andycyca / tracery-lists

Lists for use in Tracery

tracery-lists

Lists for use in Tracery

Kate Compton (@galaxykate) created Tracery: a story-grammar generation library for javascript. In order to use it, you need lists of words.

Here are some of mine. I use them with Allison Parrish's pytracery. Please let me know if you use these, makes me happy to know this work can help others.

It goes more or less like this:

import os
import tracery
from tracery.modifiers import base_english

# This assumes you only have .txt files in this path.
#
# And also that your base rules are contained in a file called 'origin.txt'
#   This will be the be-all, end-all of your creative engine
path = 'your/ad/could/be/here'

rules = {}
txtfiles = os.listdir(path)

for file in txtfiles:
    basename = file[:-4]  # remove the extension
    #print(basename)
    with open(path+file) as f:
            temp_list = f.read().splitlines()
            rules[basename] = temp_list

grammar = tracery.Grammar(rules)
grammar.add_modifiers(base_english)

for _ in range(5):
    print(grammar.flatten("#origin#"))

Don't forget to read some documentation at the wiki!

A note on fictional characters

1. The lists of fictional characters come exclusively from Public Domain sources. I don't know how copyright law applies to projects like this, storing and using copyright-protected characters for pseudo-random text creation. As such, they won't be admitted.
2. The first contributor (@andycyca) is well aware that women are under-represented in lists like these, in no small part because of centuries of biases in society at large and, in consequence, literature. I'm trying to add as many female characters as possible and would appreciate comments and suggestions to improve this list.
3. The above point also applies to other groups that have been marginalized and/or discriminated against throughout history; including (but not limited to) LGBTQ persons, native peoples, people of African descent and countless others. If anyone knows a better way of representing these groups, please let me know.
4. It's also worth noting that Public Domain characters can be re/written as having different sex, gender, nationality, background. Reimagining the Public Domain is another way of appropriating it. If anyone decides to do this, please don't just change the descriptions of a character's physical aspect and call it 'diverse'

See also:

• dariusk/corpora for a different, deeper approach to generating things