在bash中使用awk命令从记录中提取单个字段

Question

I want to use awk to extract a single field from a list of records. For example,

Assignment1:/home/dir/:Admin:08-07-12
Assignment2:/home/dir/:Paul:09-22-13

I want to extract the 1st field from the second line, or the third field from the first line. Any ideas?

Answer 1

awk -F: 'NR == 2 { print $1 }'
awk -F: 'NR == 1 { print $3 }'

For the given line number, print the specified field.

It doesn't work

My apologies, I am using the Bourne shell and this does not work.

As noted in the comment, I don't believe that the Bourne shell has anything directly to do with the problem. (If $IFS is set to something odd, or some other peculiar setup applies, maybe that has an effect. But in a normal Bourne shell, there should be no problem.)

Here's what I get on my machine (an Ubuntu 14.04 derivative, but I'm confident I'd get the same result on Mac OS X 10.10.3, and pretty much any other Unix-like system, too). This is using Bash, but I don't think that's a factor — I'll go out on a limb and say that Korn shell, Zsh, Dash, and Heirloom Shell would all work the same on this; heck, it should work in the C shell family of shells too since there's nothing special about the notations used). It is using GNU awk too, but I don't think that's a factor either.

$ cat data
Assignment1:/home/dir/:Admin:08-07-12
Assignment2:/home/dir/:Paul:09-22-13
$ awk -F: 'NR == 2 { print $1 }' data
Assignment2
$ awk -F: 'NR == 1 { print $3 }' data
Admin
$

The output looks like record 2, field 1 and record 1, field 3 to me. Please find a way to demonstrate what you're doing and the result you get — probably, add something to the question. We can clean it up later when we've worked out what's going wrong with your setup. Please identify your platform reasonably clearly, too.

Answer 2

You can always use 'cut' from bash like so

cut -d ":" -f "1,3" < asdf
Assignment1:Admin
Assignment2:Paul