On nondifferentiable wrinkled functions (Q1119760)

The author refers to \textit{V. Ene}'s paper (reviewed above) and observes that in this study it is not clear that the function in question is nondifferentiable. Therefore he constructs a continuous wrinkled nondifferentiable function \(f:\quad [0,1]\to [0,1]\) such that, if Y is a measurable subset of \((0,1)\), then \(f(x)\in Y\) almost everywhere; each subinterval of \((0,1)\) meets Y in a set of positive measure and \(f(0)=0\), \(f(1)=1\). Moreover, if Z is a measurable subset of \((0,1)\) of positive measure, then there exists a continuous wrinkled nondifferentiable function \(f:\quad [0,1]\to [0,1]\) such that \(f^{-1}(Z)\) is a measurable set of measure greater than \(1-\epsilon\) for given \(\epsilon >0\). This result is obtained by repeated application of the lemma, which states that there is a proper continuous wrinkled function f, mapping the perfect set \[ P=\{x| x=\sum^{\infty}_{i=2}c_ ie_ i,\quad c_ i\in \{0,1\}\quad with\quad e_ i=2^{-i}+2^{-2i}\} \] into X, a measurable set with positive measure, such that no extension of f to \([0,1]\) can have a finite or infinite derivative at any point of P.