On the generalization of Forelli's theorem (Q303603)

Forelli's theorem about the minimum requirements for a function to be holomorphic says the following: Let \(F:B^n\to\mathbb C\) be a complex-valued function defined on the unit ball of \(\mathbb C^{n}\). \(F\) is holomorphic if the following two conditions are satisfied: {\parindent=0.6cm \begin{itemize}\item[(i)] For every positive integer \(k\) there exists an open neighborhood of the origin on which the function is \(\mathcal C^{k}\). \item[(ii)] \(F\) is directionally holomorphic, that is, for every direction \(v\in\mathbb C^{n}\), \(\| v\| =1\), the function \(\varphi_{v}(\zeta)=F(\zeta v)\) is holomorphic on the unit disc. \end{itemize}} Since its first appearance, there have been a lot of generalizations to Forelli's theorem, among others by Chirka, the authors and Kim-Poletsky-Schmalz. This paper aims to solve the final ``remaining'' case and the result reads: In dimension \(2\), consider the vector field \(X=\alpha z\frac{\partial}{\partial z}+\alpha(m w+\beta z^{m})\frac{\partial}{\partial w}\), where \(m\) is a positive integer, \(\alpha\in \mathbb C\setminus\{0\}\), \(\beta\in\mathbb C\). If a complex-valued function \(F:B^2\to\mathbb C\) satisfies the conditions: {\parindent=0.6cm \begin{itemize}\item[(i)] for every positive integer \(k\) there exists an open neighborhood of the origin on which the function is \(\mathcal C^{k}\), \item[(ii)] \(F\) is holomorphic along every complex integral curve of the vector field \(X\), then \(F\) is holomorphic in a neighborhood of the origin. \end{itemize}} In arbitrary dimension, {\parindent=0.6cm if a complex valued function \(f:B^n\to\mathbb C\) satisfies the conditions: \begin{itemize}\item[(i)] for every positive integer \(k\) there exists an open neighborhood of the origin on which the function is \(\mathcal C^{k}\), \item[(ii)] \(F\) is holomorphic along every complex integral curve of an aligned holomorphic vector field \(X\), then \(F\) is holomorphic in a neighborhood of the origin. \end{itemize}} The notion of an aligned holomorphic vector field is somewhat involved. These are known to be contracting holomorphic vector fields satisfying certain assumptions on the parameters in their Poincaré-Dulac normal forms. In particular not every contracting vector field is aligned and such vector fields yield counterexamples to Forelli's theorem under milder assumptions. There is also a real-analytic version of the theorem presented in the paper.