functions

Defining functions enable you to execute the same code snippet in a shell script repeatedly without
having to write the entire code again. You first define a function and then call (execute) the function in
the main program as shown below.

$ cat funcs

function function1

echo "enter your name : "

read a

echo "Hello, $a... Good to see you"

func2()

echo "enter your name : "

read a

echo "Please visit us again $a !!!"

function1

func2
$ ksh funcs

enter your name :

one

Hello, one... Good to see you

enter your name :

two

Please visit us again two !!!

Exit status functions / Return value functions – are functions that return an output when called
(executed) much similar to the “$?” shell variable
$ cat funcsion

func1()

number=$1

if [ "$number" -lt 5 ]

then

echo "$number is less than 5..." > /dev/null

echo 0

else

echo "$number is not less than 5..." > /dev/null

echo 1

fi

func1 4

func1 5

func1 6
$ ksh funcsion

Functions as comments

$ cat one

echo "this is 1"

echo "this is 2"

func()

echo "this is 3"

echo "this is 4"

echo "this is 5"

$ ksh one

this is 1

this is 2

Awk
Extracting fields using awk, awk is used to extract a field(column) from a given output in tabled
format,awk can also extract using delimiters as shown below. By default awk takes the white space as
the delimiter.

-F : use a specified field seperator(delimiter) to extract field(s)

-v : initialize a variable which can be used inside the BEGIN and END block
-f : read the awk program source from a program-file instead of the first command line

argument

$ cat rows

one two three

four five six

seven eight nine

ten eleven twelve

$ awk '{print $1}' rows

one

four

seven

ten

$ awk '{print $3}' rows

three

six

nine

twelve

using awk with hyphen(-) as the delimiter.

$ cat rows

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve

$ awk -F "-" '{print $2}' rows

two

five

eight

eleven

exercise : use awk to extract the filenames from the output of ls –ltr command, use the pipe (|) operator
to pass the output of ls command to awk - the file name field in the ls –ltr command is the 9 field.

$ cat rows1

one two three

four five six

seven eight nine

ten eleven twelve

$ awk -v var=Iam '{print var" " $1}' rows1

Iam one

Iam four

Iam seven

Iam ten
$ awk -v var='this is' '{print var" " $1}' rows1

this is one

this is four

this is seven

this is ten

$ cat rows1

one two three

four five six

seven eight nine

ten eleven twelve

$ awk -v var="this is" '{print var" " $1}' rows1

this is one

this is four

this is seven

this is ten

$ awk –f comamnd_file input_file

$ cat cmd.awk

{print $1}

$ awk –f cmd.awk rows1

one

four

seven

ten
$ cat cmd.awk

{print var1" "$1}

$ awk -v var1="this is" -f cmd.awk rows1

this is one

this is four

this is seven

this is ten

$ cat cmd1.awk

{print var1" "$1" "var2" "$2}

$ awk -v var1="this is" -v var2="and" -f cmd1.awk -F "-" rows

this is one and two

this is four and five

this is seven and eight

this is ten and eleven

Cut
Cut is another unix binary/utility that lets you extract fields from a given output, cut does not
have any default delimiting case behavior

-d : use a specified delimiter to extract fields

-f : extract only these fields; also print any line that contains no delimiter character,

unless the -s option is specified

-s : do not print lines not containing delimiters

-c : extract only these characters

$ cat rows|cut -d"-" -f1-2

one-two

four-five

seven-eight

ten-eleven

$ cat rows|cut -d"-" -f1,3

one-three

four-six

seven-nine

ten-twelve

$ cat rows|cut -d"-" -f1-3

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve

$ cat rows

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve
$ cut -c 1-4 rows

one-

four

seve

ten-

$ cut -c 1,4 rows

o-

fr

se

t-

$ cat rows

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve

$ cut -c 5-12 rows

two-thre

-five-si

n-eight-

eleven-t
$ cat rows1

one two three

four five six

seven eight nine

ten eleven twelve

$ cut -f1 rows1

one two three

four five six

seven eight nine

ten eleven twelve

$ cut -d " " -f2 rows1

two

five

eight

Eleven

$ cat rows2

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve

one two three

four five six

seven eight nine

ten eleven twelve

$ cut -s -d "-" -f2,3 rows2

two-three

five-six

eight-nine

eleven-twelve

$ cat rows2

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve

one two three

four five six

seven eight nine

ten eleven twelve

$ cut -s -d " " -f1,3 rows2

one three

four six

seven nine

ten twelve
$ cat rows2

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve

one two three

four five six

seven eight nine

ten eleven twelve

$ cut -d " " -f1,3 rows2

one-two-three

four-five-six

seven-eight-nine

ten-eleven-twelve

one three

four six

seven nine

ten twelve

Sed
sed like awk is another language in itself, sed stands for stream editor. sed has so many applications but
the most predominantly used are search and replace

number: a number in the script ensures that the patterns are changed only on that line of

the file

w : ‘w’ flag in the script writes a new file with the resulting output but only for the

matching lines

g : replaces the pattern globally in the entire line

-e : with this option multiple scripts for search and replace can be included

-f : takes a command file with search and replace scripts

: or | : using these two characters will change the syntax of the sed command to include

