We investigate the problem of efficiently supporting location-aware Publish/Subscribe (Pub/Sub for short), which is essential in many applications such as location-based recommendation and advertising, thanks to the proliferation of geo-equipped devices and the ensuing location-based social media applications. In a location-aware Pub/Sub system (e.g., an e-coupon system), subscribers can register their interest as spatial-keyword subscriptions (e.g., interest in nearby iphone discount); each incoming geo-textual message (e.g., geo-tagged e-coupon) will be delivered to all the relevant subscribers immediately. While there are several prior approaches aiming at providing efficient processing techniques for this problem, their approaches belong to spatial-prioritized indexing method which cannot well exploit the keyword distribution. In addition, their textual filtering techniques are built upon simple variants of traditional inverted indexes, which do not perform well for the textual constraint imposed by the problem. In this paper, we address the above limitations and provide a highly efficient solution based on a novel adaptive index, named AP-Tree. AP-Tree adaptively groups registered subscriptions using keyword and spatial partitions, guided by a cost model. AP-Tree also naturally indexes ordered keyword combinations. Furthermore, we show that our techniques can be extended to process moving spatial-keyword subscriptions, where subscribers can continuously update their locations. We present efficient algorithms to process both stationary and moving subscriptions, which can seamlessly and effectively integrate keyword and spatial partitions. Our extensive experiments demonstrate that AP-Tree and its variant AP+-Tree can achieve up to an order of magnitude improvement on efficiency compared with prior state-of-the-art methods.