who | sed 's/ .*//'
-	delete everything after the first space
-	By default sed command executes the command for every line
-	You can specify a range before specifying the command
-	The command will be executed only for those lines mentioned
	in the range
	Ranges can be line numbers such as
	12,17
-	Ranges can be patterns as well such as
	/^$/
	or
	/^$/!
	First means all the blank lines
	Second means all the non-blank lines
-	Mix line numbers with pattern
	sed '1,/^$/d'
	Delete from line 1 upto first occurrence of a blank line

	sed '/^$/,/^end/p'
	Print group of lines strating from a blank line going to
	a line which starts with the string end.
$ who | sed '/lingras/p'
lingras    pts/1        Feb 14 19:16
lingras    pts/1        Feb 14 19:16
stevenso   pts/8        Feb 14 18:52
avstud     pts/6        Feb 14 18:16
phone      pts/38       Feb 13 23:32
newbiggi   pts/59       Feb 14 15:38
peltsch    pts/60       Feb 14 11:34
lingras    pts/46       Feb 14 10:23
lingras    pts/46       Feb 14 10:23
lingras    pts/49       Feb 14 10:23
lingras    pts/49       Feb 14 10:23
$ who | sed '/lingras/'
UX:sed: ERROR: Unrecognized command: /lingras/
$ who | sed -n '/lingras/p'
lingras    pts/1        Feb 14 19:16
lingras    pts/46       Feb 14 10:23
lingras    pts/49       Feb 14 10:23
$ who | sed -n '/lingras/!p'
stevenso   pts/8        Feb 14 18:52
avstud     pts/6        Feb 14 18:16
phone      pts/38       Feb 13 23:32
newbiggi   pts/59       Feb 14 15:38
peltsch    pts/60       Feb 14 11:34
$

-	We can emulate greps using sed
-	Even though we have been using single line programs typed
	directly on the command line, sed can accept a program from
	a file. The option is -f
	sed -f program-file input-file

				Awk
-	awk programs have the following format

	pattern{action}
$ who | awk '$1=="lingras" {print $1}'
lingras
lingras
lingras
$

-	awk program has a lot of special variables
-	NR means line number
$ who | awk '$1=="lingras" {print NR, $1}'
1 lingras
7 lingras
8 lingras
$
-	lingras occurred at line number 1,7 and 8
-	$1 corresponds to first field, $2 corresponds second field
	$0 corresponds to entire line
$ who | awk -F: '{print $1}'
lingras    pts/1        Feb 14 19
stevenso   pts/8        Feb 14 18
avstud     pts/6        Feb 14 18
phone      pts/38       Feb 13 23
newbiggi   pts/59       Feb 14 15
peltsch    pts/60       Feb 14 11
lingras    pts/46       Feb 14 10
lingras    pts/49       Feb 14 10
$ who
lingras    pts/1        Feb 14 19:16
stevenso   pts/8        Feb 14 18:52
avstud     pts/6        Feb 14 18:16
phone      pts/38       Feb 13 23:32
newbiggi   pts/59       Feb 14 15:38
peltsch    pts/60       Feb 14 11:34
lingras    pts/46       Feb 14 10:23
lingras    pts/49       Feb 14 10:23
$ who | awk -F: '{print NF,$1}'
2 lingras    pts/1        Feb 14 19
2 stevenso   pts/8        Feb 14 18
2 avstud     pts/6        Feb 14 18
2 phone      pts/38       Feb 13 23
2 newbiggi   pts/59       Feb 14 15
2 peltsch    pts/60       Feb 14 11
2 lingras    pts/46       Feb 14 10
2 lingras    pts/49       Feb 14 10
$

-	We can specify a different field separator by using -F option
	Previous command uses : as the field separator that's why we have
	NF giving us a value of 2
-	NF corresponds to number of fields on the line
-	$NF will ofcourse correspond to the last field
$ who | awk '$1=="lingras" {print NR, NF, $NF}'
1 5 19:16
7 5 10:23
8 5 10:23
$
$ last -10 | egrep '\(' | awk '{print $NF}' | sed 's/(//' | sed 's/)//'\
> | awk -F: '{print $1,$2}'
00 00
00 00
00 00
00 10
00 00
00 00
00 00
00 00
$
-	The store the output from last command in a file called say
	__temp__

-	run the previous series of commands by greping the file __temp__
	for tty to get the modem statistics
	for pts to get the campus statistics
$ egrep "tty" __temp__
$ egrep "pts" __temp__
-	awk allows two special patterns called BEGIN and END

egrep tty __temp__ |  egrep '\(' | awk '{print $NF}'\
| sed 's/(//' | sed 's/)//'\
| awk -F: '{print $1,$2}' | awk -f count.awk

count.awk
BEGIN{logins=0;time=0;max=0}
{
	logins = logins+1;
	time = time+60*$1+$2;
	if(max < (60*$1+$2))
		max = 60*$1+$2;
}
END{
	hours=(time/logins)/60;
	minutes=(time/logins)%60;
	mhours=max/60;
	mminutes=max%60;
	printf("Campus\t%d:%d\t%d:%d\n",hours,minutes,mhours,mminutes);
}
$ schedule | awk '$1=="95W2"{printf("%s%s\n",$2,$3)}' | tr '[A-Z]' '[a-z]'