replace a string

Shell Programming and Scripting
1

hello

I have a file where an entry needs to be modified. an example of this entry is:

install_location=/xena/sybase/ase12_5

I need to replace xena with another hostname (`hostname` output) for example:

install_location=/dx1/sybase/ase12_5

how to do that?

thanks

2 
 
eval sed 's/xena/`hostname`/' file

should work

3

no need for eval

sed "s/xena/`hostname`/" file
4

true - bad habit of mine overusing eval :slight_smile:

5

thanks. but the string may change:

install_location=/xena/sybase/ase12_5

I want everything between =/ and /sybase to be replaced with the hostname

thx

6

bit clunky but:

 sed "s_=/.*/sybase_=/`hostname`/sybase_" file

should work

7 
sed -e "s/\(install_location=.\)[^/]*\(\/.*\)/\1`hostname`\2/g" file
8

wow.

thanks.

one more thing please:

server_name=XENA
hostname=xena

need to change also these 2 lines:

everything after = to `hostname` in capital lettter for server_name and `hostname` for hostname.

could you please help?
thanks

9

Perhaps something like this

sed -e "s/=\(\/\)[^/]*\(\/\)/=\1`hostname`\2/g" -e "s/=[^/][^/]*/=`hostname`/g" input.txt

I think you are better off with putting install_location, server_name and hostname in the regex part of the sed.

10

this will change all entries to have `hostname` after the equal (=) sign

11

Yes it does. That is why I suggested you put in those strings in the regex part of sed so that you can achieve your desired output.

sed -e "s/install_location=....../g" -e "s/hostname=....../g" -e "s/server_name=....../y" input.txt

Try awk. I think awk may have a toUpper functionality.

12

hello,

So far all is working for replacing all needs strings.

sed -e "s/install_location=....../g" -e "s/hostname=....../g" -e "s/server_name=....../y" input.txt

but frankly I have no idea how to integrate toUpper and awk in the above sed.

any hints please?