Skip to content

Latest commit

 

History

History
80 lines (58 loc) · 1.59 KB

README.md

File metadata and controls

80 lines (58 loc) · 1.59 KB

NanGram

Negative Nancy's expression generator & parser! XD

Write grammars, generate random expressions, parse expressions.

Installation

pip install nangram

Usage

First write yourself a nice little grammar like this!:

grammar.bnf:

determiner = "the " | "a " .
noun = "cat " | "dog " | "chair " .
verb = "runs " | "snacks " | "plays " .
adverb = "quickly" | "ferociously" | "sneakily" .

noun_phrase = determiner noun .
verb_phrase = verb adverb .
sentence = noun_phrase verb_phrase .

main = sentence .

Then import nangram and load the grammar source:

import nangram

grammar = nangram.Grammar.load_bnf('grammar.bnf')

Then generate some expressions!:

for expression in grammar.generate():
    print(expression)

And you get...

the cat plays ferociously
the dog plays ferociously
the chair snacks quickly
a cat snacks quickly
a dog runs quickly
a dog snacks ferociously
a dog plays quickly
a chair plays ferociously

Then try parsing something:

for parse_tree in grammar.parse('a cat snacks sneakily'):
    print(parse_tree)

And you get...

sentence = 'a cat snacks sneakily'
├ noun_phrase = 'a cat '
│ ├ determiner = 'a '
│ └ noun = 'cat '
└ verb_phrase = 'snacks sneakily'
  ├ verb = 'snacks '
  └ adverb = 'sneakily'

(The parser will yield as many parse trees as are valid, so if you have an ambiguous grammar, for example, you can parse all variations. If the expression is not in the language, you won't get any parse trees.)

See examples/english.bnf and examples/english.py for another example.