Shell Programming and Scripting

Hi

I am a beginner to sed command, here I have a question about using sed to add a few characters into a token of a string.

For example, I have a file,
sqw:qqq:123124:uiqe
dfd:ccc:12390:dfjis
cde:aaa:21311:dfjsid

and, I want the output to be,
sqw:qqq:123124:uiqe
dfd:*WW*ccc:12390:dfjis
cde:aaa:21311:dfjsid
then result is redirected to a new file.

Looking for dfd as pattern, then add *WW* to its second field by using sed.

I have tried to run below statement, which is working on the terminal but not in the script.
cat file | awk -F: '{print $2}' | sed 's:^:\*WW\*:'

I have read a couple of sed threads at this forum, but I am still not able to solve above question by myself. Pls help me out. Thanks.

Code:

sed 's/^dfd:/&*WW*/' myFile.txt

oh, excellent. Thanks a lot

Hi again, I have one more question here. How can I use sed look the Kth field in the line and then replace it. eg. I want to look up 'cde' then replace the line
cde:aaa:21311:dfjsid
to
cde:aaa:00000:dfjsid

in this case the only field you are manipulating is nueric so:

Code:

sed '/^cde/s/[0-9][0-9]*/00000/'

should work.

Having aid that, awk would be a much better tool in this situation:

Code:

awk 'BEGIN { FS=OFS=":" } /^cde/ { $3 = "00000" ; print }'