I don't think this behavior still occurs. Random tests have been updated some while ago. Is this still an issue?

Agree, this problem is hopelessly out of whack for C#. The rules are applied so inconsistently it is nearly impossible to figure out the correct solution for every random case.

For example: (M.-z vghqG&3kG]ptVQ;xZ/l:\oTjc&P'A*dN0iBU|Qnjw9>1k(]eP.p}?B'3{-@S+J=wBcj/cy)>!ACPB|ZGiU37d.NG;A DN7kl 7ICpVj2g^=

This one has multiple pipes so should fail, but doesn't, I guess because there is a random parenthetical in the middle?? It makes no sense how this is being parsed successfully when other similar ones fail.

Up!

And update: I partially said sh.t: the parentheses are correctly placed (I had forgotten the way the kata worked). So, the problem should only be that your internal function stopes parsing the input string in some cases when this one is invalid and your function doesn't considere it invalid. Either that, or you have to change your random string generator so that your solution deals finally with those strings.

C# version has some problems, for example:

parse '.tGn\e&[UC=e[0vqHNt5m1"#6]>5I<A&m+tAV&lbG&n1 4Oig'7^W]Zbmyv$';?NUkp<W?(nG+gy3%_HmI_4HL0N2QZ.gZ+e?Y3A21' = '' shouldBe '(.tGn\e&[UC=e[0vqHNt5m1"#6]>5I<`A&m+tAV&lbG&n1 4Oig'7^W]Zbmyv$';?NUkp<W?)'

this is invalid regexp because of parentheses