摘要
在现有研究的基础上提出了一种基于式样单(stylesheet)划分的XML数据并行转换方法,并针对该方法所涉及的以下关键问题进行了讨论:1)如何从式样单中获得多个相互独立的转换单元,它们可以并行执行而互不影响;2)如何根据运行环境自动调整转换任务的数量和负载;3)如何将多个并行转换得到的结果进行有效的合并。将提出的算法应用于实际的Open XML-UOF文档格式转换项目,取得了很好的效果。该方法对于在并行环境下有效地提高XML的数据转换性能具有较大的应用价值。
Based on the exploration of current research of XSLT acceleration,a new method to transform XML data in parallel mode based on XSLT Stylesheet partitioning was proposed.The related topics were discussed including:1)how to obtain the independent transform units which can be executed concurrently without affecting each other,2)what is the proper number of parallel tasks according to the underlying platform being used and how to balance their workload,3)how to merge the individual intermediate results correctly.The proposed method was applied into the real project,i.e.the Open XML-UOF document format translation project,achieving fairly good result.This research is significant to the improvement of XML transform performance under parallel processing environment.
出处
《计算机科学》
CSCD
北大核心
2016年第3期231-237,共7页
Computer Science
基金
北京市属高等学校创新团队建设与教师职业发展计划项目(IDHT20130519)资助