Shell Programming and Scripting

You can't. And, by the way, using OFS wouldn't solve your problem either. How about

Code:

while IFS=";" read -r -a arr; do printf "%013.f" ${arr[0]}; unset arr[0]; printf ";%s" ${arr[*]}; printf "\n"; done < file

Using awk printf %f works fine for this for numeric fields up to about 16 digits. If you have longer fields or if you have fields that contain alphanumeric strings, you could also try something like:

Code:

#!/bin/ksh
# Usage: script_name [ file [ min_field1_width ]]

# We don't want an old file hanging around if no lines are padded this time.
rm -f padded

# Pad first field to N characters with leading zeros.
awk -v N="${2:-13}" '
BEGIN {	FS = OFS = ";"
	Z = sprintf("%0*d", N, 0)
}
(L = length($1)) < N {
	$1 = substr(Z, 1, N - L) $1
	print > "padded"
}
1' "${1:-file}"

This was written and tested using a Korn shell but should work with any POSIX-conforming shell. If you want to try this on a Solaris/SunOS system, change awk to /usr/xpg4/bin/awk or nawk.

If you invoke this script with no arguments, it defaults to using file as the name of the file to be processed and 13 characters minimum for the first field producing the output you requested. If you invoke it with:

Code:

./script_name file 20

where file is the name of a file containing the input data you provided in post #1 in this thread, it will pad the 1st field of the contents of file with leading zeros to produce no less than twenty characters in that field:

Code:

00000000790069375798;94019;DIR-505/E;8;15.12;D-Link;CMT
00000004039784518926;94045;2.633-002.0;2;13.00;Kärcher;CMT
00000008806085027084;92798;CLT-W406/SEE;1;9.16;Samsung;CMT
00000005025232683680;94246;DMC-FZ200;10;251.26;Panasonic;CMT
00000000885370404449;94571;3LR-00001;10;30.25;Microsoft;CMT

and the file named padded will contain a copy of the above as well (since every line is padded with seven or eight zeros).

Without an array it becomes easier

Code:

while IFS=";" read -r a1 a0
do
  printf "%013.f;%s\n" "$a1" "$a0"
done < file