Kazutaka Matsuda

Bidirectionalization transformation based on automatic derivation of view complement functions

Bidirectional transformation is a pair of transformations: a view function and a backward transformation. A view function maps one data structure called source onto another called view. The corresponding backward transformation reflects changes in the view to the source. Its practically useful applications include replicated data synchronization, presentation-oriented editor development, tracing software development, and view updating in the database community. However, developing a bidirectional transformation is hard, because one has to give two mappings that satisfy the bidirectional properties for system consistency. In this paper, we propose a new framework for bidirectionalization that can automatically generate a useful backward transformation from a view function while guaranteeing that the two transformations satisfy the bidirectional properties. Our framework is based on two known approaches to bidirectionalization, namely the constant complement approach from the database community and the combinator approach from the programming language community, but it has three new features: (1) unlike the constant complement approach, it can deal with transformations between algebraic data structures rather than just tables; (2) unlike the combinator approach, in which primitive bidirectional transformations have to be explicitly given, it can derive them automatically; (3) it generates a view update checker to validate updates on views, which has not been well addressed so far. The new framework has been implemented and the experimental results show that our framework has promise.

Bidirectionalizing graph transformations

Bidirectional transformations provide a novel mechanism for synchronizing and maintaining the consistency of information between input and output. Despite many promising results on bidirectional transformations, these have been limited to the context of relational or XML (tree-like) databases. We challenge the problem of bidirectional transformations within the context of graphs, by proposing a formal definition of a well-behaved bidirectional semantics for UnCAL, i.e., a graph algebra for the known UnQL graph query language. The key to our successful formalization is full utilization of both the recursive and bulk semantics of structural recursion on graphs. We carefully refine the existing forward evaluation of structural recursion so that it can produce sufficient trace information for later backward evaluation. We use the trace information for backward evaluation to reflect in-place updates and deletions on the view to the source, and adopt the universal resolving algorithm for inverse computation and the narrowing technique to tackle the difficult problem with insertion. We prove our bidirectional evaluation is well-behaved. Our current implementation is available online and confirms the usefulness of our approach with nontrivial applications.

Combining syntactic and semantic bidirectionalization

Matsuda et al. [2007, ICFP] and Voigtländer [2009, POPL] introduced two techniques that given a source-to-view function provide an update propagation function mapping an original source and an updated view back to an updated source, subject to standard consistency conditions. Being fundamentally different in approach, both techniques have their respective strengths and weaknesses. Here we develop a synthesis of the two techniques to good effect. On the intersection of their applicability domains we achieve more than what a simple union of applying the techniques side by side delivers.

Three complementary approaches to bidirectional programming

This paper surveys three distinct approaches to bidirectional programming. The first approach, syntactic bidirectionalization, takes a program describing the forward transformation as input and calculates a well-behaved reverse transformation. The second approach, semantic bidirectionalization, is similar, but takes the forward transformation itself as input rather than a program describing it. It requires the transformation to be a polymorphic function and uses parametricity and free theorems in the proof of well-behavedness. The third approach, based on bidirectional combinators, focuses on the use of types to ensure well-behavedness and special constructs for dealing with alignment problems. In presenting these approaches, we pay particular attention to use of complements, which are structures that represent the information discarded by the transformation in the forward direction.

Refactoring pattern matching

Defining functions by pattern matching over the arguments is advantageous for understanding and reasoning, but it tends to expose the implementation of a datatype. Significant effort has been invested in tackling this loss of modularity; however, decoupling patterns from concrete representations while maintaining soundness of reasoning has been a challenge. Inspired by the development of invertible programming, we propose an approach to program refactoring based on a right-invertible language rinv-every function has a right (or pre-) inverse. We show how this new design is able to permit a smooth incremental transition from programs with algebraic datatypes and pattern matching, to ones with proper encapsulation, while maintaining simple and sound reasoning.

A grammar-based approach to invertible programs

Program inversion has many applications such as in the implementation of serialization/deserialization and in providing support for redo/undo, and has been studied by many researchers. However, little attention has been paid to two problems: how to characterize programs that are easy or hard to invert and whether, for each class of programs, efficient inverses can be obtained. In this paper, we propose an inversion framework that we call grammar-based inversion, where a program is associated with an unambiguous grammar describing the range of the program. The complexity of the grammar indicates how hard it is to invert the program, while the complexity is related to how efficient an inverse can be obtained.

Enhancing Semantic Bidirectionalization via Shape Bidirectionalizer Plug-ins

Matsuda et al. (2007) and Voigtlander (2009) have introduced two techniques that given a source+to-view function provide an update propagation function mapping an original source and an updated view back to an updated source, subject to standard consistency conditions. Previously, we developed a synthesis of the two techniques, based on a separation of shape and content aspects (Voigtlander et al. 2010). Here, we carry that idea further, reworking the technique of Voigtlander such that any shape bidirectionalizer (based on the work of Matsuda et al. or not) can be used as a plug-in, to good effect. We also provide a data-type-generic account, enabling wider reuse, including the use of pluggable bidirectionalization itself as a plug-in.

Gradual refinement: blending pattern matching with data abstraction

Pattern matching is advantageous for understanding and reasoning about function definitions, but it tends to tightly couple the interface and implementation of a datatype. Significant effort has been invested in tackling this loss of modularity; however, decoupling patterns from concrete representations while maintaining soundness of reasoning has been a challenge. Inspired by the development of invertible programming, we propose an approach to abstract datatypes based on a right-invertible language RINV--every function has a right (or pre-) inverse. We show how this new design is able to permit a smooth incremental transition from programs with algebraic datatypes and pattern matching, to ones with proper encapsulation (implemented as abstract datatypes), while maintaining simple and sound reasoning.

FliPpr: a prettier invertible printing system

When implementing a programming language, we often write a parser and a pretty-printer. However, manually writing both programs is not only tedious but also error-prone; it may happen that a pretty-printed result is not correctly parsed. In this paper, we propose FliPpr, which is a program transformation system that uses program inversion to produce a CFG parser from a pretty-printer. This novel approach has the advantages of fine-grained control over pretty-printing, and easy reuse of existing efficient pretty-printer and parser implementations.

Polynomial-time inverse computation for accumulative functions with multiple data traversals

Inverse computation has many applications such as serialization/deserialization, providing support for undo, and test-case generation for software testing. In this paper, we propose an inverse computation method that always terminates for a class of functions known as parameter-linear macro tree transducers, which involve multiple data traversals and the use of accumulations. The key to our method is the observation that a function in the class can be regarded as a non-accumulative context-generating transformation without multiple data traversals. Accordingly, we demonstrate that it is easy to achieve terminating inverse computation for the class by context-wise memoization of the inverse computation results. We also show that when we use a tree automaton to express the inverse computation results, the inverse computation runs in time polynomial to the size of the original output and the textual program size.