hindent 2.0.0

Contains the class `Hindent`. Import and use this class to format and execute Scheme code.

Hindent is a syntax wrapper around lisp/scheme. The difference between Hindent and lisp/scheme is very simple: wherever lisp/scheme uses parenthesis, Hindent uses hanging indents. For example:

# Hindent

+
  2
  2

# lisp

(+ 2 2)

And some more examples:

display
  +
    2
    2

(display (+ 2 2))

display
  car
    quote
      1
        2
        3

(display (car (quote (1 2 3))))

define
  x
  4

+
  2
  x

(define x 4)
(+ 2 x)

How it works

Hindent simply parses the Hindent code and inserts parenthesis where appropriate to convert it to lisp/scheme. So, it is mostly a parser, but a fancy word for it could be a transpiler.

Installation

1. pip install hindent
2. Install chez scheme (brew install chezscheme) and make sure it's on your path by running chez.
   • This also works with other lisps or schemes. If you want to do that, the notes in the initialize method will explain how to do it.
3. The syntax highlighter for vs code is Hindent Lang. You can find it in the normal extensions menu / marketplace.

Usage

I think a nice way to interactively run programs is with jupyter. So, I recommend opening up a jupyter notebook. Then, simply start to import Hindent with from Hindent, and the docstrings/code-hover will guide you from there.

One heads up is that when running, the newline and display functions are helpful. If you don't use display, then lisp will not output anything. So, if you are experimenting, and you want to see results of what you run, you need to use display. And regarding newline, if you don't use it, all of the output will be smushed together.

Parsing Spec

• the transpiler supports literate programming. It does so by putting block comments on equal footing with code. To toggle back and forth between code and block comment, put a lone period on a line.
  • For reasons of textmate grammars and syntax highlighting, we had to make the file start as code, rather than starting as block comment. We would have preferred it the other way, but we think it's okay. We're thankful for how awesome vs code and textmate are.
• the parser/transpiler will put a new line after the source code to ensure the final dedent is correct
• any amount of whitespace that contributes to a blank line (other than the last line) should be replaced with a single new line character. For example, \n\n\n should be replaced with \n, and \n \n should also be replaced with \n
• lines that start with ; are comments and will be ignored
• calculate the number of spaces each line starts with.
• Then divide by 2 to determine the indentation level
• then, for each line find the difference in indentation level between that line and the following line
  • if the indentation level goes up one, then add an opening parenthesis before the line
  • if the indentation level goes down, then add a closing parenthesis after the line according to the number of dedents.
  • if the indentation level remains the same, then we have to check one more thing:
    • if the current indentation level (and also the indentation level of the following line) is at 0, then wrap the current line in parenthesis
    • if the current indentation level (and also the indentation level of the following line) is nonzero, then do nothing

ToDo

• allow for end-of-line comments -- this one is important because the syntax highlighter already does this correctly, and it will be confusing if the syntax highlighter works, but the code doesn't.
• maybe get it out of running as a subprocess so it returns things more easily
• make a sphinx site
  • convert readme to rst (change in pyproject.toml as well)
• push to conda (in addition to pypi)

Notes For Grant

flit build --format wheel twine upload dist/*