Shell Programming and Scripting

Thanks a lot ajcannon , I ll try it ..

But Is there any way to change the file permanently without any inbetween files.
If I m not wrong I think in the above sed command I ve to redirect to file then move to the original file name like

sed "/^[ ^I]*TAS2[ ^I]*([ ^I]*8'/s/TAS2/TAS/" file1 > file2
mv file2 file1

Like in perl if we give the command like below it changes the input file

grep "^[ ^I]*TAS2[ ^I]*([ ^I]*8' "| perl -p -i -e 's/TAS/TAS2/' $ttt

Is there any same way to do it in sed commmand

Thanks Robotonics again for your kind help
Regards,
user_prady

If you're using gnu sed, there's the "-i" option which edits the input file in place:

Code:

sed -i "..." input_file.txt

Try the man page for more details. If you don't have gnu sed you have to use an intermediate file like you said.

Thanks a LOOOOOOOT To Allllllllllllllllllllllllll..

Hello Experts,

Please help me to cope with the following problem

I ve patterens like

Input
Noptx(5) // remain the same
-*Nop(3);
Nop(9);
--Nop(8); // remain the same d3
**---Nop(7); //remain the same d3
**---Nop(7);
*--Nop(6);
--**Nop(5);
-Nop(4);
Nop(3);

- represents a space
* represents a tab space
Desired OUPUT

Noptx(5) // remain the same
-*Nop(6);
Nop(18);
--Nop(8); // remain the same d3
**---Nop(7); //remain the same d3
**---Nop(14);
*--Nop(12);
--**Nop(10);
-Nop(8);
Nop(6);


I want to change the no to two times but which matches the below patteren

CASE-1: Zero or more spaces/tabs then Nop AND
CASE-2: The line should not end with d3 .

Basically if both patteren matches then double the no inbetween ()
Previously I ve a suggestion but that wont work for the first statement in my input file

awk -F '\\([-@ ]*|\\)' 'NF==3{sub($2,$2*5)};1' file1

Now my code only does the CASE-1 but how to overcome both I ve no idea .

gawk '
/^[ \t]*Nop\(/{
split($0, a, /\(|\)/);
print a[1] "(" a[2]*2 ")" a[3] "// d3" >> "/tmp/my_tmp";
next
}
{ print $0 > "/tmp/my_tmp" } ' file


pls help.......
Regards
User_Prady