Check for duplicates while scraping

[Deleetdk]

Not re-scraping users has a cost, namely that one cannot examine changes for a profile over time. This could be interesting to examine, but it would complicate the data gathering, so perhaps best to postpone that for later.

[onbjerg]

Using maxogden/dat would also save the historical data. It does not add duplicates, but it updates the existing row with new information, while retaining the old information in the log (much like git). This would also fix #3

[Deleetdk]

Sounds good. :)