------------------------------------------------------------------------
PROBLEM: Identify files with zero length and delete them
------------------------------------------------------------------------

$ touch a b c d
$ cat > 1
Hello Kitty
cntrl D

$ ls -l
total 8                      #this line appears only with "ls -l" or "ls -l *"
-rw-r--r--  1 srk  staff  7 Nov 12 21:39 1
-rw-r--r--  1 srk  staff  0 Nov 12 22:13 c
-rw-r--r--  1 srk  staff  0 Nov 12 22:13 b
-rw-r--r--  1 srk  staff  0 Nov 12 22:13 a
-rw-r--r--  1 srk  staff  0 Nov 12 22:13 d

$ ls -l | 'awk $5==0{print "rm " $9}' | sh
$ ls -l | 'awk $5==0{print "rm " $9}' | sh -x   (tracing on)


------------------------------------------------------------------------
PROBLEM: You wish to number a file but skipping headers etc.  (Why
number lines? Useful when you wish to do a complicated merge+sort).
------------------------------------------------------------------------

$ cat > Rabbits
%Rabbit_Name Rabbit_Index
Parvi 10
Laxmi 5
Tubby 8
Padma 6
Rusty 6
cntrl D

$ nl -b p"^[^%]" Rabbits
      	%Rabbit_Name Rabbit_Index
     1	Parvi 10
     2	Laxmi 5
     3	Tubby 8
     4	Padma 6
     5	Rusty 6

#Note that what follows after "p" is a regular expression. Note the
#regular expression. (While at it, the expression "[^]" is not only
#not what you think(!) but also makes no sense. [Worth pondering].

------------------------------------------------------------------------
PROBLEM: Many data files come up with header lines. You want to
proces the data but skip the header lines. The header lines are
are usually identified by a distinct starting character or characters
(e.g. "%" or "#").
------------------------------------------------------------------------

$ cat Rabbits
%Rabbit_Name Rabbit_Index
Parvi 10
Laxmi 5
Tubby 8
Padma 6
Rusty 6

	#Below only data lines are piped to AnalysisProgam 
	#sed, grep, awk can use complex regular expressions 
	#tail requires you to know and specify number of header
	lines

$ sed '/^%/d' Rabbits | AnalysisProgram
	#delete all lines with % as first character

$ sed -n '/^%/!p' Rabbits | AnalysisProgram
	#also possible using two negatives

$ grep -v '%' Rabiits | AnalysisProgram
	#pass through lines which do not have % (anywhere) in the line

$ awk '!/^%/' Rabbits | AnalysisProgram 

$ tail -n +2 Rabbits | AnalysisProgram
	# normal usage is $tail -n 5 Rabbits (last 5 lines) # GNU
	when the number is explicitly "+" then counting is from top
	# Thus $tail -n +2 file means start at the second line