Parsing the output of
w is probably a better approach than
who. Here are some representative data, which shows the login time:
tom pts/1 2015-11-15 06:39 (michener:S.0)
06:40:10 up 1:04, 1 user, load average: 0.00, 0.01, 0.05
USER TTY FROM LOGIN@ IDLE JCPU PCPU WHAT
tom pts/1 michener:S.0 06:39 2.00s 0.03s 0.00s w
Those are more widely available than
finger. Since this is a classroom exercise, parsing the data is left to OP. As a hint, awk can do more than print its fields in a one-liner:
- Typically, one would handle the output of
w by having in the awk script a
BEGIN section (to set a line-number or state).
- Then, a default action for each line (just curly braces with no pattern) would increment the line number.
- Using the line number, handle the first line specially (skip it in this case: OP may need the number of users for a report header, but that is not used in OP's example), and skip the line with
- After that, each line can be printed as OP needs.
awk will quit when there is no more data; it is not necessary to know the number of users to do this.
If OP is told to use
who, that has options to list more information, e.g.,
$ who -l -u
LOGIN tty5 2015-11-15 05:36 3670 id=5
LOGIN tty6 2015-11-15 05:36 3671 id=6
LOGIN tty4 2015-11-15 05:36 3669 id=4
LOGIN tty3 2015-11-15 05:36 3668 id=3
LOGIN tty2 2015-11-15 05:36 3667 id=2
LOGIN tty1 2015-11-15 05:36 3666 id=1
tom pts/1 2015-11-15 06:39 00:06 5780 (michener:S.0)
tom pts/2 2015-11-15 06:52 . 6078 (michener:S.1)
again, showing the terminal name and the login times.