UNIX for Dummies Questions & Answers

Hi all,
I have a script that scan files, find old templet and replace it with new one.

#!/bin/ksh
file_name=$1
old_templet=$2
new_templet=$3

# Loop through every file like this
for file in file_name
do
cat $file | sed "s/old_templet/new_templet/g" > $file.new
#do a global searce and replace and put the results
mv $file.new $file
#rename $file.new to original name of file
done


How can I replace the old templet with few rows templet ?
for example:
the old templet =
xxxx
the new templet =
yyyy
zzzz

<script_name> <file_name> xxxx yyyy??zzzz

What has to be instead the ?? in order to split the yyyy and zzzz for tow separate rows ?

10x
Amir

modify the sed part to

sed 's/old_template/new_templ1\^Jnew_templ2\^Jnew_templ3/' tst


here the \^J
is \ then a ctrl-V and hit the return key, u ll c a ^J character

besides instead of doing the cat and redirecting to one file and then moving the new file to old u can do

perl -i -ne "s/old_template/new_templ1\nnew_templ2\nnew_templ3/" file1

file1 will be overwritten with the desired result , use the \n instead of \^J in perl

I change my script, but it does not work ......

#!/bin/ksh
file_name=$1
old_templet=$2
new_templet=$3

# Loop through every file like this
for file in file_name
do
cat $file | sed 's/old_template/new_templ1\^Jnew_templ2\^Jnew_templ3/' tst > $file.new
#do a global searce and replace and put the results
mv $file.new $file
#rename $file.new to original name of file
done

What is worng ??

I am not clear with your question, still I suppose you want to use *sed* to replace some string,

old_template=$2
new_template=$3
filename=$1


cat $filename | sed "s/$new_template/$old_template/g" > file.new
mv file.new $filename

The thing is to interpret $ we need to give the whole expression under double quotes.

Cheers
JK

#!/bin/ksh
file_name=$1
old_templet=$2
new_templet=$3

# Loop through every file like this
for file in $file_name
do
cat $file | sed "s/$new_template/$old_template/g" > $file.new
#do a global searce and replace and put the results
mv $file.new $file
#rename $file.new to original name of file
done


i try to run it from the command line, but it does not work.
any idea ???

Originally posted by amir_yosha

#!/bin/ksh
file_name=$1
old_templet=$2
new_templet=$3

# Loop through every file like this
for file in $file_name
do
cat $file | sed "s/$new_template/$old_template/g" > $file.new
#do a global searce and replace and put the results
mv $file.new $file
#rename $file.new to original name of file
done
-----------------------------------------------------------------------------------

Why are you starting a new thread each time, you can reply or post new questions in your previous mail thread itself by clicking the "post reply" button at the bottom of the screen.

Well coming back to your queries first you are having

new_templet, old_templet and file_name variables, and you are using

new_template, old_template and file variables in your sed, did you intend to use new_templet, old_templet and file_name then put it there.

Also while running the script give the arguments correctly

first argrument should be a filename which you want to edit.
second argument is the old value.
third argument is the new value.

Cheers
JK

#!/bin/ksh
file_name=$1
old_template=$2
new_template=$3

# Loop through every file like this
for file in $file_name
do
cat $file | sed "s/$old_template/$new_template/g" > $file.new
#do a global searce and replace and put the results
mv $file.new $file
#rename $file.new to original name of file
done

<script_mame> <flie_name> <old_template> <new_template>

hi,
i change it but it still does not work.
thanks for your help ....
amir