如何正确使用SED更改文件中的行？

Question

I'm trying to use bash's SED command in OS X and failing.

I need to change this line #"phpunit/phpunit:3.7.*", to the same but without the # (so like this "phpunit/phpunit:3.7.*", in my file.

What is the best way to do that in sed?

I tried this: sed -i -e "s/#\\"phpunit/phpunit:3.7.*\\",/\\"phpunit/phpunit:3.7.*\\"," file.txt

but that didn't work :(

Answer 1

I can see an obvious error, that you use the substitution separator character (the slash) inside the regular expression part. You must escape it, like:

sed -e "s/#\"phpunit\/phpunit:3.7.*\",/\"phpunit\/phpunit:3.7.*\",/" file.txt

An improvement could be to group what you want to keep, like:

sed -e "s/#\(\"phpunit\/phpunit:3.7.*\",\)/\1/" file.txt

Both solutions yield:

"phpunit/phpunit:3.7.*",

I don't use OS X , so I omitted the -i switch. In linux is enought -i with an space following it, but in your system I think you must provide a blank string, like: -i'' but not sure, so test this part.

Answer 2

Why not just use something like

sed 's/#//' file.txt

And if you need more specificty then

sed 's/#"php/"php/' file.txt

Answer 3

$ cat input
#"phpunit/phpunit:3.7.*",

Some serious back-slashing later, due to the special characters * , . and / :

$ sed  's/#"phpunit\/phpunit:3\.7\.\*",/"phpunit\/phpunit:3\.7\.\*",/' input
"phpunit/phpunit:3.7.*",