Language design

Question

So I’ve been doing some research into building a programming language and now know that they usually have separate lexers, parsers, abstract syntax trees and action trees. Could anyone explain about these in detail, particularly the differences between lexers and parsers and between ASTs and action trees? If I succeed in making one (both interpreted and compiled) I plan to create a course for it here! XD

Answer 1

Lexical analysis converts the characters found in the source code into tokens (e.g. 'i', 'n', 't' becomes a symbol "int" or '1', '2', '3' becomes a number 123.) Parsing deals with those tokens to recognize syntax (e.g. symbol equal number semicolon becomes an assignment statement [x=1;].) Abstract Syntax Trees is a way to store the tokens to reflect the source syntax. Using these trees make it easier for scope checking during semantic checking.

Don't know any action trees. Could you possibly mean activation trees?

Answer 2

This a more detailed course in compilers.
https://www.tutorialspoint.com/compiler_design/index.htm

Answer 3

He actually stated he made up the term as he didn't know what to call it. You were right about the article. AST in my mind are his action trees. He didn't add type information to his AST making the action tree by combining types with his AST. I'd leave the types empty while building the AST and fill them in his action tree step.

Answer 4

no i believe they say action trees (possibly an unrecognized term for activation trees?). Everything I know on this topic comes from these to articles:

https://medium.freecodecamp.org/the-programming-language-pipeline-91d3f449c919

https://blog.usejournal.com/writing-your-own-programming-language-and-compiler-with-JUMP_LINK__&&__python__&&__JUMP_LINK-a468970ae6df

I believe ATs were covered in the first