search word and comment or delete

asim.mcp · 08-09-2009, 11:20 AM

Hi,

I want to know that how can i search for a word in a file and then comment it or delete it.

I know this command works.
sed 's/word/#word/'g inputfile > outputfile

But the lines in my input file are as under:

zone "abc.com" { ---- This line can be comment
type master; ----- This line repeat again and again in the file
file "abc.txt";
allow-query { any; }; -This line repeat again and again in the file
};

I want to know how can i comment a certain zone. Any help please.

Regards

ghostdog74 · 08-10-2009, 08:38 PM

show more input samples. show your final output.

asim.mcp · 08-10-2009, 11:01 PM

input is:

Script will ask me:
Enter a domain: abc.com <- input

output - I don't want it to show me output... I want, When i enter a domain name it should search it into a file called "named.conf" and then comment it as I've done below.

//zone "abc.com" {
// type master;
// file "abc.file";
// allow-query { any; };
//};

original zone record in file "named.conf" is as under:

zone "abc.com" {
type master;
file "abc.file";
allow-query { any; };
};

Regards

pixellany · 08-10-2009, 11:01 PM

You might be looking for an address range.....

Suppose you wanted to start with a line containing "fred", and end with a line containing "lulu". Within this range, you want to add "#" to the beginning of each line:

Code:

sed '/fred/,/lulu/s/^/#/' oldfile > newfile

Best SED tutorial on the planet here:
http://www.grymoire.com/Unix/Sed.html

asim.mcp · 08-11-2009, 12:56 PM

Dear Thanks
It worked for me but there is one problem
It comment from:

"zone "abc.com" {
type master;
file "abc.file";
allow-query { any; };"
}; <-- This line is still not commented..

I've given the beginning zone "abc.com" and ending };

e.g.:
sed '/zone "abc.com"/,/};/s/^/\/\//' name.conf > name.conf-new

Any help please?

pixellany · 08-11-2009, 01:05 PM

When using an address range, it will start at the first occurence of address1 and end on the first occurence of address2. In your example, you can simply change the second address to "^};" meaning "};" at the beginning of the line.

ghostdog74 · 08-11-2009, 08:01 PM

Code:

input="abc.com"
awk -v input="$input" 'BEGIN{ RS=""}
$2 ~ input { 
    gsub(/\n/,"\n#")
    sub(/^/,"#")
}
{print $0 "\n"}
' file

output

Code:

# more file                                                                                                                           
zone "123.com" {
type master;
file "123.file";
allow-query { any; };
};

zone "abc.com" {
type master;
file "abc.file";
allow-query { any; };
};

zone "def.com" {
type master;
file "def.file";
allow-query { any; };
};

# ./shell.sh
zone "123.com" {
type master;
file "123.file";
allow-query { any; };
};

#zone "abc.com" {
#type master;
#file "abc.file";
#allow-query { any; };
#};

zone "def.com" {
type master;
file "def.file";
allow-query { any; };
};

or

Code:

awk -v input=$input '$0 ~ input,/^$/ { $0="#"$0}1' file

asim.mcp · 08-13-2009, 07:00 AM

Dear ,

I'm so much thankful to you. What i wanted, i got.

Again many many thanks

God Bless All.

asim.mcp · 08-15-2009, 02:20 AM

Hi Dear,,

Is there any possibility to give filename as input? because currently i've to enter domain name one by one.