sed: How to delete everything in a string after two slashes //?

Question

I have a line with the code

require_once(PATH_ROOT).'/calls/inumber.php'); //this is a comment<br>

I want to delete everything with SED after the //. My first try was

sed -i 's/[//].*//' file;

But that deletes everthing after (PATH.ROOT).'/
I want to remove the comment, not the PATH. Ir is not in the sample above, but how can I exclude SED, not to delete after http:// cause there are two // too.

EDIT: Ok, the quest is, to remove all One-Line-Comments that starts with at least two slashes. It doesnt matter what letters/numbers/signs follow, replace it with nothing. The only exception is http(s):// that should be skipped. Examples and results:
$a=5; //first comment
$a=5;

$b=10; ////// second comment
$b=10;

$c=15; /// /*&/$%§$%&/& third comment
$c=15;

/////////////////////////////
should be empty string

/*test comment*/
/*test comment*/ --> no change cause there are no TWO slashes

Summary: everything after // should be removed (incl the two //) except the http(s)://

Answer 1

You can use greedy nature of quantifiers to always delete only the last occurrence

$ cat ip.txt
require_once(PATH_ROOT).'/calls/inumber.php'); //this is a comment<br>
http://foo/123 //commenting stuff
a//b/c/d 1//23/4/5 //commented

$ sed 's|\(.*\)//.*|\1|' ip.txt
require_once(PATH_ROOT).'/calls/inumber.php'); 
http://foo/123 
a//b/c/d 1//23/4/5 

• sed allows different delimiters to be used, this helps to avoid having to escape //
  • [//] is same as [/], meaning it matches a single /
• \(.*\)//.* use capture group for portion of the line before last set of // so that you can put it back in replacement section using \1

Answer 2

Now that you've changed the question (a lot) here's a sed that shouldn't remove any URLs (file:// or http:// or https:// or anything:// ) - it ignores :// but otherwise deletes everything after two slashes:

sed 's|\([^:]\)//.*$|\1|'

It matches anything not a : (saving that character) followed by // and anything to the end of the line, only putting back the first non-: character.

---

sed 's|//[^/]*$||'

search for // then anything not a slash [^/] zero or more times * then the end of the line $, and replace it with nothing.

If you wanted to match & delete any whitespace before a comment too, you could use the whitespace character class \s

sed 's|\s*//[^/]*$||'

Note that you can't have a slash inside a comment, since that would match URL's as comments, unless you recognized and excluded URL's.

---

Just in case you want to keep the //'s (since you say "delete everything... after the //") you could just put them back:

sed 's|//[^/]*$|//|'

---

Note: To just remove all PHP comments, follow this answer Best way to automatically remove comments from PHP code