Closed ryantmoore closed 5 years ago
@kurtawirth Tested this using search #MAGA
and timeout = 30
, finding 43 records and one duplicate. For the repeated user ID, the elements that should be constant from the bot-identification were constant, and the elements that should be different (tweet text) were different.
Currently, we scrape tweets, store all their handles, and search each handle in sequence. We should store only the distinct handles, then search that subset, then (likely) merge the results back into the tweet data.