FIELDWIDTH & versions of awk

[Sivaram Neelakantan]

Hello,

FIELDWIDTH, I believe is specific to gawk which I find enormously
useful in dealing with fixed width fields without any field
delimiters. How do I simulate the same on other versions of awk?
Sometimes, I'm not allowed to load gawk on some sytems and hence this
portability question.

regards,
Sivaram

[Charles Demas]

In article <a7043d61.03072...@posting.google.com>,

use the substr() function to define fields.

e.g.:

a=$0
$1=substr(a,1,2)
$2=substr(a,3,3)
$3=substr(a,7,2)
.
.
.

Chuck Demas

--
Eat Healthy | _ _ | Nothing would be done at all,
Stay Fit | @ @ | If a man waited to do it so well,
Die Anyway | v | That no one could find fault with it.
de...@theworld.com | \___/ | http://world.std.com/~cpd

[Peter S Tillier]

[...]

When using awk on platforms where gawk is not available (for whatever
reason) and where I need to use fixed width fields I use the following
awk function (sorry that lines wrap in posting):

function setFieldsByWidth( i,n,FWS,start,copyd0) {
# Licensed under GPL Peter S Tillier, 2003
# NB corrupts $0
copyd0 = $0 # make copy of $0 to work on
if (length(FIELDWIDTHS) == 0) {
print "You need to set the width of the fields that you require" >
"/dev/stderr"
print "in the variable FIELDWIDTHS (NB: Upper case!)" >
"/dev/stderr"
exit(1)
}

if (!match(FIELDWIDTHS,/^[0-9 ]+$/)) {
print "The variable FIELDWIDTHS must contain digits, separated" >
"/dev/stderr"
print "by spaces." > "/dev/stderr"
exit(1)
}

n = split(FIELDWIDTHS,FWS)

if (n == 1) {
print "Warning: FIELDWIDTHS contains only one field width." >
"/dev/stderr"
print "Attempting to continue." > "/dev/stderr"
}

start = 1
for (i=1; i <= n; i++) {
$i = substr(copyd0,start,FWS[i])
start = start + FWS[i]
}

}

Note that the "/dev/stderr" entries in some lines have wrapped.

I then call setFieldsByWidth() in my main awk code as follows:

BEGIN {
FIELDWIDTHS="7 6 5 4 3 2 1" # for example
}
!/^[ ]*$/ {
saveDollarZero = $0 # if you want it later
setFieldsByWidth()
# now we can manipulate $0, NF and $1 .. $NF as we wish
print $0, NF
next
}

1

or something similar. It has worked OK for me quite a few times.

It's a workaround.

HTH
--
Peter S Tillier
"Who needs perl when you can write dc, sokoban,
arkanoid and an unlambda interpreter in sed?"