UnnecessaryApply: fix false positive when lambda has multiple member references

[t-kameyama]

Fixes #3561

[codecov]

Codecov Report

Merging #3564 (a387781) into main (aaf09dc) will decrease coverage by 0.03%.
The diff coverage is 60.00%.

❗ Current head a387781 differs from pull request most recent head a268f6d. Consider uploading reports for the commit a268f6d to get more accurate results

@@             Coverage Diff              @@
##               main    #3564      +/-   ##
============================================
- Coverage     78.22%   78.19%   -0.04%     
- Complexity     2826     2834       +8     
============================================
  Files           466      466              
  Lines          9136     9137       +1     
  Branches       1724     1726       +2     
============================================
- Hits           7147     7145       -2     
  Misses         1059     1059              
- Partials        930      933       +3     

Impacted Files	Coverage Δ	Complexity Δ
.../arturbosch/detekt/rules/style/UnnecessaryApply.kt	80.00% <60.00%> (-9.66%)	16.00 <8.00> (+8.00)	⬇️

---

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 aaf09dc...a268f6d. Read the comment docs.

[BraisGabin]

Can we wait a bit to merge this, I'm not so sure about this issue. It seems like a true positive to me.

[BraisGabin]

What do you think about what I said in the issue?

[t-kameyama]

class Foo {
    fun bar(f: () -> Unit) {}
    var baz = 1
}

fun test(foo: Foo) {
    // case 1
    foo.apply {
        println(baz)
        println(baz)
    }

    // case 2
    foo.apply {
        bar {
            println(baz)
            println(baz)
        }
    }
}

Since it's not flagged in case 1, I don't think it should be flagged in case 2 as well. I understand that this rule does not flag when lambda has multiple member references.

[BraisGabin]

Maybe we should update it and make it flag the first case too. Promoting with or let

[t-kameyama]

Maybe we should update it and make it flag the first case too. Promoting with or let

I don't think so. The first case can be replaced with let, but it's not an unnecessary apply.

[BraisGabin]

Since it's not flagged in case 1, I don't think it should be flagged in case 2 as well. I understand that this rule does not flag when lambda has multiple member references.

I agree with you here. The problem that I'm talking about is already in the rule so this PR just make it more consistent so we should merge this. I think that we need to rethink a bit this rule but that's another topic 👍.