refactor: raise violated rules as exception

[redeboer]

Violated rules are now raised as an exception. This simplifies the Result class: it only contains a formalism type and a list of StateTransitionGraphs.

[codecov]

Codecov Report

Merging #459 (cc65781) into master (22d6f67) will increase coverage by 0.00%.
The diff coverage is 78.37%.

@@           Coverage Diff           @@
##           master     #459   +/-   ##
=======================================
  Coverage   90.04%   90.04%           
=======================================
  Files          22       22           
  Lines        3285     3296   +11     
  Branches      827      830    +3     
=======================================
+ Hits         2958     2968   +10     
- Misses        169      170    +1     
  Partials      158      158           

Flag	Coverage Δ
unittests	90.04% <78.37%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/expertsystem/reaction/solving.py	91.42% <ø> (ø)
src/expertsystem/reaction/__init__.py	86.53% <78.37%> (-0.39%)	⬇️
src/expertsystem/amplitude/helicity_decay.py	90.34% <0.00%> (+0.77%)	⬆️

[redeboer]

• Extract an ExecutionInfo class from the QNResult class

@spflueger
I'm trying to find out why b99825d breaks the channel tests

[redeboer]

@spflueger
As a side note, it's also possible to split this PR into:

1. Decorating of classes with attr.s. I also tried to decorate Topology, but that breaks the tests somehow, so better to tackle that in a separate PR
2. Refactoring of the Result class

--> #460

[spflueger]

• Extract an ExecutionInfo class from the QNResult class

@spflueger
I'm trying to find out why b99825d breaks the channel tests

We already discussed that in private. But for completeness and documentation I comment here as well. So the issue is that invalid QNResults are created (I can't remember the root cause now). Irrelevant of the root cause, the conclusion is, that the design of this Result class has to be changed as a whole. Otherwise the code keeps staying messy somewhere.

[spflueger]

@spflueger
As a side note, it's also possible to split this PR into:

1. Decorating of classes with attr.s. I also tried to decorate Topology, but that breaks the tests somehow, so better to tackle that in a separate PR
2. Refactoring of the Result class

--> #460

I would like to see some plan on how to improve the Result class design first. For example maybe attr.s is not needed or things change completely anyways, that it does not help changing the current code into using attr.