Transposing Rows to Columns in Bash - bash

Question

I've trawling the internet to make sense of how to do what I want to do, but to no avail. I need to transpose the data below....
Caption=C:
Description=Local Fixed Disk
DriveType=3
FreeSpace=41265664000
ProviderName=
Size=146056146944
VolumeName=
Caption=D:
Description=Local Fixed Disk
DriveType=3
FreeSpace=125067259904
ProviderName=
Size=1073738674176
VolumeName=Data
Caption=E:
Description=Removable Disk
DriveType=2
FreeSpace=
ProviderName=
Size=
VolumeName=
To a table layout, like this...
Caption Description DriveType FreeSpace ProviderName Size VolumeName
C: Local Fixed Disk 3 41265664000 146056146944
D: Local Fixed Disk 3 125067259904 1073738674176 Data
E: Removable Disk 2
In bash. I've been exploring a myriad of awk scripts but I don't seem to understand the logic behind them that well :|
Any help would be appreciated. Thank you!

Answer 1

with gawk:
awk -v RS="Caption" -F"[=\n\"]" '
NR==2{
printf RS;
for(i=3;i<=NF;i+=2){
printf ":"$i
};
print ""
};
{
for(i=2;i<=NF;i+=2){
if ($i != "" ) printf ":"$i;
else printf ": "
};
print ""}' file | column -s":" -t

Answer 2

You can use other tools to do this such as Perl or python.
In Perl for example you could use this:
#!/usr/bin/perl -n
use Text::ASCIITable;
next unless /^(\w+)\s*=\s*(.*)$/;
$data{$1} = [] if not $data{$1};
push $data{$1}, $2;
END {
$t = Text::ASCIITable->new();
$t->setCols(keys %data);
for my $i (0..#{$data{(keys %data)[0]}} - 1) {
$t->addRow(map $data{$_}[$i], keys %data)
}
print $t;
}
Your data can be in the data.txt so you can write:
$ ./myscript.pl data.txt
.---------------------------------------------------------------------------------------------------.
| Caption | Size | DriveType | VolumeName | FreeSpace | Description | ProviderName |
+---------+---------------+-----------+------------+--------------+------------------+--------------+
| C: | 146056146944 | 3 | | 41265664000 | Local Fixed Disk | |
| D: | 1073738674176 | 3 | Data | 125067259904 | Local Fixed Disk | |
| E: | | 2 | | | Removable Disk | |
'---------+---------------+-----------+------------+--------------+------------------+--------------'
This script is generic for N number of columns given in any order. However, if the columns are known and displayed always in the same order the code can be simplified to this:
#!/usr/bin/perl -n
use Text::ASCIITable;
BEGIN {
#columns = qw/Caption Description DriveType FreeSpace ProviderName Size VolumeName/;
$t = Text::ASCIITable->new();
$t->setCols(#columns);
$re = join "\n", map "$_=(?<$_>.*)", #columns;
undef $/;
}
$t->addRow(map $+{$_}, #columns) while(/$re/g);
END { print $t; }
Or even in Python if you want:
#!/usr/bin/python
import sys
from terminaltables import AsciiTable
table_data = []
for row in ["Caption" + d for d in sys.stdin.read().split("Caption")[1:]]:
table_data.append([column.split('=')[1] for column in row.split("\n")[:-1]])
columns = [column.split('=')[0] for column in row.split("\n")[:-1]]
table_data.insert(0, columns)
print AsciiTable(table_data).table

Answer 3

$ cat tst.awk
BEGIN { FS=OFS="=" }
{ hdr = hdr sep $1; data = data sep $2; sep=OFS }
$1=="VolumeName" { if (!c++) print hdr; print data; data=sep="" }
$ awk -f tst.awk file | column -s= -t
Caption Description DriveType FreeSpace ProviderName Size VolumeName
C: Local Fixed Disk 3 41265664000 146056146944
D: Local Fixed Disk 3 125067259904 1073738674176 Data
E: Removable Disk 2