If you ever need to use the Unix/Linux sed command to insert text before or after a line of text in an existing file, here's how I just ran several sed commands to update my old Function Point Analysis tutorial to have a format that doesn't look like it was created in the 1990s.
This tutorial consists of over 40 files, and I had eight changes I wanted to make each file. So I had two choices: modify each file by hand over the next six hours, or run a series of sed commands and be done in 30 minutes. (I chose the sed commands.)
sed example - Step 1
Because you can easily screw up a lot of files with one sed command, Step 1 is to make a backup of whatever files you're thinking about modifying with sed. Then move that backup to a different directory.
Skip this step at your own risk. You've been warned.
sed example: insert a line after another line
To demonstrate a simple example, here's how I used sed to insert an HTML break tag after a line that contained the text string "header.shtml". After looking at a few files I knew this string (a) was in every file and (b) was unique in those files, so I wrote this sed command:
/header.shtml /a\ <br/>
I saved that sed command in a file named changes.sed.
Next, to make sure I remembered the sed syntax for inserting new text after an existing line, I ran a test sed command from the Linux command line like this:
sed -f changes.sed < node2.shtml | more
This command doesn’t actually change the file node2.shtml, it just reads the file, and then makes the changes as it writes the file contents to standard output. I piped that sed command output into the Linux more command so I could make sure my change worked as expected.
A Linux shell script to change all the files with sed
Satisfied that my change looked correct on this one file, I wrote a simple Linux shell script to run my sed command on every file in the current directory, like this:
# change all the "node*.html" files in the current directory
for i in `ls node*html`
do
echo "working on $i ..."
sed -f changes.sed < $i > ${i}.tmp
mv ${i}.tmp $i
done
I saved this shell script to a file named go.sh, and then ran it like this:
sh go.sh
After it ran, I checked some of my files, and this simple “sed insert after” example worked just fine.
A sed script to insert text before and after a line
That sed example demonstrates how to insert text after a given line in a text file. Next, here's a sed script I used to insert two HTML "div" tags, the first one after the opening body tag, and the second one before the closing body tag.
# insert a div tag *after* the opening body tag /<body /a\ <div id="wrap_body"> # insert the closing div tag *before* the closing body tag /<\/body>/i\ </div>
Again, I tested this sed script like this:
sed -f changes.sed < node2.shtml | more
and once I was sure everything looked right, I again ran my Linux shell script to modify all the HTML files in the current directory.
sed example - inserting text before and after lines
I hope this short example of how to use the sed command to insert text before and after lines in many text files has been helpful.
books by alvin
 |
 |
 |
 |
