hii ..
i have a problem in using sed .. awk is very complicated for me .. my task is a simple stuff .. i have to create a script in which i have to take three input
1>filename
2>first word
3> second word

in this script i have to just replace every instance of first word by second word in file supplied

here is my silly script
#!/bin/bash
ls $1 > /dev/null
if [ $? -eq 1 ]
then
echo file not found
fi
cat $1 | sed 's/$2/$3/g'


when i supply constant string instead of variable .. its doing my work .. like if i give
instead of $2 and $3 like

sed 's/dipen/ass/g'

its replacing dipen by ass .. in the file ..
so again i m stuck by not able to expand my variable ..

i tried everything .. somethings i tried were

sed 's/`echo $2`/`echo $2`/g'
sed 's/${!1}/${!2}/g'

i have tried using escape sequence almost everywhere ..

thnks in advance

try sed "s/$2/$3/g" instead of sed 's/$2/$3/g'

yea, to see what is actually happening, try this:

echo "$PATH"
and
echo "$PATH"

single quotes escape special characters in the shell.

echo '$PATH' is the same as echo "\$PATH"

thanks agrouf and sirclif ... i really need to read some tutorial on expanding variable .. its the second time i have landed into this puddle ..

... but expanding variable is one concept ... escaping special character n so on ...
got anything in ur bookmarks related to scripting ..? if u got any link then plz lemme know the link
thanks a ton guys ..

http://rute.2038bug.com/index.html.gz
http://www.tldp.org/LDP/abs/html/index.html

try these :-)