Sed, awk, Perl or other for de-interleaving text file - perl

Question

I would like a relatively compact command to perform line-by-line de-interleaving of a text file, i.e
a1
a2
a3
a4
b1
b2
b3
b4
c1
c2
c3
c4
d1
d2
d3
d4
maps to
a1
b1
c1
d1
a2
b2
c2
d2
a3
b3
c3
d3
a4
b4
c4
d4
The interleaving depth should be adjustable. The lines themselves do not contain any useful structure to assist with the process, and the example above is just a toy example for demonstration purposes. What tool can I use to do this?

Answer 1

sort can do it!
$ sort -k1.2 your_file
-k1.2 sorts by first field starting from 2nd character.
Output:
a1
b1
c1
d1
a2
b2
c2
d2
a3
b3
c3
d3
a4
b4
c4
d4

Answer 2

Basically, what you're looking at doing is reading your data into a 2D array. As you read it in, you can (for example) put the data into the array row by row.
Then when you write the data out, you traverse the array column by column. Adjusting the (de-)interleaving you do just requires a different size of array (or at least that you use a different amount of it, though you could leave the array size itself fixed, if you chose).

Answer 3

According to your new requirements, reordering elements based on their position in the file:
use strict;
use warnings;
my #sorted;
my $depth = 4; # the adjustable interleaving depth
while (<DATA>) {
my $num = ($. % $depth) - 1; # $. is input line number
push #{ $sorted[$num] }, $_;
}
for (#sorted) {
print #$_;
}
__DATA__
a1
a2
a3
a4
b1
b2
b3
b4
c1
c2
c3
c4
d1
d2
d3
d4
Note that the script can be tested on an input file by changing <DATA> to <> and running:
perl script.pl input.txt

Answer 4

Update
Having finally understood your question, thanks to TLP, I suggest this solution. It expects the depth and the input file name on the command line:
$ perl deinter.pl 4 interleaved.txt
and prints the reordered data to STDOUT.
use strict;
use warnings;
my $depth = shift;
my #data = <>;
for my $start (0 .. $depth-1) {
for (my $i = $start; $i < #data; $i += $depth) {
print $data[$i];
}
}
output
a1
b1
c1
d1
a2
b2
c2
d2
a3
b3
c3
d3
a4
b4
c4
d4
Previous solution
Here is a technique that reads the whole file into memory, builds a set of keys for comparison, and sorts the indices of the data so that they can be printed in the new order.
It can be changed for your purposes by modifying the regular expression that extracts the keys fields, and changing the sort block so that the sorted order is correct.
If your file is enormous then it may be necessary to build only the array of keys in memory, and leave the rest of the data on file to be read as it is output.
use strict;
use warnings;
open my $fh, '<', 'interleaved.txt' or die $!;
my #data = <$fh>;
my #keys = map [ /^(.)(.)/ ], #data;
my #sorted = sort {
$keys[$a][1] <=> $keys[$b][1] or
$keys[$a][0] cmp $keys[$b][0]
} 0 .. $#keys;
print $data[$_] for #sorted;

Answer 5

This might work for you (GNU sed and sort):
sed '1{x;s/^/1/;x};G;s/\n/\t/p;x;y/1234/2341/;x;d' file|sort -sk2|sed 's/\t.*//'

Answer 6

I'd like to credit Borodin and TLP for their inputs and answers, which inspired the solution. Its ugly, but I like it
awk 'BEGIN{v=4}{now=(NR-1)%v; STOR[now] = STOR[now] "\n" $0;} END {for (v in STOR) print STOR[v]}'
It also has the flaw of printing spurious newlines (well, the ones appended to the start of the array), but I can deal with that.
EDIT:
Solution for the newlines:
awk 'BEGIN{v=4}{now=(NR-1)%v; STOR[now] = STOR[now] "\n" $0;} END {for (v in STOR) print substr(STOR[v],2)}'