This function had to be refactored to return the object needed for the next function, rather than "parsed block with this object already in it"

The previous approach required only one param for getBlockTgMessage(parsedBlock), which was nice. However, better design (and more importantly easier to automate mock generation) if this is split out.

Had to be done here because unit tests failed for getBlockTgMessage on account of mock generation treating parsedBlock and parsedBlockWithTokenInfo as identical objects. JS quirk that is hard to paper over in nodejs.

Function is now called with two parameters instead of the modified parsedBlock.

You need to start tracking this in the inputs. Can't just assume sent amount less change is correct in the outputs, as many token send txs have inputs from multiple addresses.

This weird flag variable was necessary to handle complicated etoken txs with multiple address inputs and outputs

Opportunity for optimization: create an array of input addresses, then determine change by what went to any given input address, added to T2972 master list

This was, well, dumb. "oh bc js clones things, we'll just unclone one and not the other one"

just impacts them both, same as earlier approach

Added these unit tests bc needed to test a few outputScript results on txs I was debugging.