be accused of pedantry.

Sent via Deja.com http://www.deja.com/
Before you buy.

You have uncanny insight into the problem.  I've just recently received a
book on AWK, suggesting I might find it helpful.  Well........
Again, UNIX is no BIG BLUE or P.C. so my knowledge is very limited where it
is concerned.

Your explanation was very helpful in abling me to make a decision as to how
to proceed.  Writing a routine in Clipper or C++ will be alot more
productive to the end result of having a parameter driven command line
utility.  I was simply looking for an AWK program to wrap lines and allowing
the passing of parameters (filein, fileout, lrecl) from the command line.  I
felt that it would give me a more immediate understanding of the AWK
language.

Maybe you could suggest a GOOD reference of  tools/templetes for AWK.  Any
help would be appreciated.

Thank you again for your help..........jay