摘要
通过水平积聚和垂直积聚对原始数据流所对应的数据流图进行裁剪,从而将大部分查询转换成一遍遍历数据流便可执行。本文分析在几种特定情况下,数据流图不能被裁剪成经一遍遍历之后就可以被执行的情况。
By horizontal fusion and vertical fusion to prune data flow graph corresponding to initial datastream, this paper trans-forms most of XQuery queries into a single pass and executes it, analyzes some situations that our technic can not transform the queries into one pass and then executes it.
出处
《计算机与现代化》
2008年第8期60-62,65,共4页
Computer and Modernization
关键词
数据流程图
水平积聚
垂直积聚
流依赖
积聚依赖
data flow graph
horizontal fusion
vertical fusion
flow dependence
aggregate dependence