Unnecessary asterisk in regex that finds CSS comment

Question

I thought to ask this as an update to my previous similar question but it became too long.

I was trying to understand a regex given in w3.org that matches css comments and got this doubt

Why do they use

\/\*[^*]*\*+([^/*][^*]*\*+)*\/
----------------^

instead of just

\/\*[^*]*\*+([^/][^*]*\*+)*\/

?

Both are working similarly. Why do they have an extra star there?

1. Let's look at this part:

   \*+([^/*][^*]*\*+)*
   -A- --B--     -C-
   

   Regex engine will parse the A part and match all the stars until there is NO MORE stars or there is a line break. So once A is done, the next character must be a line break or anything else that's not a star. Then why instead of using [^/] they used [^/*]?

2. Also look at the repeating capturing group.

   ([any one char that's not / or *][zero or more chars that's not *][one or more stars])

   It captures groups of characters ending with atleast one or more stars. So C will take all the stars leaving B with no stars to match in the next round.

   So the B part won't get a chance to meet any stars at all. That is why I think there's no need to put a star there.

But that regex is in w3.org so I guess my understanding may be wrong. Please explain what I'm missing.

Answer 1

This has already been corrected in the CSS3 Syntax module:

\/\*[^*]*\*+([^/][^*]*\*+)*\/   /* ignore comments */

Notice that the extraneous asterisk is gone, making this expression identical to what you have.

So it would seem that it was simply a mistake on their part while writing the grammar for CSS2. I'm digging the mailing list archives to see if there's any discussion there that could be relevant.