Add CustomDebugStringConvertible conformance to ByteBuffer

[gwynne]

Declare CustomDebugStringConvertible conformance for ByteBuffer, which already implements the conformance requirement.

Motivation:

ByteBuffer implements both description and debugDescription, but conforms only to CustomStringConvertible. As a result, String(describing: ByteBuffer()) and String(reflecting: ByteBuffer()) yield the same output, when it would be expected that the latter would include the additional "readable bytes (max 1k)" debug output returned by invoking debugDescription directly. To my knowledge, there is no drawback to adding this conformance.

Modifications:

Added CustomDebugStringConvertible to the list of protocol conformances on the extension which declares the description and debugDescription properties and the existing CustomStringConvertible conformance.

Result:

Given:

print(ByteBuffer().description)
print(String(describing: ByteBuffer()))
print(ByteBuffer().debugDescription)
print(String(reflecting: ByteBuffer()))

Before (slightly edited for brevity):

ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }
ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }
ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }
readable bytes (max 1k): [ ]
ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }

After (slightly edited for brevity):

ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }
ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }
ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }
readable bytes (max 1k): [ ]
ByteBuffer { readerIndex: 0, writerIndex: 0, readableBytes: 0, ... }
readable bytes (max 1k): [ ]

[dnadoba]

Great catch! I actually always wondered why we are not printing part of the contents.

[Lukasa]

This seems like a good enhancement, thanks!