 |
 |
Heya Eric, On 5/4/20 3:14 PM, Eric Lease Morgan wrote: > Despite having a Twitter Developer Account, I still can not seem to get Twarc to retrieve tweets older than thirty days, and I'm wondering what I'm doing incorrectly. > > I created and was granted by Twitter a developer account called nfc4ds. The account has a number of subscriptions, and one of them is called "Search Tweets: Full Archive / Sandbox". I then dutifully created a twitter application complete with an API key and secret. I then authorized Twarc (a Python tool/library) to use the application. I then used Twarc to search Twitter with this query: > > from:ericleasemorgan > > The result was a set of a few tweets. > > I then used the following query, but the results were empty: > > (from:ericleasemorgan) until:2019-12-31 since:2019-01-01 > What am I doing wrong? Is Twarc unable to query things from the past? Does my subscription not really include the full archive? Sadly in the fine print on your developer account is how far back in the twitter API you are allowed to search. Pretty sure that a search for the year 2019 is not included. My rule of thumb with twarc (really Twitter's API) is a you have a 7 days and occasionally you get "lucky" with the aforementioned 7 day "Francis rule". After 7 days depending on your search parameters it is "hit or miss". For better or worse we've chosen slack as the place we keep a pulse of twarc https://docs.google.com/forms/d/e/1FAIpQLSf3E7PAXPoT-XoedpEy9UCTpDPS8kPj5JkMwpaWbuqVP0bTrQ/viewform if you'd like an a more authoritative response, and/or an issue here https://github.com/DocNow/twarc/issues if you'd rather not "join another slack workspace", but my... gut is saying "Twitter won't let you do that" anyway. These folks have an "end around" on Twitter's API" that may address your use case. Or certainly I got results. https://github.com/twintproject/twint Cheers, ./fxk > > -- > Eric Morgan > University of Notre Dame >