agronholm / cbor2 / 69

94%

Pull #13

travis-ci

General encode/decode float handling:

* Try python 3.6 native float16 struct format
* If that doesn't work, try using Numpy instead
* Failing that, use byte-twiddling

Canonical encoding will now use the minimum float size available

Performance tweak:

encoding canonical keys for sorting will use the encode_to_bytes method
to avoid creating a new instance all over again for each key.

71 of 71 new or added lines in 3 files covered. (100.0%)

474 of 492 relevant lines covered (96.34%)

3.73 hits per line