Adding the character "#" in a file using sed

kushalkoolwal · 07-11-2008, 02:28 PM

Here is how my current file (module.txt) looks like:

Code:

#Un-comment the following line to load the IDE-Floppy disk drive driver "floppy"
#floppy

# Generated by sensors-detect on Thu Jun  5 17:05:10 2008
# I2C adapter drivers
# modprobe unknown adapter CS5536 ACB0
# Chip drivers
lm90

# Generated by sensors-detect on Fri Jan 25 15:40:21 2008
# Chip drivers
dme1737

and I want it to look like this after testing a particular condition in my shell script (if [ SOMETING = TRUE ] then)
[CODE]
Here is how my current file (module.txt) looks like:

Code:

#Un-comment the following line to load the IDE-Floppy disk drive driver "floppy"
#floppy

# Generated by sensors-detect on Thu Jun  5 17:05:10 2008
# I2C adapter drivers
# modprobe unknown adapter CS5536 ACB0
# Chip drivers
lm90

# Generated by sensors-detect on Fri Jan 25 15:40:21 2008
# Chip drivers
#dme1737

Basically I need to append "#" character at the beginning of the line. (see the blue color above)

Thanks

ramram29 · 07-11-2008, 02:42 PM

You could do something like:

cp file /tmp/file2

if [ SOMETING = TRUE ]
then
sed s/^dme1737/\#dme1737/g /tmp/file2 > file
fi

Test it first by running `sed s/^dme1737/\#dme1737/g file`

kushalkoolwal · 07-11-2008, 02:49 PM

Quote:

Originally Posted by ramram29

You could do something like:

cp file /tmp/file2

if [ SOMETING = TRUE ]
then
sed s/^dme1737/\#dme1737/g /tmp/file2 > file
fi

Test it first by running `sed s/^dme1737/\#dme1737/g file`

Thanks for your reply. I do understand what you are trying to do with the above commands however I have couple of questions with regards to that:

1. Why do we use "^" character before dme1737. Simply telling "dme1737" (without "^") won't do?

2. Also why the character "\" before "#dme1737"?

In short we do want to use the substitute function (/s) won't giving the following command:

Code:

 sed s/dme1737/#dme1737/g /tmp/file2 > file

work?

This is only for my understanding.

jschiwal · 07-11-2008, 02:55 PM

The carat matches the beginning of a line.

Your input pattern has "dme1737" at the beginning of the line, so "^dme1737" matches this pattern.
Using just "dme1739" may work as well but if this pattern exists elsewhere you will have a
false positive match.

Also consider
sed '/^dme1737/s/^/#/' file >newfile.
Since we selected the line with the pattern "^dme1737", we can simply insert a # on that line.

There are often a number of solutions. You need to be accurate enough in your regular expressions
to avoid false matches or misses. So adding ^ before dme1737 would be a good idea. The pattern could be contained in a comment for example. If this where for an automated script that you will use for a while or publish with software, you may want to be even more precise ( matching only dme1737 and not dme1737b for example ) in order to future proof your script.

As far as I know, the backslash isn't needed to escape the # character.

pixellany · 07-11-2008, 02:59 PM

You don't want the "\" before the "#"

"\" is typically used to "escape" a character--which is a funny way of saying change its meaning. "#" is not special in sed, so it does not need to be escaped.

"^" means "at the beginning of the line". You don't need it if there is no other instance of the pattern.