Evaluating Untyped Lambda-Expressions in Haskell

GitHub repository for a project about comparing various methods for evaluating untyped lambda expressions. The algorithms are implemented in Haskell, whereas the lexer/parser stack chosen is Alex+Happy. The complete though process for this project, along with a brief introduction to the world of lambda calculus, is described in the article and presentation.

Table of Contents

• Introduction
• Repository Structure
• Getting Started
• Article
• Presentation
• Source Code
  • Usage
  • Test Generation
• Citation
• License

Introduction

This repository explores different methods for evaluating lambda expressions, including two algebraic and a computational approach. It includes an article detailing the theoretical aspects, a presentation summarizing key points, and the source code to experiment with the various evaluation methods described.

Repository Structure

The repository is organized as follows:

. ├── article │ ├── article.pdf │ └── src │ ├── main.tex │ └── references.bib ├── presentation │ ├── presentation.pdf │ └── presentation.pptx ├── src │ ├── AlgebraicAlphaConversion.hs │ ├── AlgebraicDeBrujin.hs │ ├── Computational.hs │ ├── LambdaTerm.hs │ ├── parser.hs │ ├── parser.y │ ├── scanner.hs │ ├── scanner.x │ └── test │ └── testGenerator.py ├── README.md ├── CITATION.cff └── LICENSE

• article/: Contains the article PDF and its source files.
• presentation/: Contains the presentation in PDF and PPTX formats.
• src/: Contains the source code for different methods of evaluating lambda expressions.
• README.md: This README file.
• CITATION.cff: Citation information for this repository.
• LICENSE: The license under which this project is distributed.

Getting Started

To explore the contents of this repository, follow the steps below:

1. Clone the Repository:

bash git clone https://github.com/dambrosidenis/Lambda_Expression_Evaluation.git cd Lambda_Expression_Evaluation

1. View the Article:

Open the article/article.pdf file to read the full article.

1. View the Presentation:

Open the presentation/presentation.pdf or presentation/presentation.pptx file to review the presentation.

1. Explore the Source Code:

Navigate to the src/ directory to review the source code implementing various methods for evaluating lambda expressions. More information below.

Article

The detailed article explaining the different methods for evaluating lambda expressions can be found in the article/ directory:

• Article PDF

The source files for the article, including LaTeX and bibliography, are located in article/src/.

Presentation

The presentation summarizing the key points of the project can be found in the presentation/ directory:

• Presentation PDF
• Presentation PPTX

Source Code

The src/ directory contains the source code for various methods of evaluating lambda expressions, including:

• Algebraic Alpha Conversion method (AlgebraicAlphaConversion.hs)
• Algebraic De Bruijn method (AlgebraicDeBrujin.hs)
• Computational Evaluation method (Computational.hs)
• Common algebraic Lambda Term Definitions to parse the terms (LambdaTerm.hs)
• Parser written in Happy (parser.y) and compiled (parser.hs)
• Scanner written in Alex (scanner.x) and compiled (scanner.hs)
• Test instance Generator (test/testGenerator.py)

Usage

To run the program on an input file containing various lambda expression to evaluate, it is sufficient to execute

bash ghc parser.hs ./parser <input-file>

If, on the other hand, you prefer to recompile the lexer and the parser, you have to execute

bash alex scanner.x

to compile the scanner into scanner.hs, and then

bash happy parser.y

to obtain parser.hs.

Test Generation

The parser takes as input a text file with a series of lambda expressions. If you want to check your own expressions, the correct syntax for the file is unambiguously described in the article. If, on the other hand, you are good with randomly generated instances, you can use the Python script testGeneration.py to generate a list of random expressions obtained through a random walk on the grammar. The usage of the file is the following:

bash python3 testGeneration.py <output-file> <number-of-expressions> <max-depth>

where: - output-file is the name of the file where you want to save the expressions - number-of-expressions is the number of examples you want - max-depth specifies the highest level of term nesting possible in order to avoid excessively lengthy expressions

Note that in absence of any of the arguments, the script will use the default values ./test.txt, 10 and 5

Citation

If you use this work in your research, please cite it as follows:

@software{D_Ambrosi_Evaluating_Untyped_Lambda_2023, author = {D'Ambrosi, Denis}, month = sep, title = {{Evaluating Untyped Lambda Expressions in Haskell}}, url = {https://github.com/dambrosidenis/Lambda_Expression_Evaluation}, version = {1.0.0}, year = {2023} }

License

This project is licensed under the MIT License. See the LICENSE file for more details.