Bug Fix: Avoid over-deletion when looking for associated comma

[ketkarameya]

Previously, we were walking up the parent to find the associated trailing/leading comma.
But we did not have any safety checks guarding this comma deletions (like we have for comments).

Lets say in the below scenario we delete isFlgEnabled := exp.BoolValue(staleFlagConst) :

 t.Run("message", func(t *Foobar.T) {
         isFlgEnabled := exp.BoolValue(staleFlagConst)
         fmt.Println("some logging statement")
     })

Our comment/comma cleanup logic would kick in and cleanup stuff like below:

 t.Run("message"
         fmt.Println("some logging statement")
     })

Note we deleted the extra , func(t *Foobar.T) {

I introduced a simple safety check that ensures that there is no node between the comma and the node intended to be deleted.

[lazaroclapp]

Fix makes sense to me, but see comment about potentially missing test case. If there isn't any case that would fit, then let me know and I'll approve as is :)

[lazaroclapp]

In both these examples, the comma is before the deleted node, but the change handles both before and after, right? Any test case case you can think of for the "after" case?

[ketkarameya]

aah I see ur point. I actually had a test case like that, but I had it slightly wrong. I reordered the statements here.
In our earlier logic we would delete }, in addition to isFlgEnabled1 := exp.BoolValue(staleFlagConst)

But now we don't, because we see that there is a node which ends between the ending of deleted node and comma

[lazaroclapp]

Ok, but this change makes the comment above outdated, not?

// Should not delete this comment and the func below

It should now say something about the trailing },

[ketkarameya]

good catch. Addressed. 0c68a1f

[ketkarameya]

addressed comments

[lazaroclapp]

👍