[priqueue.members] [queue.mod] [stack.mod] Harmonize push_range wording #6388
Conversation
Discussion of P2767 raised the point that `forward` here is redundant (ranges are take-by-forwarding-reference but use-by-lvalue, since they can't possibly dangle at this point). Be consistent. Also, harmonize the English wording of all three container adaptors. The flat container adaptors are deliberately excluded from this editorial patch, because this harmonization would not be editorial for them; they will be brought into line by P2767 section 5 once LWG has re-reviewed it.
| Equivalent to \tcode{c.append_range(std::forward<R>(rg))} | ||
| if that is a valid expression, | ||
| otherwise \tcode{ranges::copy(rg, back_inserter(c))}. | ||
| Adds the elements of \tcode{rg} to \tcode{c} |
There was a problem hiding this comment.
For other reviewers: c.append_range(rg) is equivalent to c.append_range(std::forward<R>(rg)) since append_range for a standard container doesn't care about the value category of its input_range argument.
Dropping the "Equivalent to" Words of Power is non-editorial, however, which I failed to point out when we discussed this change earlier. "via" doesn't have the same fully-specified meaning. Could we rephrase these as Equivalent to \tcode{c.append_range(rg)} if that is a valid expression, otherwise equivalent to \tcode{ranges::copy(rg, back_inserter(c))}.?
There was a problem hiding this comment.
Dropping the "Equivalent to" Words of Power is non-editorial
I was afraid of that. I do want to point out that the current wording is "Equivalent to ..." in 2 places and "Inserts ... via ..." in the 3rd place, which is again bizarrely inconsistent; and IIUC we can't harmonize all 3 in either direction without one of the changes being non-editorial in this sense. I can bring this back in a real paper if we need me to.
Once I'm not trying to be editorial, I'll go even further and suggest replacing the word "via" with "as if by".
Discussion of P2767 raised the point that
forwardhere is redundant (ranges are take-by-forwarding-reference but use-by-lvalue, since they can't possibly dangle at this point). Be consistent. Also, harmonize the English wording of all three container adaptors.The flat container adaptors'
insert_rangefunctions are deliberately excluded from this editorial patch, because this harmonization would not be editorial for them; they will be brought into line by P2767 section 5 once LWG has re-reviewed it.(Attn @CaseyCarter @var-const)