L-attributed Grammars Research Articles

Inferring Absolutely Non-Circular Attribute Grammars with a Memetic Algorithm

When valid syntactical structures are additionally constrained with context-sensitive information the Grammar Inference needs to be extended to the Semantic Inference. In this paper, it is shown that a complete compiler/interpreter for small Domain-Specific Languages (DSLs) can be generated automatically solely from given programs and their associated meanings using Semantic Inference. In this work a wider class of Attribute Grammars has been learned, while only S-attributed and L-attributed Grammars have previously been inferred successfully. Inferring Absolutely Non-Circular Attribute Grammars (ANC-AG) with complex dependencies among attributes has been achieved by integrating a Memetic Algorithm (MA) into the LISA.SI tool. The results show that the proposed Memetic Algorithm is at least four times faster on the selected benchmark than the previous method.

Teaching machines to write like humans using L-attributed grammar

Reading and writing are easy for humans. The automatic reading of handwritten characters has been studied for several decades. Machine learning algorithms for reading tasks often require a huge amount of data to perform with similar accuracy to humans, yet it is also difficult to gain sufficient meaningful data. Automatic writing tasks have not been studied as extensively. In this paper, we teach machines to write like teaching a child by telling the machine the method for writing each character using L-attributed grammar. With the aid of the proposed TMTW (Teaching Machines To Write) interacting system, a human as a teacher only needs to provide the writing sequence of parts and control lines. The proposed system automatically perceives the relationships between control lines and parts, and constructs the grammars. Top-down derivation and the stroke generation method are applied to generate varying characters based on the learned grammars. For as long as a machine can write, it can be applied in robot control or training sample generation for automatic reading tasks. The MNIST and CASIA datasets are used to demonstrate the effectiveness of the proposed system on different languages. The machine written samples are used to train a network, which is evaluated on the MNIST test set. A test error rate of 1.23% is achieved using only approximately 20 grammars on average for each digit. Using the generated and handwritten samples together as a training set can reduce the test error rate to 0.61%. Similar experiments are conducted using the CASIA data set, and the results demonstrated that the proposed method is effective in generating characters with a complex structure. The source codes and grammars used in this paper have been made publicly available in https://github.com/step123456789/TMTW.

From object algebras to attribute grammars

Oliveira and Cook (2012) and Oliveira et al. (2013) have recently introduced object algebras as a program structuring technique to improve the modularity and extensibility of programs. We analyze the relationship between object algebras and attribute grammars (AGs), a formalism to augment context-free grammars with attributes. We present an extension of the object algebra technique with which the full class of L-attributed grammars - an important class of AGs that corresponds to one-pass compilers - can be encoded in Scala. The encoding is modular (attributes can be defined and type-checked separately), scalable (the size of the encoding is linear in the size of the AG specification) and compositional (each AG artifact is represented as a semantic object of the host language). To evaluate these claims, we have formalized the encoding and re-implemented a one-pass compiler for a subset of C with our technique. We also discuss how advanced features of modern AG systems, such as higher-order and parameterized attributes, reference attributes, and forwarding can be supported.

On syntax-directed adjoint fortran code

Gradients of high-dimensional functions can be computed efficiently and with machine accuracy by so-called adjoint codes. We present an L-attributed grammar for the single-pass generation of intraprocedural adjoint code for a subset of Fortran. Our aim is to integrate the syntax directed approach into the front-end of the NAGWare Fortran compiler. Research prototypes of this compiler that build adjoint code based on an abstract intermediate representation have been developed for several years. We consider the syntax-directed generation of adjoint code as a low development cost alternative to more sophisticated algorithms. The price to pay comes in form of a very limited set of code optimizations that can be performed in a single-pass setting.

PAGELEARN: Learning Semantic Functions of Attribute Grammars in Parallel

Attribute Grammars (AGs) are a generalization of the concept of Context-Free Grammars (CFGs). The formalism of AGs has been widely used for the specification and implementation of programming languages. On the other hand there is an intimate relationship between AGs and Logic Programming. The paper presents a parallel method for learning semantic functions of Attribute Grammars (AGs) based on AGLEARN (2) using PAGE system (12). The method is more efficient in both execution time and interaction needed than the sequential one. The method presented is adequate for S-attributed grammars and for L-attributed grammars as well.

Uncle-attributed grammars

Evaluation of inherited attributes is a problem in conjunction with LR parsing, because the derivation tree is incomplete during parsing. An evaluation scheme for inherited attributes is presented based on a restricted grammar class, uncle-attributed grammars. A transformation to the uncle-attributed form is described for L-attributed grammars.