The XML Path Language (XPath) is an industry standard notation for addressing parts of an XML document. It is supported by many XML processing libraries and has been used as the foundation for several dedicated XML processing languages. Regular patterns, an alternative way of investigating and destructing XML documents, were first proposed in the XDuce language and feature in a number of its descendants. The processing styles offered by XPath and by regular patterns are each quite convenient for certain sorts of tasks, and the designer of a future XML processing language might well like to provide both. This designer might wonder, however, to what extent these mechanisms can be based on a common foundation. Can one be implemented by translating it into the other? Can aspects of both be combined into a single notation? As a first step toward addressing these questions, we show in this paper that a language closely related to the “downward axis ” fragment of XPath can be accurately translated into ambiguous XDuce-style regular patterns with a “collect all matches ” interpretation. 1

Abstract. In the design of type systems for XML programming languages based on regular expression types and patterns the focus has been over result analysis, with the main aim of statically checking that a transformation always yields data of an expected output type. While being crucial for correct program composition, result analysis is not sufficient to guarantee that patterns used in the transformation are correct. In this paper we motivate the need of static detection of incorrect patterns, and provide a formal characterization based on pattern matching operational semantics, together with locally exact type analysis techniques to statically detect them. 1

Abstract. Assertions test expected properties of run-time values without disrupting the normal computation of a program. Here we present a library for enriching programs in the lazy language Haskell with assertions. Expected properties are written in an expressive pattern logic that combines pattern matching with logical operations and predicates. The presented assertions are lazy: they do not force evaluation but only examine what is evaluated by other parts of the program. They are also prompt: assertion failure is reported as early as possible, before a faulty value is used by the main computation. 1

Abstract. Assertions test expected properties of run-time values without disrupting the normal computation of a program. Here we present a library for enriching programs in the lazy language Haskell with assertions. Expected properties are written in an expressive pattern logic that combines pattern matching with logical operations and predicates. The presented assertions are lazy: they do not force evaluation but only examine what is evaluated by other parts of the program. They are also prompt: assertion failure is reported as early as possible, before a faulty value is used by the main computation. 1

Regular expressions are used to match and extract text. It is easy for developers to make syntactic mistakes when writing regular expressions, because regular expressions are often complex and different across programming languages. Such errors result in exceptions at run time, and there is currently no static support for preventing them. This paper describes practical experience designing and using a type system for regular expressions. This type system validates regular expression syntax and capturing group usage at compile time instead of at run time — ensuring the absence of PatternSyntaxExceptions from invalid syntax and IndexOutOfBoundsExceptions from accessing invalid capturing groups. Our implementation is publicly available and supports the full Java language. In an evaluation on five open-source Java applications (480kLOC), the type system was easy to use, required less than one annotation per two thousand lines, and found 56 previouslyunknown bugs.