pycorpora 0.1.2

A Python wrapper for Darius Kazemi's Corpora Project

A simple Python interface for Darius Kazemi’s Corpora Project, “a collection of static corpora (plural of ‘corpus’) that are potentially useful in the creation of weird internet stuff.” The pycorpora interface makes it easy to use data from the Corpora Project in your program. Here’s an example of how it works:

import pycorpora
import random

# print a random flower name
print random.choice(pycorpora.plants.flowers['flowers'])

# print a random word coined by Shakespeare
print random.choice(pycorpora.words.literature.shakespeare_words['words'])

Allison Parrish created the pycorpora interface. Python 3 is not yet supported. The source code for the package is on GitHub. Contributions are welcome!

Installation

Installation by hand:

python setup.py install

Installation with pip:

pip install pycorpora

The package does not include data from the Corpora Project; instead, the data is downloaded when the package is installed (using either of the methods above). By default, the “master” branch of the Corpora Project GitHub repository is used as the source for the data. You can specify an alternative URL to download the data from using the argument --corpora-zip-url on the command line with either of the two methods above:

python setup.py install --corpora-zip-url=http://example.com/corpora.zip

… or, with pip:

pip install pycorpora --install-option="--corpora-zip-url=http://example.com/corpora.zip"

(The intention of --corpora-zip-url is to let you install Corpora Project data from a particular branch, commit or fork, so that changes to the bleeding edge of the project don’t break your code.)

Update

Update Corpora Project data by reinstalling with pip:

pip install –upgrade –force-reinstall pycorpora

Usage

Getting the data from a particular Corpora Project file is easy. Here’s an example:

import pycorpora
crayola_data = pycorpora.colors.crayola
print crayola_data["colors"][0]["color"] # prints "Almond"

The expression pycorpora.colors.crayola returns data deserialized from the JSON file located at data/colors/crayola.json in the Corpora Project (i.e., this file). You can use this syntax even with more deeply nested subdirectories:

import pycorpora
mr_men_little_miss_data = pycorpora.words.literature.mr_men_little_miss
print mr_men_little_miss_data["little_miss"][-1] # prints "Wise"

You can use from pycorpora import ... to import a particular Corpora Project category:

from pycorpora import governments
print governments.nsa_projects["codenames"][0] # prints "ARTIFICE"

from pycorpora import humans
print humans.occupations["occupations"][0] # prints "accountant"

You can also use square bracket indexing instead of attributes for accessing subcategories and individual corpora (just in case the Corpora Project ever adds files with names that aren’t valid Python identifiers):

import pycorpora
import random
fruits = pycorpora.foods["fruits"]
print random.choice(fruits["fruits"]) # prints "pomelo" maybe

Additionally, pycorpora supports an API similar to that provided by the Corpora Project node package:

import pycorpora

# get a list of all categories
pycorpora.get_categories() # ["animals", "archetypes"...]

# get a list of subcategories for a particular category
pycorpora.get_categories("words") # ["literature", "word_clues"...]

# get a list of all files in a particular category
pycorpora.get_files("animals") # ["birds_antarctica", "birds_uk", ...]

# get data deserialized from the JSON data in a particular file
pycorpora.get_file("animals", "birds_antarctica") # returns dict w/data

# get file in a subcategory
pycorpora.get_file("words/literature", "shakespeare_words")

As an extension of this interface, you can also use the get_categories, get_files and get_file methods on individual categories:

import pycorpora

# get a list of files in the "archetypes" category
pycorpora.archetypes.get_files() # ['artifact', 'character', 'event', ...]

# get an individual file from the "archetypes" category
pycorpora.archetypes.get_file("character") # returns dictionary w/data

# get subcategories of a category
pycorpora.words.get_categories() # ['literature', 'word_clues']

Examples

Here are a few quick examples of using data from the Corpora Project to do weird and fun stuff.

Create a list of whimsically colored flowers:

from pycorpora import plants, colors
import random

random_flowers = random.sample(plants.flowers["flowers"], 10)
random_colors = random.sample(
    [item['color'] for item in colors.crayola["colors"]], 10)
for pair in zip(random_colors, random_flowers):
    print " ".join(pair).title()

# outputs (e.g.):
#   Maroon Bergamot
#   Blue Bell Zinnia
#   Pink Flamingo Camellias
#   Tickle Me Pink Begonia
#   Burnt Orange Clover
#   Fuzzy Wuzzy Hibiscus
#   Outer Space Forget Me Not
#   Almond Petunia
#   Pine Green Ladys Slipper
#   Shadow Jasmine

Create random biographies:

from pycorpora import humans, geography
import random

def a_biography():
    return "{0} is a(n) {1} who lives in {2}.".format(
        random.choice(humans.firstNames["firstNames"]),
        random.choice(humans.occupations["occupations"]),
        random.choice(geography.us_cities["cities"])["city"])

for i in range(5):
    print a_biography()

# outputs (e.g.):
#   Jessica is a(n) ceiling tile installer who lives in Grand Forks.
#   Kayla is a(n) substance abuse social worker who lives in Torrance.
#   Luis is a(n) hydrologist who lives in Saginaw.
#   Leah is a(n) heating installer who lives in Danville.
#   Grant is a(n) building inspector who lives in Vineland.

Automated pizza topping-related boasts about your inebriation:

from pycorpora import words, foods
import random

# "I'm so smashed I could eat a pizza with spinach, cheese, *and* hot sauce."
print "I'm so {0} I could eat a pizza with {1}, {2}, *and* {3}.".format(
    random.choice(words.states_of_drunkenness["states_of_drunkenness"]),
    *random.sample(foods.pizzaToppings["pizzaToppings"], 3))

The possibilities… are endless.

History

• 0.1.2: Python 3 compatibility (contributed by Sam Raker); vastly improved build process (contributed by Hugo van Kemenade).

License

The pycorpora package is MIT licensed (see LICENSE.txt). The data in the Corpora Project is itself in the public domain (CC0).

Acknowledgements

Thanks to Darius Kazemi and all of the Corpora Project contributors!

This package was developed as part of my Spring 2015 research fellowship at ITP. Thank you to the program and its students for their interest and support!