* op-node: Add channel frame serialization

This creates a struct & serialization & deserialization methods
for it. This is to replace the current serialization code which
is embedded in more complex methods.

The reading is done with a reader API instead of a bytes API because
the frame is variable length & is originally merged together with
other frames without a clean division. The writing API uses a
writer for simplicity (but is not required).

This lays the groundwork for easily switching to fixed int sizes.

* op-node: Use channel frame deserialization

This uses the new channel frame object for deserialization. Some
of the API interaction is a little weird in the channel_bank
IngestData loop, but the code is not able to be easily tested and
upgraded.

* op-node: Use channel frame in serialization

This uses the default implementation (through a struct) rather
than the custom logic. It might make sense to use a slightly
different API for serialization than deserialization given the
inputs, but splitting out the logic into it's own function is
still an improvement.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>