add write_all_vectored to AsyncWriteExt
#7768
Conversation
leshow
changed the title
write_all_vectored to AsyncWriteExt (#3679)write_all_vectored to AsyncWriteExt
leshow
force-pushed
the
write_all_vectored
branch
from
16eed02 to
34aa2a3
Compare
|
We could implement |
|
I thought that as well but I saw |
martin-g
left a comment
martin-g
left a comment
There was a problem hiding this comment.
If this PR will be finished then it also needs an IT test.
| /// &mut self, | ||
| /// bufs: &mut [IoSlice<'_>], | ||
| /// ) -> io::Result<()> { | ||
| /// let mut pos = 0; |
There was a problem hiding this comment.
The pos variable is never used. Maybe return it as a result to make the example more complete ?!
|
|
||
| pub(crate) fn write_all_vectored<'a, 'b, W>( | ||
| writer: &'a mut W, | ||
| bufs: &'a mut [IoSlice<'b>], |
There was a problem hiding this comment.
The signature in std is a bit different:
| bufs: &'a mut [IoSlice<'b>], | |
| bufs: &'a mut &'a mut [IoSlice<'b>], |
| let me = self.project(); | ||
| while !me.bufs.is_empty() { | ||
| let n = ready!(Pin::new(&mut *me.writer).poll_write_vectored(cx, me.bufs))?; | ||
| if n == 0 { |
There was a problem hiding this comment.
I think this needs extra care for the case when all bufs are empty. In this case 0 is not really an error.
Same improvement should be done in the (ignored) sample code in async_write_ext.rs
3 participants
Motivation
Closes #3679
Was recently looking for
write_all_vectoredand found it missing, went to add it before realizing the MSRV is not there yet.I thought I'd submit this anyway, when the MSRV version moves ahead you can accept, or we could copy in the
IoSlice::advance_slicesfn until that happens?Solution
Tried to follow existing pattern for
write_xmethods