Grafting-Trees/graft_a_tree.py

#!/usr/bin/env python
# -*- coding: <utf-8> -*-

'''
It automatises the Oulipo constraint Litterature Definitionelle:
http://oulipo.net/fr/contraintes/litterature-definitionnelle

invented by Marcel Benabou in 1966
In a given phrase, one replaces every significant element (noun, adjective, verb, adverb) 
by one of its definitions in a given dictionary ; one reiterates the operation on the newly received phrase, 
and again.


Copyright (C) 2018 Constant, Algolit, An Mertens, Gijs de Heij

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details: <http://www.gnu.org/licenses/>.

'''

import nltk
from nltk.corpus import wordnet as wn
from nltk import word_tokenize
import nltk.data
import random
import re
import time
import wikipedia
from trees import trees
import os
import os.path

import sys
from io import StringIO

def choose_a_tree(dict):
	collection = random.choice(list(trees.items()))
	# Give name gardener
	gardening = collection[0]
	cut = " in "
	if cut in gardening:
		index = re.search(cut, gardening).start()
		gardener = gardening[:index]
		source = gardening[index+3:]
	else:
		gardener = gardening
		source = 'Unknown'
	# Trees (s)he planted
	tree = collection[1]
	# if (s)he planted more than 1 tree, pick 1
	if type(tree) == list:
		tree = random.choice(tree)
	return gardener, source, tree

# find bio gardener on Wikipedia
def bio(gardener):
	try:
		bio = wikipedia.page(gardener)
		short_bio = bio.summary
	except:
		short_bio = "There is no English Wikipedia page about this person."
	return short_bio

def bio_offline(gardener):
  filename = os.path.join('bios', re.sub(r'\W', '', gardener))

  if os.path.exists(filename):
    with open(filename, 'r') as h:
      return h.read()
  
  return "There is no English Wikipedia page about this person."


"""
  Return blossoming version of the knot.
  knot string
  return string
"""
def grow_new_branch (bud):
  branch = wn.synsets(bud)

  if branch:
    return branch[0].definition()
    
  return bud

# Turn nltk tree back into a sentence
def trim_tree (branches):
  tree = " ".join([branch[0] for branch in branches])
  # Remove spaces in html tags
  tree = re.sub(r'\<\s*(\/?\w+)\s*\>', '<\\1>', tree)
  return re.sub(r'\s[\.,:;\(\)]', '', tree)

def graft_a_tree (years = 3):
  gardener, source, tree = choose_a_tree(trees)
  short_bio = bio_offline(gardener)

  print(
    '<section>'
    + 'Using a quote by <em>{}</em><br><br>'.format(gardener) \
    + short_bio
    + '</section>'
  )

  year = 1

  print('<p class="tree first">{}</p>'.format(tree))

  while year <= years:
    branches = nltk.pos_tag(word_tokenize(tree))
    # Filter out nouns, pick one
    position, bud = random.choice([(position, bud[0]) for position, bud in enumerate(branches) if bud[1] == 'NN'])

    tree_with_bud = '{} <em>{}</em> {}'.format(
      trim_tree(branches[:max(0, position)]),
      bud.strip(),
      trim_tree(branches[position+1:]))

    print('<p class="tree" data-year="{}">{}</p>'.format(year, tree_with_bud))

    # Make new tree placeholder for the bud show we can show it in two states and then grow out
    grown_tree = '{} <strong>{}</strong> {}'.format(
      trim_tree(branches[:max(0, position)]),
      grow_new_branch(bud).strip(),
      trim_tree(branches[position+1:]))
    # Let it grow again

    print('<p class="tree" data-year="{}">{}</p>'.format(year, grown_tree))

    grown_tree = re.sub(r'\<\/?\w+\>', '', grown_tree)

    if grown_tree == tree:
      break
    else:
      tree = grown_tree

    year += 1

if __name__ == '__main__':
  graft_a_tree(3)