BytesMut lacks an implementation of std::io::Write

[NeoLegends]

Copied from tokio-rs/tokio-io#28

I'm currently using the codec module to implement a framed tansport.

Before the refactoring, the encode-method of the old Codec trait received a Vec as parameter to serialize the message into. Vec directly implements std::io::Write which makes it very easy to put any kind of data into it, since most implementations have some way of writing into an instance of Write.

Now the Vec has been replaced by BytesMut, which lacks a direct implementation of that trait. Write-compatibility is only archievable through into_writer, which consumes the BytesMut. Consuming is not possible, though, because the method only receives the BytesMut by &mut.

Is this desired behavior?

[carllerche]

This seems fine to me. My only question is if BytesMut should grow as needed or return an error when it is full.

Thoughts @alexcrichton

[alexcrichton]

Sounds like a good idea to me! I'd grow the buffer internally

[carllerche]

@alexcrichton So BufMut has a writer adapter: BufMut::writer(). That implementation does not grow the buffer. It's built on BufMut::bytes_mut and remaining_mut.

[vorner]

I don't think a writer that doesn't grow the buffer is of much use. Usually, when I need a writer, I delegate the encoding or writing to some other code. And often I don't know in advance how much such code will write.

An example of code that would make sense:

use serde_cbor;
use tokio_io::io::Encoder;

struct E;

impl Encoder for E {
  ...
  fn encode(&mut self, item: Self::Item, dst: &mut BytesMut) -> Result<(), Self::Error> {
    serde_cbor::to_writer(dst)
  }
}

Now I have to wrap dst into a newtype that implements writer because it doesn't know how to be a writer (which isn't much work). If it didn't extend the dst as needed, I'd still need to wrap it, because I don't know the size in advance.

[carllerche]

This is now done: https://docs.rs/bytes/0.5.6/bytes/struct.BytesMut.html#impl-Write

[ColinFinck]

@carllerche BytesMut still lacks an implementation of std::io::Write.
What your URL refers to is an implementation of std::fmt::Write, which is something different.

I'd say this issue is still unresolved.

[taiki-e]

@ColinFinck please see #478

[ColinFinck]

Thanks for the reference, @taiki-e
In the meantime, I'm using BufMut::writer on BytesMut.