Reorder of fields

Shell Programming and Scripting
1

I need to reorder the fields of an input file. I was using the following awk statement but I am stumped on how to get it exactly as I need.

I need the output to be tab delimited except for the CITY, NAME and ID. Those last set of columns should be a single field separated by a space. And they are variable in length, ie, can be 1 or 5+ columns in length.

So basically fields 1-7 and the last 2 fields put in their proper positions tab delimited, then put everything else in the last field space delimited.

awk '{ print $1,$8,$2,$(NF-1),$(NF),$3,$4,$5$6$7,$1="",$2="",$3="",$4="",$5="",$6="",$7="",$8="",$0 }'

Input

500 XYZ3849 2 30 1964 09 17 ST CITY NAME AP ID 37 -84
500 XYZ3850 3 40 1998 10 27 ST CITY NAME AP 37 -84
500 XYZ3851 4 50 2013 05 7 ST CITY NAME 37 -84

Needed Output

500    ST    XYZ3849    37    -84    2    30    19640917    CITY NAME AP ID
500    ST    XYZ3850    37    -84    3    40    19981027    CITY NAME AP 
500    ST    XYZ3851    37    -84    4    50    20130507    CITY NAME

Any help is appreciated.

Thanks!

2

Where is the "KY" coming from?

You can "cheat" by pasting together multiple separators without commas. They'll be concatenated instead of separated by OFS.

awk -v OFS="\t" '{ print "tab", "separated", "space separated " $1 " " $2 " " $3 }' input

I don't know why you're printing all the $8="" and things. If you just leave them out, they won't be printed at all, you don't need to unset them.

$0 is a special variable meaning 'entire line'. Not sure why that's in there either.

3

Oversight that has since been corrected.

4 
awk -v OFS="\t" '{
        S=""
        for(N=9; N<=(NF-2); N++) S = S " " $N # Paste fields 9 - (NF-2)
        print $1, $8, $2, $(NF-1), $NF, $3, $4, $5 $6 $7, substr(S,2);
}' spacesep
5

for 2 digits $6 $7:

print $1, $8, $2, $(NF-1), $NF, $3, $4, $5 sprintf("%02d%02d",$6,$7), substr(S,2);
6

Thanks to both of you.

I incorporated both your suggestions and had to create a tab separation between

37 -84

This is what I think will work for the long term.

awk -v OFS="\t" '{
        S=""
        for(N=9; N<=(NF-2); N++) S = S " " $N # Paste fields 9 - (NF-2)
        print $1, $8, $2, $(NF-1),"\t"$NF, $3, $4, $5 sprintf("%02d%02d",$6,$7), substr(S,2);
}'
1 Like