How can we rearrange a table with bash? - bash

Question

I have some ascii file like:
color green
color black
color yellow
os Linux
os android
os windows
os mac
drink coffee
drink water
number 1
number 0
number 100
I want to make a table something like:
color os drink number
-----------------------------------
green Linux coffee 1
black android water 0
yellow windows 100
mac
I wrote this code and it does what I want.
Are there some better solutions?
#!/bin/bash
inpfile=$1
keys=$(cat $inpfile | cut -d" " -f 1 | uniq)
for key in $keys
do
values=$(cat $inpfile | grep $key | cut -d" " -f 2)
result=$(echo "$key ----- $values")
echo $result | datamash -W transpose > /tmp/$key.table
done
paste -d "\t" /tmp/*.table
rm /tmp/*.table

Answer 1

The shell script in your question would fail given various input values (substrings, regexp metachars, spaces, etc.) and/or environment settings and/or even the contents of the directory you run it from. Copy/paste it into http://shellcheck.net and it'll tell you about some of the issues. It'll also be extremely slow.
Here's how to do it robustly, efficiently, and portably using any awk in any shell on every Unix box:
$ cat tst.awk
BEGIN { OFS="\t" }
!($1 in tag2colNr) {
tag2colNr[$1] = ++numCols
rowNr = ++numVals[numCols]
vals[rowNr,numCols] = $1
}
{
colNr = tag2colNr[$1]
rowNr = ++numVals[colNr]
vals[rowNr,colNr] = $2
numRows = (numVals[colNr]>numRows ? numVals[colNr] : numRows)
}
END {
for (rowNr=1; rowNr<=numRows; rowNr++) {
for (colNr=1; colNr<=numCols; colNr++) {
printf "%s%s", vals[rowNr,colNr], (colNr<numCols ? OFS : ORS)
}
}
}
$ awk -f tst.awk file
color os drink number
green Linux coffee 1
black android water 0
yellow windows 100
mac
There are various ways to add the line of dashes under the tags (header) line and/or change the spacing between fields if you feel there's some value in it but since you said "I want to make a table something like..." I assume the above is like enough.

Answer 2

Another possibility is using tput:
#!/bin/bash
set -euo pipefail
function handleFile() {
local data_file="${1}"
local x=0
local y=0
local column_width=20
# we are using tput below to place the cursor,
# so clear the screen for clean output.
clear
while read -r header ; do
# each header is a new column so reset y coordinate to 0
y=0
# place header with underline
printf "$(tput cup "${y}" "${x}")$(tput smul)%s$(tput sgr0)" "${header}"
while read -r row ; do
# move down 1 row by incrementing y coordinate value
((y+=1))
# place the current row value without underline
printf "$(tput cup "${y}" "${x}")%s$(tput sgr0)" "${row}"
# get row values for each header using grep/awk
done < <(grep "${header}" "${data_file}" | awk '{ print $2 }')
# move to next column by incrementing x coordinate by desired column width
((x+=column_width))
# use awk/sort to get unique header values
done < <(awk '{ print $1 }' "${data_file}" | sort -u)
# move cursor down 2 lines and then remove all cursor settings
((y+=2)) && tput cup "${y}" 0 && tput sgr0
}
handleFile "$#"
Sample usage is:
./script.sh data.txt
color drink number os
green coffee 1 Linux
black water 0 android
yellow 100 windows
mac