Recent work to incorporate a Unicode-friendly version of neotoma has revealed a minor test issue that results in inconsistencies when testing on R16 vs 17/18. Fix this so we can update our dependencies.
In short, under R16 with the latest tagged neotoma this test fails to generate a unicode error as expected, because the non-latin1 character is split into two integers that fall within the latin1 space.
Recent work to incorporate a Unicode-friendly version of neotoma has revealed a minor test issue that results in inconsistencies when testing on R16 vs 17/18. Fix this so we can update our dependencies.
In short, under R16 with the latest tagged neotoma this test fails to generate a unicode error as expected, because the non-latin1 character is split into two integers that fall within the latin1 space.