As more and more data have both spatial and textual attributes, spatial keyword query (SKQ) has been proposed to enhance the data search. An SKQ takes a spatial location and a set of keywords as arguments, and returns objects satisfying spatial and textual constraints which are then probably ranked according to certain functions. This paper investigates the current spatial keyword search techniques. It first defines the problem and analyzes the challenges. Then, it discusses the basic spatial keyword query processing techniques. Specifically, it categorizes the proposed queries according to their components, and then classifies the existing query processing techniques into three groups. For each group, the paper reviews the proposed indexing and query processing techniques. The paper also compares these techniques from several perspectives. In addition, it discusses extensions to the basic SKQ. Finally, it addresses research work related to SKQ, as well as some future research directions.