Enhancement: Yield more mutations

[localheinz]

This PR

• yields more mutants from UnwrapArray* mutators

💁‍♂️ I'm unsure where to draw the line, although I remember that we briefly discussed this in previous PRs. Maybe worth taking a look? Maybe it would also make sense to start configuring levels? Also take a look at #597.

[maks-rafalko]

array_diff($a, $b); -> $a;, $b; seems like a good idea at first, but it is not hard to notice that only a very lousy test will notice$bs instead of $as because array_diff is supposed to return everything but elements of $b.

Otherwise put, these two mutations are equivalent. If you have a test against mutation $a, it'll surely catch mutation $b. That's why I decided against both when editing #514.

Originally posted by @sanmai in #544 (comment)

---

^ This is why we decided to have only one mutation for array_*diff* functions. I think we shouldn't change anything for them.

[maks-rafalko]

For the latest 2 PRs we changed 1 to N mutations because they were array_*intersect* functions:

• Enhancement: Implement UnwrapArrayUintersectAssoc mutator #628
• Enhancement: Implement UnwrapArrayUintersectUassoc mutator #633

which is totally makes sense

[sanmai]

I don't think these mutations are a very great idea, but not exactly bad one. There could be some merit I haven't yet found. So it'll be best if these extra mutations would not be enabled by default.

[maks-rafalko]

Let's make a decision about this one, to not keep it open without any activity.

I'm 👎 on adding multiple mutations by default (current state of PR) since it produces too many mutants that will likely be killed.

I'm 👍 on adding it by demand (enabling explicitly in settings).

I'm also 👍 to just close it and keep the number of mutations balanced.

what do you think?

[maks-rafalko]

Ok, closing for now. Feel free to reopen / continue discussion