We have grown accustomed to performing elaborate queries on textual data, e.g. via online search engines, file system managers and word processors. In the past decade, retrieval methods that also work on non-textual data have become mainstream (e.g. face recognition software). Sadly, these developments have so far not caught on for data mining within geometrical data, e.g. 3D meshes generated in the course of architectural work. Specifically during data exchange, such a search functionality would be handy, as it often happens that geometry is exported but object identity is lost - think, for example, of generative geometry or exported BIM data. In this paper, we present an example of such a search functionality, based on angular search. Our method is inspired by regular expressions, a string matching technique commonly employed for matching substrings.