Crawling actual map *names*, but not yet pulling in maps

cat-town
Dan Buch 13 years ago
parent 6d9ae87557
commit 607a4158e1

@ -2,3 +2,4 @@ log/*.log
tmp/* tmp/*
db/*.sqlite3 db/*.sqlite3
config/redis.conf config/redis.conf
db/*.sqlite3-journal

@ -1,2 +1,5 @@
class Map < ActiveRecord::Base class Map < ActiveRecord::Base
def self.from_city_name(city_name)
self.find_or_initialize_by_name(city_name).save!
end
end end

@ -3,3 +3,6 @@ resque_web:
redis: redis:
port: 16379 port: 16379
map:
base_url: 'http://en.wikipedia.org/wiki/List_of_towns_and_cities_with_100,000_or_more_inhabitants/cityname:_{FIRST_LETTER}'

@ -0,0 +1,49 @@
require 'logger'
require 'nokogiri'
require 'typhoeus'
class MapCrawler
attr_accessor :base_map_url, :log, :request_pool
def initialize(base_map_url)
@base_map_url = base_map_url
@log = Logger.new(
File.expand_path('../log/map-crawler.log', File.dirname(__FILE__))
)
@log.level = Logger::INFO
@log.formatter = lambda do |severity, time, prog, message|
"#{time} - #{severity} - #{message}\n"
end
end
def crawl(city_name_callback = nil)
city_name_callback ||= lambda { |n| puts n }
hydra = Typhoeus::Hydra.new(:initial_pool_size => 26)
('A'..'Z').each do |letter|
letter_request = Typhoeus::Request.new(
@base_map_url.gsub(/\{FIRST_LETTER\}/, letter)
)
letter_request.on_complete do |response|
handle_cities(response, city_name_callback)
end
hydra.queue(letter_request)
end
hydra.run
end
def handle_cities(response, city_name_callback)
@log.info("Handling cities at url #{response.effective_url}")
doc = Nokogiri::HTML(response.body)
doc.css('div.mw-content-ltr ul')[3].css('li a').each do |anchor|
@log.info("Found city: #{anchor.text}")
city_name_callback.call(anchor.text.strip)
end
end
end

@ -0,0 +1,6 @@
namespace :maps do
desc 'Index the maps!'
task :index => :environment do
MapCrawler.new(Setting.map(:base_url)).crawl(Map.method(:from_city_name))
end
end

@ -0,0 +1,16 @@
require 'spec_helper'
describe MapCrawler do
let(:subject) { MapCrawler.new(Setting.map(:base_url)) }
describe 'when crawling for actual maps', :integration => true do
before(:each) do
Map.destroy_all
end
it 'should create Map entries for each map found' do
expect { subject.crawl }.to change{ Map.count }.by_at_least(1)
end
end
end
Loading…
Cancel
Save