[move.iterators] Various (hopefully) editorial fixes #2582
Conversation
JohelEGP
changed the title
JohelEGP
changed the title
JohelEGP
changed the title
|
Increased the scope of the PR. My first attempt to the fix in the second commit was wrong, so there's some noise. |
source/iterators.tex
Outdated
| @@ -4406,7 +4406,7 @@ | |||
| \begin{itemdescr} | |||
| \pnum | |||
| \constraints | |||
| The expression \tcode{x + n} shall be valid and have type \tcode{Iterator}. | |||
| The expression \tcode{x + n} shall be valid. | |||
There was a problem hiding this comment.
The intent here is "The expression \tcode{x.base() + n} shall be valid and have type \tcode{Iterator}." Admittedly it's weird to constrain n + move_iterator when move_iterator + n isn't constrained; I'll open an LWG issue to constrain the member operators similarly.
There was a problem hiding this comment.
...and while I'm here, thanks for the incredibly detailed review of and corrections to the Ranges wording.
There was a problem hiding this comment.
The intent here is "The expression
\tcode{x.base() + n}shall be valid and have type\tcode{Iterator}." Admittedly it's weird to constrainn + move_iteratorwhenmove_iterator + nisn't constrained; I'll open an LWG issue to constrain the member operators similarly.
Right. I hoped that [move.iter.requirements] did that job.
...and while I'm here, thanks for the incredibly detailed review of and corrections to the Ranges wording.
Sure :)
source/iterators.tex
Outdated
| template<class Iterator, SizedSentinel<Iterator> S> | ||
| constexpr iter_difference_t<Iterator> operator-( | ||
| const move_iterator<Iterator>& x, const move_sentinel<S>& y); | ||
|
|
There was a problem hiding this comment.
These operators are very much intentionally hidden friends for the effect on name lookup. This change is not editorial.
There was a problem hiding this comment.
Thank you. I guess that's the case for all the friends added?
| @@ -4406,7 +4406,7 @@ | |||
| \begin{itemdescr} | |||
| \pnum | |||
| \constraints | |||
| The expression \tcode{x + n} shall be valid and have type \tcode{Iterator}. | |||
| The expression \tcode{x.base() + n} shall be valid and have type \tcode{Iterator}. | |||
|
|
|||
| \pnum | |||
| \returns \tcode{x + n}. | |||
There was a problem hiding this comment.
This change creates an inconsistency between the constraints and the returned value.
move_iterator:::operator+(difference_type) has no constraints, so the expression in question is always valid (even though instantiation of the operator might fail), so presumably we're missing a constraint around [move.iter.nav]p7? Whether or not this expression is SFINAEable is not editorial, and the current specification appears to be normatively wrong, so I think this part needs to go to LWG.
There was a problem hiding this comment.
http://wg21.link/LWG3293 attempts to solve this.
There was a problem hiding this comment.
Thanks for the reminder that I need to fix this issue's P/R.
|
The former change is handled by https://wg21.link/LWG3293 (P3) and the latter is in the draft. |
Hopefully the intention is crystal clear and this can be handled editorially.