Print a range of lines from standard input or one or more files.

lines is a command line filter that either prints a range of lines, or skips some specified number of lines from the header, some specified number of lines from the footer, or both.

While use of the head and tail POSIX command line programs is already easy to use, how does one go about skipping the initial M lines of a file, or skipping the final N lines of a file? How does one go about skipping both M lines from the top and N lines from the bottom?

Every time I need to do this I spend time doing research on the Internet, judging among a handfull of solutions using awk, perl, python, ruby, or some other scripted solution. Each of the proposals has a slightly different syntax, and some of them don't even work. However all I really want is a filter that does what I need without having to search and study those respective man pages. Is that too much to ask?

I do not claim any of this is an original idea. But I have not found a similar program elsewhere, so I decided to write it myself. I hope it serves others well.

All of the examples assume the following input file, sample.txt.

1: test
2: test
3: test
4: test
5: test
6: test
7: test
8: test
9: test
10: test

lines accepts either --range STRING or -r STRING to specify a range of lines to print. In these examples, I always use the short flag.

In general, lines -r M-N is equivalent to sed -n M,Np, although lines allows omitting M, in which case it will default to the first line, and N, in which case it will default to the last line.

$ lines sample.txt -r 4-7
4: test
5: test
6: test
7: test

Interestingly, when using the short flag name, one may omit the space between the short flag letter and the argument. Therefore -r 4-7 is the same as -r4-7.

Either or both of the ends of the range parameter may be omitted. When the first number is omitted, printing starts at the first line. When the final number is omitted, printing ends at the final line.

As previously described, the intervening space between the flag letter and the argument may be omitted, causing -r -3 to have the same meaning as -r-3, printing the first 3 lines of the file. Printing the first 3 lines of the file is equivalent to lines --top 3.

$ lines sample.txt -r -3
1: test
2: test
3: test

Both -r 7- and -r7- both print lines 7 thru the end of the file. Note this is different than printing the final 7 lines of the file, as one might do with lines --bottom 7.

$ lines sample.txt -r 7-
7: test
8: test
9: test
10: test

Equivalent to sed -n Np.

$ lines sample.txt -r 3
3: test

Equivalent to (( M+=1 )) ; sed -n "$M,\$p", although that modifies M along the way.

$ lines sample.txt --skip-top 2
3: test
4: test
5: test
6: test
7: test
8: test
9: test
10: test

Equivalent to dying your hair gray, because I have not found a way to do this with sed. Maybe I should give awk a swing...

$ lines sample.txt --skip-bottom 2
1: test
2: test
3: test
4: test
5: test
6: test
7: test
8: test

See above, regarding omitting lines from the footer.

$ lines sample.txt --skip-top 3 --skip-bottom 2
4: test
5: test
6: test
7: test
8: test

Equivalent to head -n N. Also note this will have the same effect as calling -r -N.

$ lines sample.txt --top 3
1: test
2: test
3: test

Equivalent to tail -n 3.

$ lines sample.txt --bottom 3
8: test
9: test
10: test

brew install karrick/homebrew-repo/lines

If you don't have the Go programming language installed, then you'll need to install a copy from https://golang.org/dl.

Once you have Go installed:

$ go get github.com/karrick/lines