IMO random tests in JS are incorrect
Input str:
(A(JKL)K(T)12(@!)89(9#@)%^(M)z(hij)(C(|>)z(z)$%^(*_)AB(XY)nop(FG)234(jkl)y(gh)567(M)(%(w)23(6)WX(+|>)>(o)u(T)xy(2)D(+)(?|(y)r(wxy)FGH(hi)wxy(pq)7(2)abc(|~)STU(|~`)O(BC)(i(MN)^&(y)H(23)FGH(WXY)o(12)+(3)u(EFG)CD(789)!$(M)+|(JKL)r(nop)V(56)(jkl(de)MN(KL)wxy(%^&)b(EFG)456(E)xyz(23)op(K)TU(_)$%(CDE)ghi(st)jkl(st)56(k)gh(r)(f(89)+|>(BCD)ab(EFG)$%(zAB)ghi(WXY)ghi(5)>(9#)PQR(BCD)A(2)ij(g)k(tuv)|>(l)abc(GHI)89#(+)cde(bc))))))))
Input idx: 0
Expected: -1
My answer: 421

Don't we already have enough to/from camelCase kata? Is this really significantly different from these?

https://www.codewars.com/kata/break-camelcase
https://www.codewars.com/kata/split-camelcase

Thank you. Approved :)

Maybe we should add test with double/multiple lowercase sign: This__Is______Example