OpenHTF data module's convert_to_base_types() now actually leaves uni… (#769)

* OpenHTF data module's convert_to_base_types() now actually leaves unicode types alone like its doc string suggests.

I ran into this issue when trying to reprocess a test run that had unicode characters in a measurement's unit's "suffix" field.
It's reproducible within this fix using the following command:
./blaze-bin/third_party/car/laser/gbr/v3/testers/amplifier_calibration/amplifier_calibrator_reprocess --dut-serial=QCAG01809300010 --reprocess=QQABUUNBRzAxODA5MzAwMDEwAAEGAWL9ry8k --local=1 calibrate_gbr

RELNOTES:
PiperOrigin-RevId: 194878170

* re-removing history.py

Author: kdsudac

openhtf/util/data.py

@@ -30,12 +30,13 @@
 
 from mutablerecords import records
 from past.builtins import long
+from past.builtins import unicode
 
 from enum import Enum
 import six
 
 # Used by convert_to_base_types().
-PASSTHROUGH_TYPES = {bool, bytes, int, long, type(None), str}
+PASSTHROUGH_TYPES = {bool, bytes, int, long, type(None), unicode}
 
 
 def pprint_diff(first, second, first_name='first', second_name='second'):