Extension theorems for analytic objects associated to foliations (Q661955)

Let \(\mathcal{F}\) be an arbitrary germ of an analytic foliation at an isolated singularity \(0\in\mathbb{C}^{2}\). If \(U\) is a neighbourhood of \(0\in\mathbb{C}^{2}\) where each not dicritical separatix of \({\mathcal{F}}|_{U}\) corresponds to an analytic curve passing through \(0\), \(\text{Sep}(\mathcal{F},U)\) is defined as the set of not dicritical separatrices of \(\mathcal{F}\) through \(0\), and the fundamental domain of \(\mathcal{F}\) at \(0\) is defined as the following subset \(\mathcal{D}\): (1) If \(0\) is a nondegenerated reduced singularity, \(\mathcal{D}\) is \(U\setminus \text{Sep}(\mathcal{F},U)\) along with a neighbourhood of an annulus around \(0\) contained in one of the separatrices if \(0\) is resonant (i.e. the quotient of the eigenvalues of \(\mathcal{F}\) is \(\mathbb{Q}^{-}\)); (2) If \(0\) is a saddle-node singularity, \(\mathcal{D}\) is \(U\setminus\text{Sep}(\mathcal{F},U)\) along with a neighbourhood of an annulus around \(0\) contained in the strong separatrix; (3) If \(0\) is not reduced, \(\mathcal{D}\) is \(U\setminus \text{Sep}(\mathcal{F},U)\) along with neighbourhoods of loops, one for each loop contained in a separatrix with a holonomy map given by a resonant diffeomorphism. The main result of the article under review concerns the problem of extension to a neighbourhood of \(0\) of a \(q\)-form \(\xi\) defined on \(U\setminus \text{Sep}(\mathcal{F},U)\) (Theorem 1). Concretely, it is proved that if \(\xi\) extends, after the resolution \(\tilde{\mathcal{F}}\) of \(\mathcal{F}\), in a fundamental domain of \(\tilde{\mathcal{F}}\) at any reduced singularity and along the dicritical components, \(\xi\) extends in \(U\). As a consequence of this interesting result, an application is obtained for a foliation \(\mathcal{F}\) in a neighborhood \(0\in U\) with an isolated singularity at \(0\) defined by an holomorphic \(1\)-form \(\omega\), which is transversely projective in \(U\setminus \text{Sep}(\mathcal{F},U)\) (Theorem 2). Concretely, it is proved that if \((\omega,\eta,\xi)\) is a meromorphic triple of \(1\)-forms in \(U\setminus \text{Sep}(\mathcal{F},U)\) (and then, \(d\omega= \eta \wedge \omega\), \(d\eta=\omega \wedge \xi\) and \(d\xi=\xi\wedge \eta\) in \(U\setminus \text{Sep}(\mathcal{F},U)\)), with \(\eta\) meromorphic in \(U\) and \(\xi\) meromorphic in \(U\setminus \text{Sep}(\mathcal{F},U)\), then \(\xi\) can be extended meromorphically to a neighbourhood of \(0\) if it can be extended to some fundamental domain of \(\mathcal{F}\) at \(0\). The paper is well written with ideas that are interesting for further applications and related problems.