How can I adjust the length of a column field in bash using awk or sed? - linux

Question

I've an input.csv file in which columns 2 and 3 have variable lengtt.
100,Short Column, 199
200,Meeedium Column,1254
300,Loooooooooooong Column,35
I'm trying to use the following command to achieve a clean tabulation, but I need to fill the 2nd column with a certain number of blank spaces in order to get a fixed lenght column (let's say that a total lenght of 30 is enough).
awk -F, '{print $1 "\t" $2 "\t" $3;}' input.csv
My current output looks like this:
100 Short Column 199
200 Meeedium Column 1254
300 Loooooooooooong Column 35
And I would like to achieve the following output, by filling 2nd and 3rd column properly:
100 Short Column 199
200 Meeedium Column 1254
300 Loooooooooooong Column 35
Any good idea out there about awk or sed command should be used?
Thanks everybody.

Answer 1

Use printf in awk
$ awk -F, '{gsub(/ /, "", $3); printf "%-5s %-25s%5s\n", $1, $2, $3}' file input.csv
100 Short Column 199
200 Meeedium Column 1254
300 Loooooooooooong Column 35
What I have done above, is set the IFS,field separator to ,; since the file has some white-spaces in the 3rd column alone, it mangles, how printf processes the strings, removing it with gsub and formatting in C-style printf.

Answer 2

Rather than picking some arbitrary number as the width of each field, do a 2-pass approach where the first pass calculates the max length of each field and the 2nd prints the fields in a width that size plus a couple of spaces between fields:
$ cat tst.awk
BEGIN { FS=" *, *"; OFS=" " }
NR==FNR {
for (i=1;i<=NF;i++) {
w[i] = (length($i) > w[i] ? length($i) : w[i])
if ($i ~ /[^0-9]/) {
a[i] = "-"
}
}
next
}
{
for (i=1;i<=NF;i++) {
printf "%"a[i]w[i]"s%s", $i, (i<NF ? OFS : ORS)
}
}
$ awk -f tst.awk file file
100 Short Column 199
200 Meeedium Column 1254
300 Loooooooooooong Column 35
The above also uses left-alignment for non-digit fields, right alignment for all-digits fields. It'll work no matter how long the input fields are and no matter how many fields you have:
$ cat file1
100000,Short Column, 199,a
100,Now is the Winter of our discontent with fixed width fields,20000,b
100,Short Column, 199,c
200,Meeedium Column,1254,d
300,Loooooooooooong Column,35,e
$ awk -f tst.awk file1 file1
100000 Short Column 199 a
100 Now is the Winter of our discontent with fixed width fields 20000 b
100 Short Column 199 c
200 Meeedium Column 1254 d
300 Loooooooooooong Column 35 e

Answer 3

Solution using perl
$ perl -pe 's/([^,]+),([^,]+),([^,]+)/sprintf "%-6s%-30s%5s", $1,$2,$3/e' input.csv
100 Short Column 199
200 Meeedium Column 1254
300 Loooooooooooong Column 35