Evaluating an XPath Query on a Streaming XML Document

We present an efficient algorithm for evaluating an XPath query Q (involving only child and descendant axes) on a streaming XML document D. Previously known in-memory algorithms for XPath evaluation use O(|D|) space and O(|Q||D|) time. Several previous algorithms for the streaming version use Θ(d +c) space and Θ(d|D|) time in the worst case; d is the depth of D, n is the number of location steps in Q, and c is the maximum number of candidate elements for output at any one time. In the worst case, the exponential Θ(d) space alone could well exceed the O(|D|) space used by the in-memory algorithms. Our algorithm uses O(d|Q|+cn) space and O((|Q|+d+n)|D|) time in the worst case. So, our algorithm is runtime competitive with the in-memory algorithms, while using much less memory space.