Set alignment of numeric columns when columnating data

Question

I use column -t to format data for easy viewing in the shell, but there seems to be no option to specify column alignment (e.g. align to the right).

Any Bash one-liners to do it? I have arbitrary number of columns.

Answer 1

UPDATE: added a script (not a one liner, though) which allows you to choose which columns you want justified... It caters for Left (default) and Right (not Center).. As-is, it expects TAB delimited fields. You can change the column output seperator via $S.

RJustCol=(2 3 5)  # Set columns to be right justified.
RJustRex=; Q=$'\x01'; R=$'\x02'; S=" | "
for r in ${RJustCol[@]} ;do  # Build the Right-justify regex.
  RJustRex="${RJustRex}s/^(([^\t]+\t){$((r-1))})([^\t]+)\t/\1\3$R\t/; "
done
sed -r "s/$/\tZ/g; s/^/$Q/; s/\t/\t$Q/g; $RJustRex" file |
  column -t -s $'\t' | 
    sed -r "s/  $Q/$Q/g; s/$Q([^$Q$R]*)$R([^$Q]*)/$S\2\1/g; s/$Q/$S/g; s/Z$//"

Typical output:

| The Lost Art       |   +1255 |  789 | Los                 |     -55 | 
| of the Idle Moment | -159900 | 0123 | Fabulosos Cadillacs | +321987 | 

---

Note:column doesn't work as you might expect when you have empty cells.

Option -n
        By default, the column command will merge multiple adjacent
        delimiters into a single delimiter when using the --t option;
        this option disables that behavior. This option is a Debian
        GNU/Linux extension.

---

From here on is the original answer which is related to but doesn't specifically address tne main issue of th question..

Here is the "one-liner" which suits integers (and allows +/- signs) .. The "X" place-holder forces column to right-pad the last cell.

sed 's/$/\tX/g' file |column -t |sed -r 's/([-+]?[0-9.]+)( +)/\2\1/g; s/^  //; s/X$//'

Typical output

  +1255  789  011      -55       34
-159900   33  022  +321987  2323566

---

If you have float values, or floats mixed with integers, or just integers, (optional leading +/- signs), a bit more shuffling works.

sed -r 's/$/\tX/; 
        s/([-+]?[0-9]+\.[0-9]+)\t/\1@\t/g
        s/([-+]?[0-9]+)\t/\1.@\t/g
        s/\./\t./g' file |
  column -t |
    sed -r 's/  \././g
            s/([-+]?[0-9.]+)( +)/\2\1/g
            s/\.@/ /g
            s/@//g
            s/ +X$//'

Typical output

+1255       789   0.11   -55           34
  -15.9900   33   0.22  +321.987  2323566

Answer 2

From a simple point, I'd use awk

column -t filename | awk '{for (i=0; i<NF; i++){l=index($0, $i)+1;printf "%*s ", index($0, $(i+1))-l, $i}printf "%*s\n", index($0, $NF)-l, $NF}'

This makes for slightly wider