Mary F. Fernández

UnQL: a query language and algebra for semistructured data based on structural recursion

Abstract. This paper presents structural recursion as the basis of the syntax and semantics of query languages for semistructured data and XML. We describe a simple and powerful query language based on pattern matching and show that it can be expressed using structural recursion, which is introduced as a top-down, recursive function, similar to the way XSL is defined on XML trees. On cyclic data, structural recursion can be defined in two equivalent ways: as a recursive function which evaluates the data top-down and remembers all its calls to avoid infinite loops, or as a bulk evaluation which processes the entire data in parallel using only traditional relational algebra operators. The latter makes it possible for optimization techniques in relational queries to be applied to structural recursion. We show that the composition of two structural recursion queries can be expressed as a single such query, and this is used as the basis of an optimization method for mediator systems. Several other formal properties are established: structural recursion can be expressed in first-order logic extended with transitive closure; its data complexity is PTIME; and over relational data it is a conservative extension of the relational calculus. The underlying data model is based on value equality, formally defined with bisimulation. Structural recursion is shown to be invariant with respect to value equality.

Optimizing regular path expressions using graph schemas

Query languages for data with irregular structure use regular path expressions for navigation. This feature is useful for querying data where parts of the structure is either unknown, unavailable to the user, or changes frequently. Naive execution of regular path expressions is inefficient however, because it ignores any structure in the data. We describe two optimization techniques for queries with regular path expressions. Both rely on graph schemas for specifying partial knowledge about the datas structure. Query pruning uses this structure to restrict navigation to only a fragment of the data; we give an efficient algorithm for rewriting any regular path expression query into a pruned one. Query rewriting using state extents can eliminate or reduce navigation altogether; it is reminiscent of optimizing relational queries using indices. There may be several ways to optimize a query using state extents; we give a polynomial space algorithm that finds all such optimizations. For restricted forms of regular path expressions, the algorithm is provably efficient. We also give an efficient approximation algorithm that works on all regular path expressions.

Declarative specification of Web sites with S

Abstract. Strudel is a system for implementing data-intensive Web sites, which typically integrate information from multiple data sources and have complex structure. Strudels key idea is separating the management of a Web sites data, the specification of its content and structure, and the visual representation of its pages. Strudel provides a declarative query language for specifying a sites content and structure, and a simple template language for specifying a sites HTML representation. This paper contains a comprehensive description of the Strudel system and details the benefits of declarative site specification. We describe our experiences using Strudel in a production application and describe three different, but complementary, systems that extend and improve upon Strudels original ideas.

An Algebra for XML Query

This document proposes an algebra for XML Query. The algebra has been submitted to the W3C XML Query Working Group. A novel feature of the algebra is the use of regular-expression types, similar in power to DTDs or XML Schemas, and closely related to Hasoya, Pierce, and Vouillons work on Xduce. The iteration construct involves novel typing rules not encountered elsewhere (even in Xduce).

XQuery Streaming a la Carte

Existing work on XML query evaluation has either focused on algebraic optimization techniques suitable for XML databases, or on algorithms to efficiently process XML messages represented as a stream of parsing events. In practice, complex applications often must handle both. In this paper, we develop a physical algebra that combines streaming operators with other standard relational and XML operators. Our physical model includes marked XML streams, which permit efficient XPath evaluation, but can only be consumed once. This constraint restricts the use of streaming operators to fragments of a query plan that only access data using depth-first traversal. We develop static analysis techniques to decide which fragment of a plan can be streamed. Our experiments demonstrate the benefits of blending streaming with other evaluation techniques.

A generic programming toolkit for PADS/ML: first-class upgrades for third-party developers

Domain-specific languages facilitate solving problems in a targeted domain by providing features particular to the domain. Declarative domain-specific languages have the additional benefit that users specify what something means rather than how to do something. As a result, the language compiler is free to choose the best implementation strategies and to generate multiple artifacts from a single description. PADS/ML is a declarative data description language designed to facilitate ad hoc data management. From a single description, the compiler generates a myriad of artifacts, including data structures for the in-memory representation of the data and parsers and printers. In this paper, we describe a new generic programming infrastructure for PADS/ML that allows third-party developers to define additional useful artifacts without modifying the compiler.We report on two case studies that use this infrastructure. In the first, we build a version of PADX for PADS/ML, allowing any data source with a PADS/ML description to be queried as if it were XML. In the second, we extend Harmony with the ability to synchronize any data with a PADS/ML description.

A Semi-monad for Semi-structured Data (ICDT Version)

This document proposes an algebra for XML Query. The algebra has been submitted to the W3C XML Query Working Group. A novel feature of the algebra is the use of regular-expression types, similar in power to DTDs or XML Schemas, and closely related to Hasoya and Pierces work on Xduce. The iteration construct is based on the notion of a monad, and involves novel typing rules not encountered elsewhere.