Semantics of Programming Languages

Overview

Exam

We will write the exam using Isabelle, with the known submission system - bring your laptop! All course material (book, slides, sheets, theories) is allowed. During the exam, you may only use the internet to submit your solution.

Material

• Book: Concrete Semantics including slides and demo theories
• Some motivational slides
• Exercises

Homework

Homework is the heart and soul of this course.

• Solved homework should be uploaded to the submission system, according to the instructions on the first exercise sheet. Make sure that your submission gets a “Passed” status in the system. We will not grade it otherwise!
• The latest submission date is given on each exercise sheet. Late submissions will not be graded! If you have a good excuse (such as being very sick), you should contact the tutors before the deadline.
• Each homework will get 0 to 10 points, depending on the correctness and quality of the solution.
• Discussing ideas and problems with others is encouraged. When working on homework problems, however, you need to solve and write up the actual solutions alone. If you misuse the opportunity for collaboration, we will consider this as cheating.
• Plagiarizing somebody else’s homework results in failing the course immediately. This applies for both parties, that is, the one who plagiarized, and the one who provided his/her solution.
• Important: all homework is graded and contributes 50% towards the final grade.

Aims

The aim of this course is to introduce the structural, operational approach to programming language semantics. It will show how this formalism is used to specify the meaning of some simple programming language constructs and to reason formally about semantic properties of programs and of tools like program analyzers and compilers. For the reasoning part the theorem prover Isabelle will be used.

At the end of the course students should:

• be familiar with rule-based presentations of the operational semantics of some simple imperative program constructs,
• be able to prove properties of an operational semantics using various forms of induction and
• be able to write precise formal proofs with the theorem prover Isabelle.

Important Notice

• You must be familiar with the basics of some functional programming language like Haskell, Objective Caml, Standard ML or F# (as taught, for example, in Introduction to Informatics 2 (IN0003)). For motivated students who do not have the necessary background yet: There are many introductions to functional programming available online, for example the first 6 chapters of Introduction to Objective Caml.
• You must haven taken some basic course in discrete mathematics where you learned about sets, relations and proof principles like induction (as taught, for example, in Discrete Structures (IN0015)).
• You need not be familiar with formal logic, but you must be motivated to learn how to write precise and detailed mathematical proofs that are checked for correctness by a machine, the theorem prover Isabelle.
• At the end of the course there will be a written, oral, or remote examination, depending on the number of students. Throughout the course there will be homework assignments. They will involve the use of Isabelle and will be graded. The final grade will be a combination of the examination and the homework grades, though you need to pass the exam.
• You must be subscribed to the Zulip announcements stream. All announcements will be made there.
• Please install Isabelle 2021-1 from the website and run it once before the fist tutorial (the first start builds the base session and takes some time).

Literature

• The primary reference: Tobias Nipkow, Gerwin Klein: Concrete Semantics with Isabelle/HOL
• Two traditional alternatives not based on proof assistants:
  • Hanne Riis Nielson, Flemming Nielson: Semantics with Applications: A Formal Introduction.
  • Glynn Winskel: The Formal Semantics of Programming Languages. MIT Press.