Simpler formatting of syntax trees: ((The quick brown fox) (jumps (over (the lazy dog))))

bllip.py

# Patch to make nltk.parse.bllip.py work in Python 3

def _ensure_ascii(words):
    try:
        for i, word in enumerate(words):
            if isinstance(word, bytes):
                word.decode('ascii')
            else:
                word.encode('ascii')
    except UnicodeDecodeError:
        raise ValueError("Token %d (%r) is non-ASCII. BLLIP Parser "
                         "currently doesn't support non-ASCII inputs." %
                         (i, word))

syntaxbrackets.py

"""
pip install nltk bllipparser prompt_toolkit
python -m nltk.downloader bllip_wsj_no_aux
"""

from nltk.data import find
from nltk.parse.bllip import BllipParser
from nltk.compat import string_types
from nltk.tree import Tree

from prompt_toolkit import prompt


def format_flat(tree, parens='()'):
    childstrs = []
    for child in tree:
        if isinstance(child, Tree):
            childstrs.append(format_flat(child, parens))
        elif isinstance(child, tuple):
            childstrs.append("/".join(child))
        elif isinstance(child, string_types):
            childstrs.append(child)

    if len(childstrs) == 0:
        return ''
    elif len(childstrs) == 1:
        return childstrs[0]
    else:
        return ''.join((parens[0], " ".join(childstrs), parens[1]))


def main():
    print('Loading parser..')
    model_dir = find('models/bllip_wsj_no_aux').path
    bllip = BllipParser.from_unified_model_dir(model_dir)

    print('Done.')
    print('Enter a sentence:')

    while True:
        text = prompt('> ')
        top_result = bllip.parse_one(text.split())
        print(format_flat(top_result))


if __name__ == '__main__':
    main()