these characters instead of slash (/) as the delimiter

$ sed 's/search_string/replace_string/' file_name

$ cat streaming

this is line 1

this is line 2

this is line 3

this is line n

$ sed 's/this/that/' streaming

that is line 1

that is line 2

that is line 3

that is line n
$ cat streaming

this is line 1

this is line 2

this is line 3

this is line n

$ sed '3 s/this/that/' streaming

this is line 1

this is line 2

that is line 3

this is line n

$ cat streaming

this is line 1

this is line 2

this is line 3

this is line n
$ sed '1,3 s/this/that/' streaming

that is line 1

that is line 2

that is line 3

this is line n

$ cat streaming

this is line 1

this is line 2

this is line 3

this is line n

$ sed '3,6 s/this/that/' streaming

this is line 1

this is line 2

that is line 3

that is line n
$ cat streaming

this is line 1

this is line 2

this is line 3

this is line n

$ sed 's/this/that/w file1' streaming

that is line 1

that is line 2

that is line 3

that is line n

$ cat file1

that is line 1

that is line 2

that is line 3

that is line n
$ cat streaming

this is line 1

this is line 2

this is line 3

this is line n

$ sed -e 's/this/that/' -e 's/is/was/' streaming

that was line 1

that was line 2

that was line 3

that was line n

$ cat streaming

this is /line 1

this is /line 2

this is /line 3

this is /line n
$ sed 's/\/line/phrase/' streaming

this is phrase 1

this is phrase 2

this is phrase 3

this is phrase n

$ cat streaming

this is /line 1

this is /line 2

this is /line 3

this is /line n

$ sed 's:/line:phrase:' streaming

this is phrase 1

this is phrase 2

this is phrase 3

this is phrase n
$ cat streaming

this is /line 1

this is /line 2

this is /line 3
.

this is /line n

$ sed 's|/line|/phrase|' streaming

this is /phrase 1

this is /phrase 2

this is /phrase 3

this is /phrase n

$ cat streaming

this is /line 1

this is /line 2

this is /line 3

this is /line n
$ cat cmd.sed

s/this/that/

$ sed -f cmd.sed streaming

that is /line 1

that is /line 2

that is /line 3

that is /line n

$ cat streaming

this is /line 1

this is /line 2

this is /line 3

this is /line n

$ cat cmd.sed

s/this/that/

s/is/was/

s/1/11/

s/2/22/
$ sed -f cmd.sed streaming

that was /line 11

that was /line 22

that was /line 3

that was /line n

$ cat streaming

this is line 1 and this is 1st line

this is line 2 and this is 2nd line

this is line 3 and this is 3rd line

this is line n and this is nth line

$ sed 's/this/that/' streaming

that is line 1 and this is 1st line

that is line 2 and this is 2nd line

that is line 3 and this is 3rd line

that is line n and this is nth line

$ cat streaming

this is line 1 and this is 1st line

this is line 2 and this is 2nd line

this is line 3 and this is 3rd line

this is line n and this is nth line

$ sed 's/this/that/g' streaming

that is line 1 and that is 1st line

that is line 2 and that is 2nd line

that is line 3 and that is 3rd line

that is line n and that is nth line

$ cat streaming

this is line 1 and this is 1st line

this is line 2 and this is 2nd line

this is line 3 and this is 3rd line

this is line n and this is nth line

$ cat cmd.sed

s/this/that/g

s/is/was/

s/1/11/g

s/2/22/

$ sed -f cmd.sed streaming

that was line 11 and that is 11st line

that was line 22 and that is 2nd line

that was line 3 and that is 3rd line

that was line n and that is nth line

Grep
grep stands for global regular expression and print, is used to search for patterns in a given output/file.
grep extracts the horizontal lines (rows) unlike awk which prints the vertical columns (fields)

-c : print a count of matching lines for the input file/text

-i : ignore case, prints all the lower/upper matching patterns from a given input file/text

-n : prefix each line of output with the line number

-v : inverts the sense of matching, to select non-matching lines

$ cat regexp

There was a fisherman named Fisher

who fished for some Fish in a fissure.

Till a fish with a grin,

pulled the FISHerman in.

Now they're fishing the FIssure for Fisher.

Is FISHer now done with fISHing ?

because now the FISH need to be sold on the fiSH market

$ grep "fish" regexp

There was a fisherman named Fisher

who fished for some Fish in a fissure.

Till a fish with a grin,

Now they're fishing the FIssure for Fisher.

$ grep -c "fish" regexp

$ grep -n "fish" regexp

1:There was a fisherman named Fisher

2:who fished for some Fish in a fissure.

3:Till a fish with a grin,

5:Now they're fishing the FIssure for Fisher.

$ grep -i "fish" regexp

There was a fisherman named Fisher

who fished for some Fish in a fissure.

Till a fish with a grin,

pulled the FISHerman in.

Now they're fishing the FIssure for Fisher.

Is FISHer now done with fISHing ?

because now the FISH need to be sold on the fiSH market

$ grep -v "fish" regexp

pulled the FISHerman in.

Is FISHer now done with fISHing ?

because now the FISH need to be sold on the fiSH market

$ cat regexp

There was a fisherman named Fisher

who fished for some Fish in a fissure.

Till a fish with a grin,

pulled the FISHerman in.

Now they're fishing the FIssure for Fisher.

Is FISHer now done with fISHing ?

because now the FISH need to be sold on the fiSH market

$ grep -nv "fish" regexp

4:pulled the FISHerman in.

6:Is FISHer now done with fISHing ?

7:because now the FISH need to be sold on the fiSH market

$ grep -cn "fish" regexp

$ grep -ci "fish" regexp

Streaming
Writing to files in a loop can be achieved with “>” and “>>”, but how do you read from a file and
manipulate the line for further processing ?

Use the “while construct” by taking the output from “cat” command

$ cat streaming

this is line 1

this is line 2

this is line 3

this is line n
$ cat streaming|while read line

do

echo "$line"

done

this is line 1

this is line 2

this is line 3

this is line n