craigslister 0.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 2f7c43f5d674ae551b03f42dc93f75576890281a
+  data.tar.gz: ad9657c211afc293e8691c63a27dc2c955761e4e
+SHA512:
+  metadata.gz: 3485dfec92f4f3b9b866fc02d003a656606c6450aef079e4f8e65864a83671168b458ac2a4236affecc67eecdb3bd50a6c43c1b09c0a918f5aff4da513ff4f83
+  data.tar.gz: 34d344fd148d581997485196755c1df024a6dca845152979cf3320addfbf46b096786c805d016b2c7f5eb7242b338750e5b3d19535db14ffdb3404084ca15167

data/lib/craigslister.rb

@@ -0,0 +1,87 @@
+require 'mechanize'
+
+class InvalidRangeError < StandardError
+end
+
+class Craigslister
+  attr_reader :area, :item, :high, :low, :results
+
+  def initialize args
+    @results = []
+    @area = args.fetch(:area, 'sfbay')
+    @item = args[:item]
+    @high = args.fetch(:high, nil)
+    @low  = args.fetch(:low, nil)
+    validate_price_range
+
+    @mech = Mechanize.new
+    configure_mech
+  end
+
+  def scrape
+    links.each do |link|
+      get_data_from(link)
+    end
+  end
+
+  def url
+    "#{base_url}"\
+    "search/sss?sort=rel&"\
+    "#{price_query}"\
+    "query=#{item.downcase.split(' ') * '+'}"\
+  end
+
+  def links
+    @mech.get(url)
+    @mech.page.search('.hdrlnk').map {|link| link['href']}
+  end
+
+
+  private
+    def base_url
+      "https://#{area}.craigslist.org/"
+    end
+
+    def get_data_from link
+      @mech.get(link)
+      @results << Item.new(scrape_item_data) rescue p 'no image'
+    end
+
+    def scrape_item_data
+      {
+        image: @mech.page.images[0].src,
+        title: @mech.page.at('span.postingtitletext').text.gsub(/ ?- ?\$\d+ ?\(.+\)/, ''),
+        price: @mech.page.at('span.postingtitletext span.price').text.gsub(/\$/,'').to_i,
+        location: @mech.page.at('span.postingtitletext small').text.gsub(/ ?[\(\)]/,''),
+        description: @mech.page.at('section#postingbody').text
+      }
+    end
+
+    def price_query
+      result = ''
+      result += "min_price=#{low}&" if low
+      result += "max_price=#{high}&" if high
+      result
+    end
+
+    def configure_mech
+      @mech.robots = false
+      @mech.user_agent_alias = 'Mac Safari'
+    end
+
+    def validate_price_range
+      raise InvalidRangeError if (low && high) && low > high
+    end
+end
+
+class Item
+  attr_reader :title, :image, :price, :location
+
+  def initialize args
+    @title = args[:title]
+    @image = args[:image]
+    @price = args[:price]
+    @location = args[:location]
+  end
+end
+

metadata

@@ -0,0 +1,44 @@
+--- !ruby/object:Gem::Specification
+name: craigslister
+version: !ruby/object:Gem::Version
+  version: 0.0.0
+platform: ruby
+authors:
+- Chris Scott
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2015-10-26 00:00:00.000000000 Z
+dependencies: []
+description: all you need is an item title and you can scrape item objects from craigslist
+email: christo247@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/craigslister.rb
+homepage: http://rubygems.org/gems/craigslister
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.4.5
+signing_key: 
+specification_version: 4
+summary: Scrape Craigslist!!!!! heh heh heh
+test_files: []