You can clone with
HTTPS or Subversion.
This is a smaller story broken out from #562.
This issue is just to make /search.json work like the twitter API. This route doesn't currently exist at all, but there is a nice update json decorator for you to use.
Ignore the issue of client authentication for the moment; just make this work as a normally authenticated user in the browser for now (see #564 about authentication)
Description: Returns relevant tweets that match a specified query.
Definitely interpret anything in the twitter docs so that it makes sense for rstat.us, and definitely ask if you have any questions!
Leave a comment if you're working on this.
Things in the Twitter docs that seem to "make sense" for rstat.us as it stands now:
q: A UTF-8, URL-encoded search query of 1,000 characters maximum, including operators. Queries may additionally be limited by complexity.
callback - maybe? Only available for JSON format (which we'd be using). If supplied, the response will use the JSONP format with a callback of the given name.
page: optional The page number (starting at 1) to return, up to a max of roughly 1500 results (based on rpp * page).
show_user: optional When true, prepends ":" to the beginning of the tweet. This is useful for readers that do not display Atom's author field. The default is false.
until: optional. Returns tweets generated before the given date. Date should be formatted as YYYY-MM-DD. Example Value: 2010-03-28
The above is what made sense to me based off of the current way rstat.us is built.
@jrgifford you may have seen it, but Carol created a twitter-api branch earlier today. Probably worth working against that if you're not already. :-)