Why isn't sed working?

Question

I have a file that has a single line with a lot of ! characters. I want to remove all of the ! characters.

I tried this:

sed s/!// myfile

and this:

sed 's/!//' myfile

and this:

sed 's/"\!*"//' myfile

But they all just print out all of the "!".

I must be missing something obvious. Any ideas?

Answer 1

Add a g to your regexp, for global replacement. Otherwise, only the first occurrence will be substituted:

sed s/\!//g myfile

Answer 2

Try this:

$ sed s/\!//g myfile

Answer 3

Don't forget poor old tr

tr -d '!' < filename

tr only operates on stdin, so you have to pipe data into it.

Answer 4

add the g at the end in order to replace all occurances.

sed 's/"!*"//g' myfile

Answer 5

You have to specify you want every occurence removed: sed 's/!//g' myfile
Note the g, which mean 'greedy'.
Without the 'g', you only have the first '!' removed