NAMD / pypln.api

Python library to access PyPLN's API.

PyPLN Application Programming Interface

PyPLN is a distributed pipeline for natural language processing, made in Python. Learn more at the PyPLN website.

pypln.api is a package that interacts with PyPLN HTTP API to do everything programatically, in a Pythonic way. Basically, you are able to add/list corpora, add/list documents and retrieve documents' properties (resulted from the pipeline processing by the backend).

Installation

pypln.api is available at Python Package Index. So, to install it, just execute:

pip install pypln.api

Example - Usage

You can see docstrings inside pypln.api.PyPLN, but the general usage will be something like this:

from pypln.api import PyPLN

# Start an authenticated session to PyPLN demo server
pypln = PyPLN('http://fgv.pypln.org/', ('username', 'password'))

# You could also use your authentication token:
#pypln = PyPLN('http://fgv.pypln.org/', 'my-auth-token')

# Add a new corpus to your account
new_corpus = pypln.add_corpus(name='test', description='my new corpus')

# Add a document to this new corpus
with open('my-file.pdf') as fp:
    new_doc = new_corpus.add_document(fp)
print('Document added: {}'.format(new_doc))

# Retrieve all available (processed) properties for your brand new document
print('Processed properties:')
for document_property in new_doc.properties:
    print(' - {}'.format(document_property))

# Retrieve one document property:
print('Extracted text from our PDF:')
print(new_doc.get_property('text'))

# Retrieve a document using it's url:
from pypln.api import Document
# Make sure you replace this url for the url of a document you have access to!
my_doc = Document.from_url('http://fgv.pypln.org/documents/1/',
    ('username', 'password'))
print(my_doc.get_property('text'))

# Retrieve wordcloud image built from the document
with open("wordcloud_{}.png".format(doc_id), 'w') as fd:
    fd.write(base64.b64decode(my_doc.get_property("wordcloud")))

ProTip™: use ipython to discover all methods available at PyPLN, Corpus and Document classes - they are very simple and straightford to use.

License

pypln.api is free software, released under the GPLv3.