Inspired by the wonderful @NYT_first_said, this project uses the MIT Center for Civic Media's Media Cloud project to find the earliest mention of a word in major English language newspapers. It uses this to issue a smugly superior tweet.
I do not have a setup.py for this, so you need to install from source:
$ git clone git@github.com:ColCarroll/seen_it_news.git
$ cd seen_it_news # and probably activate a virtual environment
$ pip install -r requirements.txt
There should be a file called .config.json configuring the project. The useful
grift library will complain intelligently about
each setting that you need to fill in. At the very least, you will need an API key for Media Cloud,
which you get by signing up for a free account, and
credentials for twitter, which
digital ocean's tutorial
does a good job of describing.
After installing and configuring, just python seen_it_news should check up to 20 recent tweets,
and respond to those it can. Data is stored in a sqlite database (history.db, by default), and
everything is logged to stdout.
Yes! You can get the string it plans to tweet with first_mention:
>>> from seen_it_news import FindFirstMention
>>> finder = FindFirstMention('vernatsch')
>>> print(finder.first_mention())
@NYT_first_said The Mercury News was saying "vernatsch" for 8 years before the New York Times did:
https://www.mercurynews.com/2009/03/10/wine-why-isnt-this-one-on-every-restaurants-list/
You can also get all historical uses as a generator of tuples (which makes it easier to use elsewhere):
>>> for mention in finder.all_mentions():
...: print(mention)
...:
(Media(name='The Wall Street Journal', id=1150, netloc='www.wsj.com'), 'Fall Wines: Go Red, but Light', 'https://www.wsj.com/articles/SB10000872396390443686004577635632300446956', datetime.datetime(2012, 9, 15, 0, 39, 5, tzinfo=tzutc()))
(Media(name='The Mercury News', id=35, netloc='www.mercurynews.com'), "Wine: Why isn't this one on every restaurant's list?", 'https://www.mercurynews.com/2009/03/10/wine-why-isnt-this-one-on-every-restaurants-list/', datetime.datetime(2009, 3, 10, 11, 44, 39, tzinfo=tzutc()))
Ouch! But fair! See Media Cloud's "Press and Publications" page for more academic uses of the project.