139 characters missing from flex

[rfarley3]

Fixes #139 adds edges cases for start and end offsets to avoid missing start characters in FLEX messages.

Also include some changes that:

• remove compiler warnings/type ambiguity when on Raspberry Pi architecture
• refactors mw1 and len calculations into single place
• centralize hardcoded sizes and offsets
• handle long vs short address and single vs group decoding, particular as it pertains to:
  • changing the start and end offset of message content
  • decoding the capcode
  • decode the frag bit
• refactors the message printing to encode printable whitespace modifiers so messages are printing on a single line, avoid empty messages, filter non-printables, convert % into literal (%%).

[EliasOenal]

Thanks a lot for the PR. I just have a small additional request: Given that this is a non-trivial change and contribution, would you mind adding your name (and optionally your email) to the source file? Establishing correct authorship and verifying the licensing is important to some distributions (e.g. Debian) and I would not want to cause issues in their packaging process.

[rfarley3]

Can do, just saw this. I'm glad that this is useful.

[rfarley3]

Authorship added

[rfarley3]

Fixes #159