Clever Geek Handbook

Add a block of text after another with sed

I want to add the following block of text // More text here "mysite.com" zone {type master; file "mysite.com.fwd"; allow-update {none; }; };

right after the block of text shown below

zone "." IN {
    type hint;
    file "named.ca";
};

Ultimately the file will look like this

 //There are more lines up here
 zone "." IN {
        type hint;
        file "named.ca";
    };

  zone "mysite.com" {
       type master;
       file "mysite.com.fwd";
         allow-update { none; };
         };
            //There are more lines below here

The tricky thing here is what \n

in the pattern to match doesn't work with sed.

Thanks for the help in advanced

+3


source to share


4 answers


I would do it like this:

sed "/^};/a \ \nzone \"mysite.com\" {\n type master;\n file \"mysite.com.fwd\";\n allow-update { none; };\n};\n" test.conf

Before

zone "." IN {
 type hint;
 file "named.ca";
};

After

zone "." IN {
 type hint;
 file "named.ca";
};

zone "mysite.com" {
 type master;
 file "mysite.com.fwd";
 allow-update { none; };
};

^};

- this is a template after which you add a new block with a

, you can change the template if you need it.



Another approach with sed, which is more flexible, is to use a file in which you have a new block:

sed "/^};/r block.conf" test.conf

Here we simply add text from block.conf

to test.conf

using the read command r

. /^};/

again your template.

This range template should be uniqe /file \"named.ca\";/,/^};/

I guess this should work for you pretty much

sed "/file \"named.ca\";/,/^};/ {
/file*/n
a \ \nzone \"mysite.com\" {\n type master;\n file \"mysite.com.fwd\";\n allow-update { none; };\n};\n
}
" test.conf
Structure

//,//

represents a sample of a range from a line with a matching pattern to another. In our case this is a 2 line pattern and we want to skip the first line, here goes /file*/n

. And then we add our block on the second and second lines, starting with a blank line.

Some useful approaches are also here: How to compile a block of text with a specific pattern?

+1


source


sed

is not a good tool for this. it's easy to do in the shell:

cat file1 >> file2

it means adding content to file1

the end file2

.

UPDATE:



sed -e '$r file1' file2 > file2

will work.

still fails why is sed

more secure than shell.

0


source


If you need a correct and reliable solution, there perl to parse the bind8 config file.

Check Unix :: Conf :: Bind8 :: Conf

0


source


Aspirating the following codes into "my_file" and replacing "^" instead of new lines to get one big line. We can now match the pattern "file "named.ca";

and then replace with new codes after that. Then you convert "^" back to newline characters "\ n".

cat 'my_file'|tr '\n' '^'|sed 's#//There are more lines up here- zone.*\(file "named.ca"\).*#\1  zone "mysite.com" {\n        type master;\n        file "mysite.com.fwd";\n            allow-update { none; }; };\n#g'|tr '^' '\n'

results

//There are more lines up here
 zone "." IN {
        type hint;
        file "named.ca";
    };

  zone "mysite.com" {
       type master;
       file "mysite.com.fwd";
         allow-update { none; };
         };
            //There are more lines below here
0


source






More articles:


All Articles