Review [lex]#8440
Conversation
jensmaurer
added
ballot-comment
| \begin{bnf} | ||
| \nontermdef{class-head}\br | ||
| class-key \opt{attribute-specifier-seq} class-head-name \opt{class-property-specifier-seq} \opt{base-clause}\br | ||
| class-key \opt{attribute-specifier-seq} class-head-name \opt{class-virt-specifier} \opt{base-clause}\br |
There was a problem hiding this comment.
I think we should retain the rename to "class-property-specifier".
To be discussed in core.
There was a problem hiding this comment.
FWIW, if it's not too late, I would go further: my experimentation indicates that vendors generally do treat this thing as a -seq in order to diagnose mistakes like struct A final final {} as "repeated final" instead of "expected {". So I think it would actually be useful to keep P2786's whole grammar change here, not just the change of virt to property but also the addition of -seq and the sentence forbidding repetitions, as (I think) it better reflects existing practice.
Yes, even though it'd be a -seq of only a single possible specifier and repeating that specifier would be ill-formed — it's still existing practice. Or so I think.
|
(This seems somewhat premature; we haven't even got a motion yet.) |
AlisdairM
force-pushed
the
revert_trivial_relocation
branch
from
723c495 to
bb1af1a
Compare
AlisdairM
changed the title
5 participants
Review [lex] for odd whitespace issues