Performance improvement on nonblocking sockets

[misalcedo]

The goal of this change is to further reduce the allocations and system calls for a small number of bytes. By fully consuming the read buffer before reading more bytes, we can avoid making system calls to request less than max_length bytes.

In order to handle the streaming with a response block cases, I had to add a block argument to Socket#read. This allowed me to denote whether I wanted to wait for bytes to be avialable or not. That together with a new Socket#read_chunk that tries to read a full chunk_size bytes helped me get tests passing.