C++Now 2018: Michał Dominiak “From Parsing to sema: Making Sense of Syntax Trees”

—
—
So you are writing a compiler, or an interpreter, or something else that uses a language that has to be analyzed, and you've managed to write a correct parser for your syntax, whichever way you chose to do that. Now what? When I've previously stumbled upon this question, when writing such code myself, I didn't really know the answer. I just started poking around the problem until I got something that worked. I've tried some ways that I hated, and I tried some ways that I ended up going with. This talk will take you on a journey that follows the approaches I tried, including some failures and some success stories, and while they aren't very numerous, I believe that looking at the challenges through the eyes of somebody else will help you when you're faced with similar design decisions yourself.

This talk is by no means meant to be a comprehensive guide to writing analyzers; it's meant to explain how I approached the challenge of making sense of a syntax tree, it's meant to be fun, and I also hope to learn from your experience in this field.
—
Michał Dominiak
Nokia Networks
Engineer, Software Developer

Bachelor of Engineering in Computer Science from Wrocław University of Technology. Professional and enthusiast programmer. Interested in the foundational problems of computing, from the lowest levels of CPU architectures, through operating systems and programming language design. Sometimes also a guitar player and singer.
—
---

*--*

---

[LesleyLai]

I don't thing "static" and "dynamic" polymorphism are the right terms here. Both virtual dispatch and variant are dynamic, the difference is open vs close.