Experiment: decode a chunk! TODO: how do we decode historical data? It's likely composed of regions representing records, either with some ki

read a chunk, and call decode_chunk

given something that looks like a file path, try to get data and decode it. # first fast forward into some offset. # then report on how many

>>> date, value = format_glucose( '''P "WED","11/10/10","01:46:00 ''' ... + '''"," 076 ","N","00", 00 099C''' ) >>> dat

>>> date, value = format_glucose( '''P "WED","11/10/10","01:46:00 ''' ... + '''"," 076 ","N","00", 00 099C''' ) >>> dat

Give subclasses an opportunity to advertise their own flows.

Prepare an argument parser.

Obtain a file-like handle from something describing access to data.

My library has a convenient hexdumper, YMMV.

some boiler plate to set up logging, reproducible runs, and get our little decoder's IO up and running.

Called by decode on success.

A glucose record is a tuple of the time and glucose. ( datetime.datetime, int ) >>> len( parse_text( '''2011-01-01 01:02 076''' )[ 0 ] )

read a chunk, and normalize it's representation as a bytearray.