The download page of many a language harbors deep senses of forboding, of evil lurking in its native lair. You feel that the language is not a friend, that it means to harm you. You wonder what will happen when you click on the download link.
But don't be scared. Nanolisp is a friend in a desert of programming languages. It can fit in a clipboard or an email. A file or a project. Just 3 lines of code long, it really has earned its name. It only relies on standard, cross-platform modules. It is an oasis for many a person.
Copy and paste. Et viola! You have a complete nanolisp download! Python 3.8 or greater is needed to run the code. But I promise you, it will be worth it.
The examples following this one will assume you have copy-pasted it into a file called
python3 nl.py to open up the REPL. There, you can do many operations. Here is a sample session:
$ python3 nl.py Welcome to Nanolisp 1.0.1! > (+ 1 1) 2 > (define a 5) > a 5 > (+ a 5) 10 > (define (inc n) (+ n 1)) > (inc 4) 5 > (exit) $
As you can see, Nanolisp offers a full-featured repl. To run from a file is similarly simple:
. In fact, this repository includes a test file, titled
schemetest.scm, which implements an algorithm for square root and runs it on 10 (the algorithm itself was taken from the seminal textbook on Scheme, Structure and Interpretation of Computer Programs).
Although I made this for fun, I quickly realized that this could be applied to config files and similar goals. This leads to the fairly large topic of a Domain Specific Language (DSL). Nanolisp is designed to be extensible and easily so. Unfortunately, due to its complex implementation, it's not extensible in all of its aspects.
To be continued...
Z is the variable used to store the environment. It has a custom type, but it was based off of Peter Norvig's implementation of the environment in lis.py. Why is it called Z? I have no idea.
Evaluate an AST (tuple of tuples).
Run a string.
rf (run file)
Open a file and run it (leaks memory, I guess)
Start the actual REPL, as opposed to the banner.
The name is pretty misleading; this is a list of transformations applied to the lisp code to turn it into valid python code. Feel free to add elements to it if you know what you're doing.
initl (init loader)
Allows you to import scheme files.