On Jan 18, 2023, at 9:04 PM, Andrew Tomazos <andrewtomazos_at_[hidden]> wrote:

You can see the grammar here:

https://eel.is/c++draft/gram

The relevant productions are attribute-specifier-seq (which [[nodiscard]] matches) and decl-specifier (which friend matches).

For each attribute-specifier-seq there is also corresponding prose that states what entity the attributes pertain to.

So your proposal would need to specify what changes to the grammar you propose

Here are the rules in the grammar that currently (i.e. in the 2020 Standard) require attribute specifiers to come before decl specifiers:


simple-declaration:

decl-specifier-seq init-declarator-listopt ;

attribute-specifier-seq decl-specifier-seq init-declarator-list ;

attribute-specifier-seqopt decl-specifier-seq ref-qualifieropt [ identifier-list ] initializer ;


but perhaps this could be changed to:


simple-declaration:

decl-specifier-seq init-declarator-listopt ;

attribute-specifier-or-decl-specifier-seq init-declarator-list ;

attribute-specifier-or-decl-specifier-seq ref-qualifieropt [ identifier-list ] initializer ;


attribute-specifier-or-decl-specifier-seq:

attribute-specifier-or-decl-specifier-seq attribute-specifier-or-decl-specifier

attribute-specifier-or-decl-specifier


attribute-specifier-or-decl-specifier:

attribute-specifier

decl-specifier


which might not impose a huge burden on compiler writers.

and you would have to test that those changes don't create any new ambiguities (or if they do, you need to add a disambiguation rule which states which branch of the grammar is taken in the case of an ambiguity. Search for "resolved in favor of" for examples.)

How would I perform that test?

Thanks,

Robert





On Thu, Jan 19, 2023 at 3:51 AM Robert Allan Schwartz via Std-Proposals <std-proposals_at_[hidden]<mailto:std-proposals_at_[hidden]>> wrote:

Hello,

Today I am making my first posts to this group. Please forgive me if I make mistakes. Thank you.

Here is the context:

Header <stop_token>:

[[nodiscard]] friend bool operator==(const stop_token& lhs, const stop_token& rhs) noexcept;

I would have thought that this should be:

friend [[nodiscard]] bool operator==(const stop_token& lhs, const stop_token& rhs) noexcept;

i.e. the attribute pertains to the function, not to the friend declaration,
but I was told that an attribute cannot appear in this position,
hence it must be as it is in <stop_token>.

Perhaps the grammar should be amended to allow an attribute in this position?

What do you think?

Thank you,

Robert Schwartz

--
Std-Proposals mailing list
Std-Proposals_at_[hidden]<mailto:Std-Proposals_at_[hidden]>
https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals