Hi,

I have a address book and am making a change program for it.

I have 6 fields per record
Name:Street:City:State:Zip:Phonenumber

I have asked for the record and using an if statement asked which field they would like to change.

How do I take the answer for the new field and replace the old with the new filed.

I am trying to use:
sed `s/$name/$newname/' < addressbook

I get the following error:
Change: s/Mike/Michael/: not found

any sugestions?

Thanks
Mike

If you are using variables in sed, you must use double quotes instead of single:

#!/bin/ksh

name="Mike"
newname="Michael"
sed "s/$name/$newname/" addressbook

Are you sure sed is what you want to use? The above command replaces the first occurrence of "Mike" per line no matter where it is in the line.

I think it is. This sed statment is inside an if statement pertaining to a specific record in the address book.

This did work and printed it to the screen but did not actually replace the field in the address book.

I would take any sugestions if you have a better way of doing it.

Thanks,
Mike

That sounds similar to what I am working on.

cat<<here

Please enter the record number of the particular record to change:

here

read number

record=`sed -n ${number}p phonebook`

name=`echo $record|cur -d: -f1`
echo $name

echo Do you want to change the name: "$name"
read answer
if[ "$answer" = "y" ]
then
echo "Please enter new name"
read newname
name=$newname
fi

I am now trying to use the sed command to remove the record (the number the user entered).

I also am trying to add the new fields back into the file.

sed -d '${number}'
Is this right?

Thanks,
Eric

Sorry, this line:

name=`echo $record|cur -d: -f1`

should be:

name=`echo $record|cut -d: -f1`

Eric

Instead of sed, use awk which allows you to modify certain fields:

#!/bin/ksh

no=1 # field number
name="Mike"
newname="Michael"
awk ' BEGIN { FS="|"; OFS="|" }
{ if($v1 == v2)
$v1=v3
print }' v1="$no" v2="$name" v3="$newname" addressbook
# end script

In the script above, only field 1 is modified. Any instance of "Mike" in the other fields is ignored.