Rewrite IdenticalOperandsRule with SwiftSyntax

[marcelofabri]

[SwiftLintBot]

	5 Warnings
:warning:	If this is a user-facing change, please include a CHANGELOG entry to credit yourself! You can find it at CHANGELOG.md.
:warning:	This PR may need tests.
:warning:	This PR introduced a violation in Realm: /RealmSwift/Tests/ModernObjectAccessorTests.swift:341:27: warning: Identical Operands Violation: Comparing two identical operands is likely a mistake. (identical_operands)
:warning:	This PR introduced a violation in Realm: /RealmSwift/Tests/ModernObjectAccessorTests.swift:431:27: warning: Identical Operands Violation: Comparing two identical operands is likely a mistake. (identical_operands)
:warning:	This PR introduced a violation in Realm: /RealmSwift/Tests/ModernObjectAccessorTests.swift:527:27: warning: Identical Operands Violation: Comparing two identical operands is likely a mistake. (identical_operands)

	13 Messages
:book:	Linting Aerial with this PR took 0.15s vs 0.27s on main (44% faster)
:book:	Linting Alamofire with this PR took 0.09s vs 0.34s on main (73% faster)
:book:	Linting DuckDuckGo with this PR took 0.27s vs 0.73s on main (63% faster)
:book:	Linting Firefox with this PR took 0.56s vs 1.61s on main (65% faster)
:book:	Linting Kickstarter with this PR took 0.98s vs 2.2s on main (55% faster)
:book:	Linting Moya with this PR took 0.14s vs 0.17s on main (17% faster)
:book:	Linting Nimble with this PR took 0.09s vs 0.17s on main (47% faster)
:book:	Linting Quick with this PR took 0.06s vs 0.09s on main (33% faster)
:book:	Linting Realm with this PR took 0.51s vs 1.24s on main (58% faster)
:book:	Linting SourceKitten with this PR took 0.08s vs 0.14s on main (42% faster)
:book:	Linting Sourcery with this PR took 0.21s vs 0.49s on main (57% faster)
:book:	Linting Swift with this PR took 0.8s vs 0.94s on main (14% faster)
:book:	Linting WordPress with this PR took 0.86s vs 2.59s on main (66% faster)

Here's an example of your CHANGELOG entry:

* Rewrite IdenticalOperandsRule with SwiftSyntax.  
  [marcelofabri](https://github.com/marcelofabri)
  [#issue_number](https://github.com/realm/SwiftLint/issues/issue_number)

note: There are two invisible spaces after the entry's text.

Generated by :no_entry_sign: Danger

[jpsim]

I know this is a work in progress, but flagging so it’s not missed, this looks like a regression:

This PR fixed a violation in Realm: /RealmSwift/Tests/AnyRealmValueTests.swift:942:24: warning: Identical Operands Violation: Comparing two identical operands is likely a mistake. (identical_operands)

[jpsim]

Also, it’s really awesome to see what seem to be trustworthy performance benchmark comparisons from OSSCheck.

[marcelofabri]

I know this is a work in progress, but flagging so it’s not missed, this looks like a regression:

This PR fixed a violation in Realm: /RealmSwift/Tests/AnyRealmValueTests.swift:942:24: warning: Identical Operands Violation: Comparing two identical operands is likely a mistake. (identical_operands)

Yeah, I now remember trying to reimplement this a while ago and facing the same problem: https://forums.swift.org/t/detecting-binary-operator-precedence/33350/2

[jpsim]

If porting this rule introduces some regressions but is overall more robust, I'm not opposed to that tradeoff, but it'd be nice to avoid regressions. Let me know if you'd like a hand to look into porting the expression folding algorithm like Tony mentioned in the forums.

[codecov-commenter]

Codecov Report

Merging #3894 (fe299da) into master (1752587) will decrease coverage by 0.05%. The diff coverage is 90.29%.

@@            Coverage Diff             @@
##           master    #3894      +/-   ##
==========================================
- Coverage   92.53%   92.48%   -0.06%     
==========================================
  Files         443      447       +4     
  Lines       22292    22746     +454     
==========================================
+ Hits        20629    21037     +408     
- Misses       1663     1709      +46     

Impacted Files	Coverage Δ
...Framework/Models/SequenceExprFoldingRewriter.swift	75.00% <75.00%> (ø)
...mework/Extensions/SequenceExprSyntax+Folding.swift	85.82% <85.82%> (ø)
...ntFramework/Rules/Lint/IdenticalOperandsRule.swift	97.29% <94.59%> (+3.90%)	:arrow_up:
...ce/SwiftLintFramework/Models/OperatorContext.swift	96.35% <96.35%> (ø)
...Framework/Rules/Style/EmptyEnumArgumentsRule.swift	91.42% <0.00%> (-1.43%)	:arrow_down:
...tFramework/Rules/Lint/ValidIBInspectableRule.swift	95.94% <0.00%> (-1.36%)	:arrow_down:
...tFrameworkTests/AutomaticRuleTests.generated.swift	100.00% <0.00%> (ø)
...ntFramework/Rules/Style/CommaInheritanceRule.swift	90.32% <0.00%> (ø)

:mega: Codecov can now indicate which changes are the most critical in Pull Requests. Learn more

[jpsim]

If we add the swift syntax folding code, maybe that should go in a new SwiftSyntaxFolding module?

[jpsim]

The new violations flagged by OSSCheck should probably not be reported: https://github.com/realm/realm-cocoa/blob/95ebae968d1183decef9ac5990a7f097d02b603e/RealmSwift/Tests/ModernObjectAccessorTests.swift#L341:27

I don't think it's in the spirit of this rule to include function call expressions.

[marcelofabri]

Moving this to draft as we'd want to use https://github.com/apple/swift-syntax/pull/619 instead

[marcelofabri]

Updated it to use the new SwiftOperators package. Still need to address this:

I don't think it's in the spirit of this rule to include function call expressions.

[jpsim]

Updated it to use the new SwiftOperators package.

Nice!