This potentially replaces PRAW for some forms of Reddit bot!
Conversation
Replying to
Yep! It's intended to be a compliment to PRAW. PRAW is a great module, so hopefully this adds a lot of functionality to it.
1
Replying to
Do you plan to add an offset/page function, or should we rely using "before" with created_utc if we want more pages of results?
1
Replying to
Using the before / after parameters works best for the time being. Do you think adding pagination to the API would be helpful?
2
Replying to
BigQuery would probably be more sane and performant (just my guess - not sure though!)
2
Replying to
BigQuery does work better in some use cases, although I have worked hard to get the response times down for the API. For most simple
1
1
searches, the response time is usually under 100ms. I agree with you that the pagination is worth pursuing so I'll add it to the list of
1
features to be added to the API. Please feel free to recommend more using Github if you like. I do appreciate your feedback!
1
1
Replying to
Thanks! Here's another example of aggs being super cool. You can easily see the number of comments per hour over the past seven days ...
1
related to the Hurricane. Here is an example:
api.pushshift.io/reddit/search/
That will show, by hour, the total comments mentioning Houston.
Replying to
Uh so the link_id aggregation feature should be bundled up as a social media monitoring tool for brands... they'd pay for that.
1
1
Replying to
Yeah, I am finding that aggregation to be very powerful. There are a few different variations of it via elasticsearch but it does work ...
1
Show replies