Replace multiple lines with multiple lines after finding a context - search

Question

I have a file which contains below lines.
xyz_write =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = hostname.com)(PORT = 80))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SID = database)
)
)
I want to find xyz_write and replace the lines below it till empty line is found with below lines.
xyz_write =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = hostname2.com)(PORT = 80))
(CONNECT_DATA =
(FAILOVER = TRUE)
(SERVER = DEDICATED)
(SID = database)
(SERVICE = servicename)
)
)
I tried using sed command by storing the first set of lines in a variable and replacing it will second set of lines. But it did not help. How can we achieve this ?
I tried below method:
I assigned first set of lines to a variable as below.
VARIABLE1="xyz_write =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = hostname.com)(PORT = 80))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SID = database)
)
)"
and
VARIABLE2="xyz_write =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = hostname2.com)(PORT = 80))
(CONNECT_DATA =
(FAILOVER = TRUE)
(SERVER = DEDICATED)
(SID = database)
(SERVICE = servicename)
)
)".
Then i tried running the command
"cat source_file.txt | sed "s/$VARIABLE1/$VARIABLE2/g" > destination_file.txt".
This will not work due to line breaks. There should be some way to achieve this.

Answer 1

The file filename.txt contains the lines as below
$cat filename.txt
<blank line>
abc_write =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = hostname2.com)(PORT = 80))
(CONNECT_DATA =
(FAILOVER = TRUE)
(SERVER = DEDICATED)
(SID = database)
(SERVICE = servicename)
)
)
<blank line>
xyz_write =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = hostname2.com)(PORT = 80))
(CONNECT_DATA =
(FAILOVER = TRUE)
(SERVER = DEDICATED)
(SID = database)
(SERVICE = servicename)
)
)
<blank line>
I wanted just one entry for the context. For "xyz_write" I needed single entry in the file. While adding a new entry to the file, check if it already exists. If already a entry is present for "xyz_write", warn the user and update the entry with new value. Hence I wanted to replace. Somehow I achieved this by removing the existing entry and appending the new entry at the end of file.
Command to Remove the existing entry.
\#This value is a command line input.
SERVICE_NAME="xyz_write"
\#This will remove the existing entry. The command will delete the lines till empty line after the context. -i option will write to the same input file.
sed -i "/$SERVICE_NAME/,/^$/d" filename.txt
Append the new value at the end of file.
NEW_VALUE="bcd_write
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = hostname2.com)(PORT = 80))
(CONNECT_DATA =
(FAILOVER = TRUE)
(SERVER = DEDICATED)
(SID = database)
(SERVICE = servicename)
)
)"
echo $NEW_VALUE >>file_name.txt
Thus my purpose is accomplished. Its not actually replacing the multiple lines. Its achieved by removing existing lines and appending the new lines. Hope this helps somebody.
Thanks,
-Adithya