Introduction

Programming languages are fundamental tools for software developers. Their design shapes the way we think about problems, express solutions, and organize programs into maintainable systems.

Why study programming languages?

Every programmer needs a deep understanding of programming languages to learn them more easily, to use them more effectively, to evaluate their strengths and weaknesses (particularly when assessing a new language), and to design new ones.

Most programmers think they'll never design a new programming language. However, programming languages are one important tool of automation and abstraction. You may never need to create a full-fledged, general-purpose language, but you may need to design a special-purpose programming language (also called a domain-specific language) that eases solving a narrow class of problems.

Methodology

To deepen our understanding of programming languages, we are going to devise some. We will start with a very simple language, and incrementally build on that language to study increasingly more sophisticated ones. As we do, we will explore core concepts that appear in many modern and classic programming languages.

A programming language consists of two parts: syntax and semantics. Syntax describes the notation we use to write programs in a particular language. Semantics defines the meaning of programs written in a language; that is, what a program does when we run it.

The syntax of a language is the programmer's interface to a language. From this perspective, syntax is important. Poor syntax makes programming an onerous task. An better syntax can make a language less error prone, easier to type, easier to understand, and easier to maintain.

Semantics is the engine of a language. It establishes the expressive power of a language. It is where we decide if a language support iterative statements, conditional statements, exception handling, goto statements, functions, procedures, closures, type checking, classes and objects, out parameters, in-out parameters, passing parameters by reference, passing parameters by value, continuations, parallelism, side-effects, unification, etc.

Because semantics embodies the core of a programming language, it is where we want to spend most our time in this course. We are going to build many, rather ugly, languages. But these languages will support important and powerful features!

To ease this process, we are going to rely on a tool: PLCC - Programming Languages Compiler Compiler. It will allow us to simplify the process of building languages by automating the generation of their parsers, allowing us to focus our efforts on implementing their semantics.