chicago restaurant health inspections

chicago-inspections.rb

require 'net/http'
require 'uri'
require 'nokogiri'

url = URI.parse('http://webapps.cityofchicago.org/healthinspection/inspectionresultrow.jsp')

request = Net::HTTP::Post.new(url.path)

request.set_form_data({"REST"=>" ", "STR_NBR"=>"", "STR_NBR2"=>"", "STR_DIRECTION"=>"", "STR_NM"=>"", "ZIP"=>""})

response = Net::HTTP.new(url.host, url.port).start {|http| http.request(request)}

doc = Nokogiri::HTML.parse(response.body)

doc.search('#results tr').each do |tr|
  info = tr.text.each_line.map(&:strip).delete_if {|l| l.strip == ""}.join('\n')
  ScraperWiki.save(unique_keys=['text'], data = {'text' => info})
end

this line alone

info = tr.text.each_line.map(&:strip).delete_if {|l| l.strip == ""}.join('\n')

took me a number of iterations and many lines of code to slap into a solid c++ object. I think the new c++ std lib tools are getting higher strength string parsers, but haven't played with them yet.

pure functional transforms rule!