Iterator refactor

[OverkillGuy]

New (unused) iterators for simplifying encoding,

BufferedIterator wraps the read N bytes issue fine (based on PR #1), and ChunkIterator supposed to return a full chunk string on next(). This works, but the base64 is not integrated yet.

Unit Tests are RED on the ChunkIterator, due to comparing byte array to string.

Open question is how to integrate base64 conversion inside the process, when the base64::EncoderWriter approach is a Writer trait, not a Reader trait, so I can't make a oneliner pipeline cleanly, it seems.

[danieleades]

Open question is how to integrate base64 conversion inside the process, when the base64::EncoderWriter approach is a Writer trait, not a Reader trait, so I can't make a oneliner pipeline cleanly, it seems.

you can either wrap a Reader with base64::read::DecoderReader or wrap a Writer with base64::read::EncoderWriter depending on where in your pipeline you want to do the conversion

[OverkillGuy]

you can either wrap a Reader with base64::read::DecoderReader or wrap a Writer with base64::read::EncoderWriter depending on where in your pipeline you want to do the conversion

Yeah, but I'm annoyed because I want a Reader to do encoding but this lib doesn't have it.

Looking around I see another lib: base64-stream has ToBase64Reader, which fits more what I want, despite less downloads... To be continued...

EDIT: I want to clarify, I do like the Iterator approach to generating chunks, which means I want to stay on the side of Reader-conversion.

[OverkillGuy]

Since the actual bugfix was covered elsewhere, but the refactor is still useful alone, I opened #21 to specifically address this refactor.