oksas / doomscrape Goto Github PK

Look into just queueing up the entire directory with node-thumbnail!

Not certain how best to handle this; it would be nice to have some local testing for easier access to files when testing, but definitely want AWS set up for the real deal. Maybe there's no real need to make the distinction anyway; might be easy enough just to have it all on AWS, even when testing on local machine.

See the following two links:

http://stackoverflow.com/questions/38211170/sequelize-pagination
https://coderwall.com/p/lkcaag/pagination-you-re-probably-doing-it-wrong

tl;dr from the second post:

* Don't paginate on row offset.
* Find a stable value to paginate on, like an item's date.
* Always deduplicate list items on the client side.
* Never trust your server to know exactly what the user is looking at.

Requires both a filter on the client side and some image processing of that data when the image is downloaded/processed. Not really an MVP feature, but would be really cool to have nonetheless.

A few examples, for basic functionality:

GET /doomimages/ - get all (probably has some hard limit on how many are returned; need to start pagination properly from the beginning though I think)
GET /doomimages/?author=tango&limit=10 - get most recent 10 images from Tango

That ought to cover it. Another relevant, basic param might be date-related (before, after, on). That should cover the basics to start.