Backend for hackernews.today

Hacker News aggregator's workflow

1. On node server boot up, connects to mongo database
2. HN aggregator module runs on database connected callback
3. Fetches maxItemId from HN API and starts listening on the updates API
4. Find lastProcessedItemId from setting's collection, if not exist consider -1000 from the maxItemId as the start item
5. Get snapshots of all items in between the maxItemId and lastProcessedItemId
6. Filter the items which are just stories
7. Store the stories into the database
8. When done, update the lastProcessedItemId to the maxItemId

APIs

Get top stories for a given start and end time

GET /stories?top=10&startTime=&endTime=

Time and Timezone

• HN data timestamps are in UTC (Unix time)
• HN API endpoint queries support Unix time