random_abstract.py

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

import html
import random
import requests
import sys


def get_abstract(abstract_id):
    '''Fetch an abstract by id from pubmed's JSON API'''
    abstract_id = str(abstract_id)

    url = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi?db=pubmed&retmode=json&rettype=abstract&id=' + abstract_id
    resp = requests.get(url)
    resp.raise_for_status()

    data = resp.json()
    results = data['result']
    item = results[abstract_id]

    error = item.get('error', None)
    if error == "cannot get document summary":
        return None
    if error != None:
        raise StandardError(error)

    return item


def find_latest_abstract():
    '''Do a binary search to find the most recently published paper id'''
    first, last = 29121706, 35000000

    if get_abstract(last):
        raise StandardError(
            'find_latest_abstract: far-future abstract %d unexpectedly exists' % last)

    while first <= last:
        midpoint = int((first + last) / 2)

        if get_abstract(midpoint):
            first = midpoint + 1
        else:
            last = midpoint - 1

    return first - 1


def get_random_abstract(min_id, max_id):
    '''Downloads a random abstract in the range, skipping deleted abstracts'''
    while True:
        selected_id = random.randint(min_id, max_id)
        abstract = get_abstract(selected_id)

        if abstract:
            return abstract


def compose_tweet(abstract):
    '''Turn a pubmed abstract into tweet text'''
    title = html.unescape(abstract['title'])
    url = 'https://www.ncbi.nlm.nih.gov/pubmed/%s' % abstract['uid']

    tweet_length = 280
    gap = "\n\n"
    ellipsis = "…"

    title_len = tweet_length - len(url) - len(gap)
    if len(title) > title_len:
        title = title[:title_len - len(ellipsis)] + ellipsis

    return title + gap + url


def main():
    if (sys.version_info < (3, 0)):
        print('python 3 is required to run this script')
        return

    earliest_id = 25635855  # a paper from roughly 10 yrs go
    latest_id = find_latest_abstract()  # the latest paper

    abstract = get_random_abstract(earliest_id, latest_id)
    tweet = compose_tweet(abstract)

    print(tweet)


if __name__ == '__main__':
    main()