UNIX for Dummies Questions & Answers

First time poster, but the forum has saved my bacon more times than... Lots.

Anyway, I have a text file, and wanted to use Awk (or any other sensible program) to print out overlapping sections, or arbitrary length. To describe by example, for file

1
2
3
4
5
etc...

I want the out put to be lines 1-20, then lines 10-30, then 20-40 and so on. I can't find a sensible way to make awk "go back on itself" however.

I could do the whole thing with a horrible loop, running the file through awk separately for each one, and changing a variable, but the files I am operating on ate large and there are lots of them, so ideally I would avoid slower methods.

Any help appreciated!

Hi matfald,
welcome to unix.com!

You may try something like this:

Code:

awk '{ nr[NR] = $0 }
!(NR % s) && NR >= o {
  for (i = NR - o ? NR - o : 1 ; i <= NR; i++)
    print nr[i]
  }' o=20 s=10 infile

Looks very much like that works perfectly. Thanks very much!

I was surprised to see you can tell awk stuff outside the " ' "s but I guess you learn something new every day!

Yes,
see the GNU awk manual.