Confusion on how to read raw bytes

I'm creating a low-level, codec used by multiple languages using MsgPack.

RMP has this `encode` API:

```
rmp::encode::write_bin_len(...);
rmp::endoce::write_bin(...);
```

However, it has this `decode` API:

```
rmp::decode::read_bin_len(...);
```

I was expecting a symmetric `rmp::decode::read_bin(...)`.

What's the idiomatic way to read raw bytes using the API?