Parsing out key:value pairs from text messages and geocoding them

In this third tutorial, we scrape the incoming text messages for pairings of strings in the format “key:value value …”, which we parse out with a regular expression and store in a separate Keyvalue table. This allows us to intelligently search and manipulate the data, as well as to geocode addresses submitted along with data. This yields latitude & longitude data for a given text message.


WHOOZ Tutorial: Parsing out key:value pairs from text messages and geocoding them.

Download the code for this tutorial here: whooz-keyvalue-tutorial.zip (LGPL 3.0)

This builds on the code written in the last tutorial, Batch importing text messages from Twitter in Rails

Manhattan mapped in Cartagen html/javascript with gss


Manhattan mapped in Cartagen html/javascript with gss from Jeffrey Warren on Vimeo.

A preview of the code I’m developing – I’m calling it Cartagen. It’s a web-based vector framework for dynamic cartography. A Ruby server receives map data from OpenStreetMap and from participants’ cell phones in real-time. Data is plotted in native HTML 5 with the canvas element, and styled with a new stylesheet format, GSS: Geographic Stylesheets. I’ll be demoing a more complete system during the MIT Media Lab’s Sponsor Week.

The Diagram Prize

Books that have made the shortlist but inexplicably failed to win include “A Pictorial Book of Tongue Coatings,” “Sex After Death,” “Waterproofing Your Child” and “Cheese Problems Solved” — which, its publisher says, provides “responses to more than 200 of the most commonly asked questions about cheese,” with special emphasis on mozzarella, blue cheese and cheddar.

via The Diagram Prize – Judging a Book by its Title – NYTimes.com.