There is a difference in the code that I missed before. See the attached file with the screen shots.
In LabVIEW 2018 Example Code the top level Search Unsorted 1D Array.vim has in TSS Case 1 the normal Search 1D Array function, but the start index and element are UNWIRED - thus this case would be broken.
In LabVIEW 2019 Example Code the top level Search Unsorted 1D Array.vim has in TSS case 1 the normal Search 1D Array function, but the start index and element are WIRED - thus this case is not broken. The Assert Structural Type Match function is only configured to look at the Functor input - and thus with it unwired this case is selected.
Perhaps I'm misreading but it looks like you're saying that in the 2018 version, there are some missing wires that preclude the use of case 1, and so case 2 is correctly selected.
In 2019, those necessary wires are present and so case 1 is successfully chosen, which isn't the desired operation at all...
Maybe the cases should be in a different order (i.e. switching 1&2?). I haven't checked this with an installation and am writing only based on the pdf and post, which as I said I may have misunderstood.
That is correct, I included screen shots of the 2018 and 2019 VI at the bottom of the PDF and you can clearly see the wires missing and intact.
Yes, perhaps swapping the order would work, as the malleable VI does stop at the first viable case.
It would be nice if AristosQueue could take a look at the Source Code in the examples folder in 2018/2019 and make a comment.