Internal Code Review

[hardbyte]

@RacingTadpole has carried out an independent code review of clkhash.

Summary

No major issues identified. The code looks well structured. The two most important issues would be clearing up the python version compatibility, and checking that the type errors are not pointing to any uncovered edge cases. An architectural suggestion would be to remove the dependence on the tqdm progress bar from clk.py, and move it to cli.py.

The 3 page report is available - Clkhash code review.pdf

This issue will track the individual changes to address the identified problems.

Aha! Link: https://csiro.aha.io/features/ANONLINK-41

[hardbyte]

Open Issues:

• [x] out of date cryptography requirements - https://github.com/n1analytics/clkhash/pull/95
• [x] adding test requirements - https://github.com/n1analytics/clkhash/pull/95
• [x] automated dependency vulnerability detection - https://github.com/n1analytics/clkhash/issues/97
• [ ] attribute names dataset - https://github.com/n1analytics/clkhash/issues/98
• [ ] consider applying black to avoid some minor formatting issues - https://github.com/n1analytics/clkhash/issues/99
• [x] creating attributes outside of __init__ - https://github.com/n1analytics/clkhash/issues/100
• [x] unused imports - https://github.com/n1analytics/clkhash/issues/101
• [x] out of date docstrings - https://github.com/n1analytics/clkhash/issues/102
• [x] test CLI properly - https://github.com/n1analytics/clkhash/issues/103
• [ ] progress bars shouldn't be part of the python library only part of the CLI - https://github.com/n1analytics/clkhash/issues/104
• [ ] document our use of mypy - https://github.com/n1analytics/clkhash/issues/105
• [x] resolve python version ambiguity - https://github.com/n1analytics/clkhash/issues/106
• [x] test clk.py properly - https://github.com/n1analytics/clkhash/issues/108