hello,
I am using ETL tool informatica and informatica server reads the values of variables stored in the text file in a directory.

Problem: I have an employee table and has 4 departments namely 10,20,30,40. I have to write a shell script in such way that every time i run the shell script it sud update the text file which informatica server uses in doin its ETL task.

---------------------------------------------
TEXT FILE CONTENTS:
[FOLDERNAME.SESSIONNAME]
$$DEPTNO=10
----------------------------------------------
The script file should update the value of $$DEPTNO each time it is run and it sud update in sequence....10 first time...next time i run the script it sud be 20...so on until it is 40 and it sud loop back to 10 and start the cycle ...i mean every time it is run.

I am also supposed to create a backup files in the same directory for every value of $$deptno used.

For example :
bkp1.txt 10
bkp2.txt 20
......
Could any one help me in this regard.

Thanks,
ashton

Adapt and try this script :

The file is backuped before modification : bkp<deptno>.txt

hi,
thanks for the reply.as i am a beginner cud u xplain the following code

old_deptno=$(awk -v FILE="$work_file" '
BEGIN {
FS = "="; OFS = "=";
}
$1 == "$$DEPTNO" {
print $2;
$2 = $2 % 40 + 10
}
{
print $0 >> FILE
}
' deptno.txt )
thanks,
ashton

old_deptno=$(awk ..... )
Set the variable old_deptno to the result (stdout) of the awk command.

awk -v FILE="$work_file" ' ..... ' deptno.txt
Awk set an awk variable FILE and execute the inline script with deptno as input_file.
FILE is the name of the file where to write result of script.
The old value of $$DEPTNO will be written to stdout.

BEGIN { FS="=" ; OFS="=" }
Awk initialization, the code of the BEGIN pattern is excuted before opening the first file.

FS="=" set the internal FIELD SEPARATOR to "=", records will be split in fields that are delimited by FS.
$$DEPTNO=10 gives $1=$$DEPTNO and $2=10

OFS="=" set the internal OUTPUT FIELD SEPARATOR to "=", fields will be written with "=" as separator.
For example: print "aaa","bbb" will be printed as 'aaa=bbb'

$1 == "$$DEPTNO" { ..... }
The line select input records where field $1 is equal to "$$DEPTNO" and specify the code to execute for that record.

print $2;
Write to stdout the old value of $$DEPTNO

$2 = $2 % 40 + 10
Set the new value for $$DEPTNO, 10->20 20->30 30->40 40->10
The record ($0) is now '$$DEPTNO=new_value'

{ print $0 >> FILE }
This code is executed for every input record (no pattern specified).
The current record ($0) is written (append) to the file which name is in variable FILE.
If the record is the '$$DEPTNO' record, the new value had already be set by the previous pattern/action code.


Hope this help, like said PHV ...

thankx a lot it it helps.........

This is another solution which use 'sed' instead of 'awk'


Have a look to thread need help in modifying the contents of a text file
Homework ?

I guess this helps a lot as i am familier with sed command
Thanks a lot