Skip to content

Latest commit

 

History

History
39 lines (31 loc) · 1.84 KB

README.md

File metadata and controls

39 lines (31 loc) · 1.84 KB

osm_to_poi

Extracts Points of Interest (POI) as an array of hashes from an OSM XML file This library uses libxml, which can be installed as a gem

   (sudo) gem install libxml-ruby 

More details on LibXml Ruby may be found at http://xml4r.github.com/libxml-ruby/ We use the XML stream parsing method discussed nicely in http://coldattic.info/shvedsky/s/blogs/a-foo-walks-into-a-bar/posts/56

How to use

Currently there is no gem file. So, clone the repository or download a tarball. The example 'examples/aizu_to_poi.rb' demonstrates the usage of this library to extract the POIs of interest (amenity, shop, leisure and tourism) from an OSM XML file extracted over Aizuwakamatsu from openstreetmap.org.

In this example, we start off by streaming through the XML elements one by one. Once an element by the name 'node' is reached, we check if it is processable by POIFactory. If so, it is dispatched to POIFactory for further processing. If not, we skip to the next element and so on. The POIFactory extracts POIs that we specified and stores all attributes in a hash. Once the whole XML file is processed, we retrieve an array of hashes from the POIFactory instance. In the example, this is further processed to write a CSV file with only the attributes of interest.

What does it contain?

  1. It adds some helpers to libxml-ruby (see lib/ext/xml.rb).
  2. It adds a POIFactory Class that has two instance methods, new and <<(node) and a class method processable?(node)
    1. We can pass a list of POI keys that we are interested in while initializing POI factory
    2. The <<(node) method passes in an XML node to the POI factory instance for further processing. The results are stored in memory till needed.
    3. The class method processable?(node) simply checks if the 'node' element has children.