Shell Programming and Scripting

Hi, I am writing a script that replaces values in a file with some user input. Right now it replaces what i want it to, then cuts out the rest of the text in that tag!

Here is my script:
#!/bin/sh
MAC=$1
ID=$2
Display=$3

cat $MAC-registration.cfg | sed -e 's/displayName=".*"/displayName="'$ID'"/'

I have also tried sed -e 's#\(displayName="\).*\("\)#\1'$ID'\2#g' which does the same thing.

The output that I get is:
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<reginfo>
<reg reg.1.displayName="user input" />
</reginfo>

the output that I expect is:
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<reginfo>
<reg reg.1.displayName="user input" reg.1.address="xxxx" reg.1.label="xxxx" reg.1.auth.userId="xxxx" reg.1.auth.password="xxxx" reg.1.server.1.expires.lineSeize="30" reg.1.lineKeys="2" reg.1.callsPerLineKey="1" reg.1.server.1.address="xxxxxxxxx" />
</reginfo>


Why does it cut out the rest of the data??? Thanks.

How does the line looks like before the replacement and what is the value of $ID?

Regards

Because the the "*" in "/displayName=".*" means that the sed presumably replaces everything after the "displayName." in the line with "'$ID'" in that line.

If the source files could have the text:
reg reg.1.displayName=_ID_
then the sed line could be:

Code:

sed '/_ID_/'$ID'/'

which would then leave the other text in the line untouched.

Ok. I was thinking '.*' would act as a wildcard. This script will be used to edit many different configuration files which will have different values.
But if it replaces everything on the line after displayName="" why does it still print the '/>' at the end? Because of the foreward slash?

Before the file is processed it looks similar to what I said the expected output should be. I was using hello for the value of $ID. This variable will hold a user's name so it is just plaintext with no special characters or whatever.

So basically I need this script to replace whatever text is between the quotes with the variable that the user supplied. Thanks for your help people!

So is there any kind of "wildcard function" i could use to replace varying text between quotes?

Try this:

Code:

sed 's/\(.*displayName=\)".*" \(reg\.1\.address.*\)/\1'"$ID"' \2/'

Regards

Thanks a bunch Franklin! It worked! So what exactly does .* do?