Sieve: expression negation, nested if / mixed conditionals and actions, support for non-strings in add/add!/update

[monoidic]

This PR has three parts:

• generic negation of arbitrary expressions, along with the removal of the list match negation
• nested if statements, plus mixed actions and actions in the same scope
• the attribute manipulation actions add, add! and update support non-string (bool/int/float) values

There is also an RFC-ish last commit, with some somewhat breaking changes.

• Some operators which were made redundant by negation (!~ and :notcontains) are removed
• :in is used for list inclusion checks instead of ==
• there's a clear split between single-value and multi-value matches, with a couple of new operators due to that: :containsany and :regexin for string matching against multiple regexes or string inclusions)

The last commit is not expected to be accepted at the moment. It'll break a bunch of stuff. Though comments on it would be appreciated.

[codecov-commenter]

Codecov Report

Merging #1923 (2c75a9d) into develop (a7ecf62) will increase coverage by 0.40%.
The diff coverage is 99.48%.

@@             Coverage Diff             @@
##           develop    #1923      +/-   ##
===========================================
+ Coverage    75.15%   75.56%   +0.40%     
===========================================
  Files          400      406       +6     
  Lines        21516    21937     +421     
  Branches      2874     2924      +50     
===========================================
+ Hits         16170    16576     +406     
- Misses        4674     4679       +5     
- Partials       672      682      +10     

Impacted Files	Coverage Δ
intelmq/bots/experts/sieve/expert.py	82.63% <93.75%> (+0.69%)	⬆️
intelmq/tests/bots/experts/sieve/test_expert.py	100.00% <100.00%> (ø)
intelmq/bots/experts/rdap/expert.py	68.86% <0.00%> (-3.41%)	⬇️
intelmq/bots/experts/maxmind_geoip/expert.py	21.48% <0.00%> (-0.65%)	⬇️
intelmq/bots/experts/domain_suffix/expert.py	44.70% <0.00%> (-0.65%)	⬇️
intelmq/bots/experts/tor_nodes/expert.py	48.75% <0.00%> (-0.64%)	⬇️
...telmq/bots/experts/recordedfuture_iprisk/expert.py	43.82% <0.00%> (-0.63%)	⬇️
intelmq/tests/lib/test_utils.py	99.43% <0.00%> (-0.57%)	⬇️
intelmq/bots/experts/asn_lookup/expert.py	37.71% <0.00%> (-0.55%)	⬇️
intelmq/bin/intelmqctl.py	12.88% <0.00%> (-0.06%)	⬇️
... and 13 more

[ghost]

Wow, great work! I'm just missing the docs :D

The last commit is not expected to be accepted at the moment. It'll break a bunch of stuff. Though comments on it would be appreciated.

I have no problems with breaking changes

• as long as they are well documented
• in major release (3.0 is ahead, so that's OK)
• if behaviour doesn't change in unexpected ways, in this case the behaviour of == with lists is such a candidate. However, I'm open for discussions on this item =)

[monoidic]

Some small adjustments, like dropping the negation tests on list/set actions. I will add the breaking changes in a seperate PR, this is enough for one PR as is.

[monoidic]

@wagner-certat Thoughts?

[ghost]

Some small adjustments, like dropping the negation tests on list/set actions. I will add the breaking changes in a seperate PR, this is enough for one PR as is.

Thank you very much - again!

@wagner-certat Thoughts?

We had public holidays yesterday (and sometimes I may be busy :) )

[ghost]

changelog entry added in 7ca54fb