To get started you'll need to sign up for the app and register for a partner account. In "free mode", you can submit up to 100 requests in 24 hours. Once you obtain your API key, export it as an environment variable in your shell. (Aternatively, assign your key to a new variable once you're in python.)

In shell:

$ pip install -r requirements.txt

$ export SMMRY_API_KEY='YOUR API KEY HERE'

In Python:

import os

SMMRY_API_KEY = os.environ.get('SMMRY_API_KEY')

Altnernatively...

SMMRY_API_KEY = 'YOUR API KEY HERE'

Once your have your key, import the module and create an instance of the Smmry object.

from smmryapi import SmmryAPI

smmry = SmmryAPI(SMMRY_API_KEY)

To request the summary for an article pass an article's URL to the .summarize() method. A URL is the only required parameter, and by default, the method will return a seven-sentence summary. I'll use this this BBC article as an example with sm_length=3 to request a three-sentence summary.

(The maximum sentences retrieved is 40.)

url = 'http://www.bbc.com/news/business-43298897'

s = smmry.summarize(url,sm_length=3)

The method returns a Summary object with a varying set of attributes (depending on your parameters).

Article summaries will be store in the sm_api_content attribute.

s.sm_api_content

'Too many Lego bricks is a problem many parents will sympathise with, but now the toy firm itself has admitted it has made too many. 915,103,765 - the number of ways to combine six two-by-four Lego bricks of the same colour. 3,700 - the number of different types of Lego bricks. In September, Lego said its half-year results had suffered because it had stretched itself too thin by diversifying into products that were not toys, such as the Lego movies. Lego chairman Jorgen Knudstorp said at the time that adding complexity to the company had made it harder for the toymaker to grow further.'

If you pass an integet to sm_keyword_count it will return the requested number of keywords for that article. Note that the sm_api_keyword_array will only be available if a value for sm_keyword_count is selected.

s = smmry.summarize(url,sm_length=3,sm_keyword_count=12)

s.sm_api_keyword_array

To check out other attributes view the official documentation or or run s.__dict__.keys().

s = smmry.summarize(url,sm_length=5,sm_keyword_count=3)
    
attributes = (
    s.sm_domain,               # The domain name of the URL
    s.sm_api_title,            # The article's titled 
    s.sm_url,                  # URL of the article 
    s.sm_api_content_reduced,  # Percent by which reduced
    s.sm_length,               # Number of sentences 
    s.sm_requests_remaining,   # Number of queries left
    s.sm_api_keyword_array     # Keywords
)

result = """

SMMRY reduced %s article "%s" from url
[%s] by %s to %s sentences.

You have %s requests remaining today.

Top keywords are: %s.'
"""

print(result % attributes)

SMMRY reduced bbc.com article "Lego admits it made too many bricks" from url
[http://www.bbc.com/news/business-43298897] by 80% to 5 sentences.

You have 5 requests remaining today.

Top keywords are: ['Lego', 'sales', 'year'].'