I’ve been working with the Twitter Streaming API once more, not on the Twitter Stats project I was working on a few months back but a new project. This one uses the Twitter Streaming API to gather real-time information about particular hashtags, words and any other data you can think of.
There really isn’t that much documentation for how to get started with the Twitter Streaming API so I ended up spending a lot of time writing trial and error code. Once I managed to figure it out and gain a basic understanding of how to consume data and keep the connections open I found that it was pretty much the same as using the REST API. Soon after I discovered Phirehose which is a really handy interface designed to help connect to the Twitter Streaming API, using this already written interface allowed me to focus on the next part of the project rather than creating the interface myself. I think that once I have a basic version of the project finished I will revisit this and re-write the interface and tailor it for my needs.
Using the filter-track.php class that comes with Phirehose I’ve been able to consume from the API and pull all sorts of data. I was surprised how quickly the data is pulled from Twitter, it appears to be pretty much instant. Now that I have the data being consumed the next step is to think about storing it in a database and then having a front end display it in an interesting format.