Extended Regex clarification

Question

This is an example from the book Linux Command Line

If I am trying to match (555) 123-4567 or 555 123-4567 - the following extended regex can be used.

^\(?[0-9][0-9][0-9]\)?  [0-9][0-9][0-9]-[0-9][0-9][0-9][0-9]$

Simplified to

^\(?[0-9]{3}\)?  [0-9]{3}-[0-9]{4}$

In the first example, does the second ? applies to the whole expression enclosed in the ()? Meaning it will match either 555 or (555).

In the second example:

echo "This works." | grep -E '[[:upper:]][[:upper:][:lower:] ]*\.'

Does * star apply to entire expression or only to the second part of expression that contains Upper and lower character classes?

If the previous expression was like this:

grep -E '[[:upper:][:upper:]][:lower:]*\.'

Would the * apply only to the lower character class?

Answer 1

A summary of the comments by Theophrastus and steeldriver (feel free to edit this answer) with some remarks:

In the first example, does the second? applies to the whole expression enclosed in the () ? Meaning it will match either 555 or (555).

The second ? applies to to closing (escaped) ) and the first ? to the opening (escaped) (. In both cases it matches a literal character. Also you have two space characters instead of one following the ? in your regexes, so you would have to remove one space to match your example input.

Your regexes would match (with one space character removed):

(555) 123-4567
(555 123-4567
555) 123-4567
555 123-4567

To match a group you wouldn't escape the parentheses in a extended regular expression.

An example to match 555 and the following space in an optional group:

$ echo -e "555 123-4567\n123-4567" | grep -E '^([0-9]{3} )?[0-9]{3}-[0-9]{4}$'
555 123-4567
123-4567

Does * star apply to entire expression or only to the second part of expression that contains Upper and lower character classes?

Yes, it only applies to the bracket expression [[:upper:][:lower:] ].

grep -E '[[:upper:][:upper:]][:lower:]*\.'

Would the * apply only to the lower character class?

Yes, if it wasn't misspelled. You need two brackets for the lower character class and [[:upper:][:upper:]] wouldn't make much sense. One [[:upper:]] would be sufficient.

So it would be

grep -E '[[:upper:]][[:lower:]]*\.'