P3480R6 std::simd is a range #7998
Conversation
mattkretz
force-pushed
the
motions-2025-06-lwg-11
branch
from
c42ed8a to
b4176cc
Compare
| using iterator = \exposid{simd-iterator}<basic_simd>; | ||
| using const_iterator = \exposid{simd-iterator}<const basic_simd>; | ||
|
|
||
| constexpr iterator begin() noexcept { return {*this, 0}; } | ||
| constexpr const_iterator begin() const noexcept { return {*this, 0}; } | ||
| constexpr const_iterator cbegin() const noexcept { return {*this, 0}; } | ||
| constexpr default_sentinel_t end() const noexcept { return {}; } | ||
| constexpr default_sentinel_t cend() const noexcept { return {}; } |
There was a problem hiding this comment.
Seeing that these are entirely defined within the synopsis, we should have some \indexlibrarymember macros above.
Same for basic_simd_mask.
There was a problem hiding this comment.
We have \libmember for inline indexing inside codeblocks. "grep" for examples.
There was a problem hiding this comment.
The members of simd-iterator do not get indexed, right? (because it's an exposition-only type? OTOH its members are public API.)
There was a problem hiding this comment.
Also, looking at some more examples of indexing, there's a lot of index macros missing for [simd]. I guess I should add all those as editorial commit after the motions.
There was a problem hiding this comment.
Looking at ranges.tex, the exposition-only iterators are fully indexed. So I guess I could do that as well.
I'll close this PR in favor of a single PR that includes all 4 simd motions ASAP.
Fixes #7948
Fixes cplusplus/papers#2134