leewz / lisp.py Goto Github PK

Run main.py with no arguments to start the REPL.

python main.py

Run with one argument to run a file.

python main.py mylispcode.lsp

I'm not a real Lisper, so this isn't a real Lisp dialect. I used some of Arc Lisp's names.

Syntax:

• Case-insensitive.
• Literal strings (using ") and numbers (not fractions) are implemented.
  • Characters and backslash escapes are not implemented.
• A ' quotes (i.e. leaves unevaluated) the next item, whether a literal or a list. E.g. 'A has the value of the symbol A.
• A . before the last item makes that item the tail. E.g. (CDR '(A B)) => (B), but (CDR '(A . B)) => B.
• Comments start with ; and extend to the end of the line.

Semantics:

• This implementation is a Lisp-1, meaning functions and variables share the same namespace.

These functions are defined within the Python code:

• (READ): Read and parse a single object from standard input.
• (READ string): Parse a string into an object.
• (EVAL expr): Evaluate an item.
  • Symbols evaluate to their value in scope.
  • Lists evaluate as function/macro calls.
  • Literals evaluate to themselves.
• (PRINT val): Print a single value.
• (= name val): Assignment. Evaluates name to a symbol and val to a value, and assigns it in the current scope.
• (ATOM val): True if val is not a CONS cell.
• (CONS head tail): Creates a pair (head . tail).
• (CAR lst): Gets the first part of a CONS pair. Usually the head of a list.
• (CDR lst): Gets the second part. Usually the tail of a list.
• (FUNCTION PARAMS BODY): Make an anonymous function (AKA a lambda)
• (MACRO PARAMS BODY): Make an anonymous macro. A macro, when called, gets its args unevaluated, and its return value will be evaluated instead.
• T, NIL: Constants for True and False. NIL is also considered the empty list.
• (QUOTE val): val itself, without evaluating.
• (SYMBOL string): Creates a symbol. Two symbols with the same name are the same symbol.
• (EQ x y): Test if two values are equal.
• (IF cond then else): If cond, then then, else else. Only the successful branch is evaluated.

Basic numerical operators are defined in numerics.py. Other functions are defined in startup.lpy. main.py loads both.

Will probably hit Python's recursion limit.

Example:

(= 'pi 3.1415926)
(= 'area-circle (function (r) (* pi (* r r))))