This is the definitive syntax of "array of N T" in C++ Standard https://eel.is/c++draft/dcl.array#1 :

D1 [ constant-expression_opt ] ...

I think the subscript "opt" should not be there for two reasons:

1) The descriptions of the item seem to assume that "constant-expression" cannot be omitted. If the bracket is empty, the descriptions would make no sense.

2) If "constant-expression" is omitted, the item 1 would be reduced to item 2. There is no need to take out this special case as a separate item.

Am I right?

I think you might be right. It seems like a left-over from C++17 where it used one form to describe both cases, with and without constant expression. https://timsong-cpp.github.io/cppwp/n4659/dcl.array#1

Thank you!

Hi

Is it because the array bound may be omitted when there are initialisers? The compiler counts all the initialisers and that is the array bound.
This is clause 7 from the same section, emphasis mine:

7. In addition to declarations in which an incomplete object type is allowed, an array bound may be omitted in some cases in the declaration of a function parameter ([dcl.fct]).
An array bound may also be omitted when an object (but not a non-static data member) of array type is initialized and the declarator is followed by an initializer ([dcl.init], [class.mem], [expr.type.conv], [expr.new]).
In these cases, the array bound is calculated from the number of initial elements (say, N) supplied ([dcl.init.aggr]), and the type of the array is “array of N U”.