I just realised while reading the Wikipedia article on S-expressions that I have really overcomplicated the design of a lisp interpreter. Because sexprs have such a simple design, they can be parsed much simpler and faster than using a traditional grammar approach like this project uses.
That being said, I'm going to leave this project as is because I still think it is a solid use of grammar passing techniques, but I will implement a Lisp interpreter again, taking advantage of the simplicity of sexprs. When I do, I'll link to it here.
I've grown to have a big appreciation for languages, language design, and language implementation. At some point in the near future, I plan on building a compiler (probably C), but before that, I figured that an interpreter would be a good way to get familiar with the first half of the compilation process (lexical analysis and syntactic analysis)
Lisp in an interesting family of languages, in that they aren't very similar to most of the big languages used in production when in comes to syntax. They do however act as a great tool for teaching the fundamentals of functional programming.
I chose to write a Lisp style language because they have a simple syntax, so it would be a great starting place for someone like me trying to get involved in language development.
I've implemented the following:
- Assignment
(setq v 10) - Function definition
(defun add10 (x) (+ 10 x)) - Function calling
(func param1 param2) - Types:
- Integers:
10,-5 - Strings:
"hello world" - Functions
- Lists:
(list 1 2 3 4 5)
- Integers:
- If expressions:
(if cond then else)
There are examples in the examples directory of this repository
No language is without it's quirks
- There are no booleans, instead we use numbers to represent (similar to C)
+, -, /, * - Arithmetic operators (2+ numbers)
>, <, <=, >=, ==, != - Comparison operators (2 numbers)
print - Prints given value (1 value)
list - Creates a list (1+ values)
head - Returns first element of list (1 list)
tail - Returns list of all elements but first (1 list)
length - Returns length of list (1 list)
append - Appends given value to list (1 list, value)
splice - Returns list spliced by parameters (list, start, end)
nth - Returns value at the n-th position in the list (indexed at 0) (list, index)
There are two ways of running the interpreter. You can access a REPL by running
repl.py with no arguments. This is a traditional read-evaluate-print-loop that's
great for testing basics of the interpreter.
The other is with run.py. run.py is used to run a program stored in another file.
It takes 1 argument, which is the file you'd like to run.
- Probably will need to add more comparison operator overloading for all expression types
- type checking for ops and user defined functions
- better errors
lambdaexpr - evaluates to FuncVAl
Document new code
possibly divide into new files??
add better checking of args to PrimOps
add list functions last init reverse nth