Use random.seed for deterministic doctest outputs

[gyermolenko]

This PR was intended to add doctest for blackboard.py.
As a workaround for random outputs - random.seed was used.

• random.seed(1234) if you should run script by hand like python patterns/other/blackboard.py
• pytest-randomly plugin to run via pytest

I've added related comments to tox.ini but need to explain a bit futher:
For unittest part - plugin is disabled not because of random.* outputs, but because plugin shuffles order of tests and it has found another issue (some tests depend on their order). I haven't investigated it further for the moment, will do it a bit later.

Please let me know if someone has simpler way of testing with random.* or sees problems with current approach.
p.s. there is another script with use of random - abstract_factory. And so I can fix it after.

[codecov-io]

Codecov Report

Merging #315 into master will increase coverage by 0.33%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master     #315      +/-   ##
==========================================
+ Coverage   36.49%   36.83%   +0.33%     
==========================================
  Files          38       38              
  Lines        1307     1276      -31     
==========================================
- Hits          477      470       -7     
+ Misses        830      806      -24

Impacted Files	Coverage Δ
patterns/other/blackboard.py	0% <0%> (ø)	⬆️
patterns/creational/lazy_evaluation.py	90.9% <0%> (-0.76%)	⬇️
patterns/behavioral/observer.py	88.57% <0%> (-0.62%)	⬇️
patterns/behavioral/specification.py	83.05% <0%> (-0.56%)	⬇️
patterns/structural/mvc.py	0% <0%> (ø)	⬆️
patterns/structural/flyweight_with_metaclass.py	0% <0%> (ø)	⬆️
patterns/behavioral/catalog.py	0% <0%> (ø)	⬆️
patterns/behavioral/memento.py	0% <0%> (ø)	⬆️
patterns/behavioral/registry.py	0% <0%> (ø)	⬆️
patterns/behavioral/chain_of_responsibility.py	0% <0%> (ø)	⬆️
... and 1 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1af2089...669818d. Read the comment docs.