Efficient XPath Evaluation Using a Structural Summary Index
Xiaojie Yuan, Xin Wang, Chenying Wang
Abstract
Xiaojie Yuan, Xin Wang, Chenying Wang
Abstract
XPath has been widely used as a query language for selecting nodes from XML documents. Efficient XPath evaluation is thus a core building block in XML query processors. In the paper, we first introduce a compact structural summary index for XML data and then propose an efficient XPath evaluation algorithm based on this index. Our algorithm takes full advantage of the structural summary to prune a large unnecessary search space. By combining consecutive parent-child steps into a single path fragment, we show how the number of required index lookups and join operations can be significantly reduced. More importantly, our algorithm can process a large class of twig patterns consisting of both parent-child and ancestor-descendant relationships. Finally, we report experimental results on a range of data and queries.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
XPath has been widely used as a query language for selecting nodes from XML documents. Efficient XPath evaluation is thus a core building block in XML query processors. In the paper, we first introduce a compact structural summary index for XML data and then propose an efficient XPath evaluation algorithm based on this index. Our algorithm takes full advantage of the structural summary to prune a large unnecessary search space. By combining consecutive parent-child steps into a single path fragment, we show how the number of required index lookups and join operations can be significantly reduced. More importantly, our algorithm can process a large class of twig patterns consisting of both parent-child and ancestor-descendant relationships. Finally, we report experimental results on a range of data and queries.
Key concepts: XPath, Computer science, XML, Path expression, XML database, Twig, Block (permutation group theory), Index (typography)