Ktlint 0.44.0, Kotlin 1.6.10

[petertrr]

• Update Kotlin to 1.6.10
  • Use explicit ListSerializer in DiktatSmokeTest
• Update ktlint to 0.44.0
  • Rename all rule ids to work around breaking change in rule execution order
  • Set isUnitTestContext = false in tests, because otherwise it breaks smoke test

This pull request closes #1105

[petertrr]

java.lang.ClassCastException: class kotlin.reflect.jvm.internal.KTypeImpl cannot be cast to class kotlin.jvm.internal.TypeReference (kotlin.reflect.jvm.internal.KTypeImpl and kotlin.jvm.internal.TypeReference are in unnamed module of loader 'app')
	at org.cqfn.diktat.ruleset.smoke.DiktatSmokeTest.overrideRulesConfig(DiktatSmokeTest.kt:300)
	at org.cqfn.diktat.ruleset.smoke.DiktatSmokeTest.overrideRulesConfig$default(DiktatSmokeTest.kt:53)
	at org.cqfn.diktat.ruleset.smoke.DiktatSmokeTest.regression - should not fail if package is not set(DiktatSmokeTest.kt:96)

#1152 (comment)

That's actually funny, that diktat-snapshot workflow passes. I.e., the project can build and run at least via maven plugin, but the error is simply in our tests infrastructure.

[petertrr]

Ktlint 0.44 also has (finally) a proper logging framework integration. We should check it in our plugins, where for both maven and gradle a slf4j-compatible logger is already available.

[codecov]

Codecov Report

Merging #1215 (1da5c61) into master (41c161a) will decrease coverage by 1.48%.
The diff coverage is 100.00%.

@@             Coverage Diff              @@
##             master    #1215      +/-   ##
============================================
- Coverage     83.88%   82.40%   -1.49%     
+ Complexity     2551     2550       -1     
============================================
  Files           105      105              
  Lines          7198     7205       +7     
  Branches       1951     2018      +67     
============================================
- Hits           6038     5937     -101     
+ Misses          352      351       -1     
- Partials        808      917     +109     

Flag	Coverage Δ
unittests	82.40% <100.00%> (-1.49%)	⬇️

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

Impacted Files	Coverage Δ
...tlin/org/cqfn/diktat/ruleset/utils/AstNodeUtils.kt	81.62% <ø> (-1.21%)	⬇️
...kotlin/org/cqfn/diktat/ruleset/rules/DiktatRule.kt	66.66% <100.00%> (+3.03%)	⬆️
...g/cqfn/diktat/ruleset/rules/chapter1/FileNaming.kt	75.00% <100.00%> (ø)
.../diktat/ruleset/rules/chapter1/IdentifierNaming.kt	82.23% <100.00%> (-1.02%)	⬇️
...qfn/diktat/ruleset/rules/chapter1/PackageNaming.kt	89.83% <100.00%> (+0.84%)	⬆️
...at/ruleset/rules/chapter2/comments/CommentsRule.kt	88.63% <100.00%> (ø)
...leset/rules/chapter2/comments/HeaderCommentRule.kt	88.00% <100.00%> (-0.80%)	⬇️
.../ruleset/rules/chapter2/kdoc/CommentsFormatting.kt	72.10% <100.00%> (-1.58%)	⬇️
...diktat/ruleset/rules/chapter2/kdoc/KdocComments.kt	87.77% <100.00%> (-0.50%)	⬇️
...ktat/ruleset/rules/chapter2/kdoc/KdocFormatting.kt	84.11% <100.00%> (-0.94%)	⬇️
... and 113 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 41c161a...1da5c61. Read the comment docs.

[petertrr]

Build passes with kotlin 1.6.10 and ktlint 0.43.2 and explicit ListSerializer.
In ktlint 0.44.0 they must've changed something important.
Also, they now have rule modifiers: RunAfterRule to establish order without the need to constantly change ruleset provider.

[petertrr]

There was a breaking change in pinterest/ktlint#1230: now rule execution order is determined not by order from RuleSetProvider (which we heavily relied on), but rather rules are sorted alphabetically by IDs with ability to tweak order with modifiers. I managed to make our smoke tests pass by copying part of our rules order using modifiers, but it probably needs more testing.

[orchestr7]

There was a breaking change in pinterest/ktlint#1230: now rule execution order is determined not by order from RuleSetProvider (which we heavily relied on), but rather rules are sorted alphabetically by IDs with ability to tweak order with modifiers. I managed to make our smoke tests pass by copying part of our rules order using modifiers, but it probably needs more testing.

very serious change, as we depend on the rule order…

[orchestr7]

May be we should assign proper IDs (aka 1,2,3) for rules to save the order?

[petertrr]

May be we should assign proper IDs (aka 1,2,3) for rules to save the order?

We should prepend all rule IDs with padded indices 001, 002, ..., 137 to preserve the original order

[petertrr]

May be we should assign proper IDs (aka 1,2,3) for rules to save the order?

We should prepend all rule IDs with padded indices 001, 002, ..., 137 to preserve the original order

@Arrgentum could you please help with this change? You can commit directly into this branch. Rules should have IDs like 001-comments, 002-single-constructor etc. Basically, we want them to be sortable lexicographically into the same order that we currently have in DiktatRuleSetProvider. This way we won't need things like RunAfterRule