New model for storing GML document into relational database
通用映射模式下GML关系数据库存储研究

This paper proposed a new model called G2RDB for storing GML data into RDBMS. Through parsing GML documents, G2RDB got geo-feature info, spatial object and nest relation described by its unique parent of every node. According to the predefined mapping rules and the mapping tables of geo-feature models, the proposed method could store GML data into RDB by the generated relational schemas that were mapping from GML document to four kinds of relational tables: feature_info, featm_mappingtab, feat_att and doc_spatialObj. A prototype system of the proposed method had been implemented based on the syntactic and semantic database to show applicability for storing GML data. As a result, the implemented system can store various GML data which have diverse XML structures and different size. Also, comparison shows the advantages that GML data stored in RDB. Furthermore, spatial, non-spatial and mixed content queries can be performed over the stored GML data quickly, obtaining results, in the form of fragments of GML data or GML document.