Skip to content

[class.dtor] Change redundant wording into note #3966

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[class.dtor] Change redundant wording into note #3966

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 4 additions & 3 deletions source/classes.tex
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2098,9 +2098,10 @@
A prospective destructor can be
declared \keyword{virtual}\iref{class.virtual}
and with a \grammarterm{pure-specifier}\iref{class.abstract}.
If the destructor of a class is virtual and
any objects of that class or any derived class are created in the program,
the destructor shall be defined.
\begin{note}
Since a pure virtual destructor will be invoked by the destructor of
a derived class, it must still be defined in most cases; see~\ref{basic.def.odr}.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We shouldn't use "must" in notes, and certainly not in this way: we're not talking about an existing, external requirement here, but we're explaining the reason for the rule in [basic.def.odr].

@jensmaurer, @opensdh, any ideas for how to phrase an "explanation of a rule"? How about:

"Since a pure virtual destructor will be invoked by the destructor of a derived class, it has a definition in most cases; see [basic.def.odr]."

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

"Since a pure virtual destructor will be invoked by the destructor of a derived class, a definition is generally required; see [basic.def.odr]."

Another alternative would be "usually required".

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@zygoloid suggested:

"Because a pure virtual destructor will be invoked by the destructor of a derived class, [basic.def.odr] requires a definition in most cases."

I'm a bit hesitant about having "requirements in notes", which we aren't allowed to, and maybe this latter form is a bit further from the forms ISO considers "requirements" than "is required" (https://www.iso.org/sites/directives/current/part2/index.xhtml#_idTextAnchor082).

Copy link
Contributor

@opensdh opensdh Nov 9, 2023 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How about

[Note: Even a pure virtual destructor is odr-used by any derived concrete class. — end note]

to avoid talking about any requirement at all? Otherwise, you can use statements like "… is necessary" that more clearly appear to be (mere) statements of fact.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think that's quite correct, since an abstract derived class will call the destructor of a direct non-virtual abstract base class. I'm also not a fan of starting the sentence with "Even", as it implies elaboration on a previous broader statement. How about:

[Note: The destructor of a derived class will odr-use ([basic.def.odr]) the selected destructor of a direct base class -- even if that destructor is pure virtual. -- end note]

Or

[Note: A pure virtual selected destructor will be odr-used ([basic.def.odr]) by any derived classes. -- end note]

Copy link
Contributor

@opensdh opensdh Nov 10, 2023 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The note doesn't have to be exhaustive; having a concrete class is a sufficient but not necessary condition for the odr-use. If you like it better, though, the first version you gave seems correct with minor edits:

[Note: A definition of a destructor of a derived class odr-uses ([basic.def.odr]) the selected destructor of a direct base class, even if that destructor is pure virtual. — end note]

I don't think the second is right, since those derived classes could all have pure virtual destructors with no definition.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The second version is intended to be more concise but less precise... I'll apply your suggested changes to the first version next week :)

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@sdkrystian Friendly ping.

\end{note}

\pnum
\begin{note}
Expand Down