XQStream++: Fast tuple extraction algorithm for streaming XML data

Query XML Streaming Data with List

“…For instance, Q1: Figure 1(c), which include "//" axis (A-D relationship), '/' axis (P-C relationship), and "[]"predicate, matches the XML document in Figure 1(a), after event SE(a1), SE(b1), SE(b2), SE(b3), element b1, b2 and b3 are candidates for query node b. If stored in stack as in XQStream++ [10], SE(d1) makes d1 as a candidate for query node d, it is a descendant of b1, b2 and b3, but only top element b3 in stack can be visited, the relationships between d1 and b1/b2 cannot maintained. Similar situation for P-C relationship between d and e query nodes, as shown in Figure 2, d2 and e1 are satisfy P-C relationship, on receiving SE(e2), d1 and e2 are also satisfy P-C relationship, but now d1 is at the bottom of stack, if operations sequence of stack are strict, d1 cannot be visited again until d2 has been pop from stack.…”

“…Some stack-based algorithms extends the array-based algorithms by exploiting the path stack technique to compactly encode a potentially exponential number of query pattern matches in polynomial space, which include TwigM [7], StreamTX [8] and LQ and EQ [9]. XQStream++ [10] addressed the problems of processing tuple extraction queries for streaming XML, but it is not suitable for deep recursive document, predicate and wildcard, this problem is analyzed in Section 2.3.…”