scrapify 0.0.7 → 0.0.8

data/lib/scrapify/base.rb

@@ -69,7 +69,9 @@ module Scrapify
       end
 
       def parse_html
-        Nokogiri::HTML(html_content)
+        doc = Nokogiri::HTML(html_content)
+        doc.css('br').each {|br| br.replace("\n")}
+        doc
       end
 
       def html_content
@@ -106,6 +108,15 @@ module Scrapify
           attributes = Hash[attribute_names.map {|attribute| [attribute, send("#{attribute}_values")[index]]}]
           self.new(attributes)
         end
+
+        define_singleton_method :where do |conditions = {}|
+          raise Scrapify::AttributeDoesNotExist.new(conditions.keys - attribute_names) unless conditions.keys.all?{|key| attribute_names.include?(key) }
+          indices = conditions.collect do |attribute, value|
+            send("#{attribute}_values").each_with_index.find_all{|attr_val, index| attr_val == value}.collect(&:last)
+          end
+          common_indices = indices.reduce {|a, b| a & b}
+          common_indices.collect{|index| find_by_index(index)}
+        end
       end
 
       def define_count(key_attribute)

data/lib/scrapify/exceptions.rb

@@ -0,0 +1,3 @@
+module Scrapify
+  class AttributeDoesNotExist < StandardError; end
+end

data/lib/scrapify/version.rb

@@ -1,3 +1,3 @@
 module Scrapify
-  VERSION = "0.0.7"
+  VERSION = "0.0.8"
 end

data/lib/scrapify.rb

@@ -6,5 +6,6 @@ require 'uri'
 require 'net/http'
 require 'scrapify/base'
 require 'scrapify/scraper'
+require 'scrapify/exceptions'
 require 'json'
 require 'jsonify'

data/spec/shared/finder.rb

@@ -0,0 +1,10 @@
+shared_examples_for '#finder' do |klass_or_object, conditions|
+  it 'should fetch objects based on conditions' do
+    pizza = klass_or_object.where(conditions).first
+    pizza.name.should == 'chicken golden delight'
+    pizza.image_url.should == 'golden.jpg'
+    pizza.price.should == '4.56'
+    pizza.ingredients.should be_empty
+    pizza.ingredient_urls.should == ['chicken.html', 'delight.html']
+  end
+end

data/spec/shared/scrapify.rb

@@ -19,7 +19,7 @@ shared_examples_for "Scrapify" do |klass_or_object|
           <span class='references'><ol><li></li></ol></span
         </li>
         <li>
-          <a>veg supreme</a><input value="veg.jpg">
+          <a>veg<br/>supreme</a><input value="veg.jpg">
           <span class='price'>(2.34)</span>
           <span class='ingredients'>
             <ol>
@@ -62,7 +62,7 @@ shared_examples_for "Scrapify" do |klass_or_object|
     end
 
     it "should parse html and fetch attributes using css" do
-      klass_or_object.name_values.should == ['chicken supreme', 'veg supreme', 'pepperoni', 'chicken golden delight']
+      klass_or_object.name_values.should == ['chicken supreme', "veg\nsupreme", 'pepperoni', 'chicken golden delight']
     end
 
     it "should parse html and fetch attributes using xpath" do
@@ -85,6 +85,10 @@ shared_examples_for "Scrapify" do |klass_or_object|
       klass_or_object.first.name.should == 'chicken supreme'
     end
 
+    it "should replace br with newline" do
+      klass_or_object.all[1].name.should == "veg\nsupreme"
+    end
+
     describe "cache headers" do
       it "should return the http headers" do
         klass_or_object.http_cache_header.should == {
@@ -120,7 +124,7 @@ shared_examples_for "Scrapify" do |klass_or_object|
   describe "last" do
     it "should fetch last matching element" do
       last_pizza = klass_or_object.last
-      last_pizza.name.should == 'chicken golden delight'
+      last_pizza.name.should == "chicken golden delight"
       last_pizza.image_url.should == 'golden.jpg'
     end
   end
@@ -129,7 +133,7 @@ shared_examples_for "Scrapify" do |klass_or_object|
     it "should fetch all objects" do
       pizzas = klass_or_object.all
       pizzas.size.should == 4
-      pizzas.map(&:name).should == ['chicken supreme', 'veg supreme', 'pepperoni', 'chicken golden delight']
+      pizzas.map(&:name).should == ['chicken supreme', "veg\nsupreme", 'pepperoni', 'chicken golden delight']
       pizzas.map(&:image_url).should == ['chicken.jpg', 'veg.jpg', 'pepperoni.jpg', 'golden.jpg']
     end
   end
@@ -169,10 +173,26 @@ shared_examples_for "Scrapify" do |klass_or_object|
       pizzas = klass_or_object.all
       pizzas.to_json.should == [
         {name: "chicken supreme", image_url: "chicken.jpg", price: '1.23', ingredients: ['corn', 'tomato'], :ingredient_urls => []},
-        {name: "veg supreme", image_url: "veg.jpg", price: '2.34', ingredients: ['mushroom', 'jalapeno'], :ingredient_urls => []},
+        {name: "veg\nsupreme", image_url: "veg.jpg", price: '2.34', ingredients: ['mushroom', 'jalapeno'], :ingredient_urls => []},
         {name: "pepperoni", image_url: "pepperoni.jpg", price: '3.45', ingredients: [], :ingredient_urls => []},
         {name: "chicken golden delight", image_url: "golden.jpg", price: '4.56', ingredients: [], :ingredient_urls => ['chicken.html', 'delight.html']},
       ].to_json
     end
   end
+
+  it_behaves_like '#finder', klass_or_object, :name => 'chicken golden delight'
+  it_behaves_like '#finder', klass_or_object, :image_url => 'golden.jpg'
+  it_behaves_like '#finder', klass_or_object, :price => '4.56'
+  it_behaves_like '#finder', klass_or_object, :ingredient_urls => ['chicken.html', 'delight.html']
+  it_behaves_like '#finder', klass_or_object, :name => 'chicken golden delight', :image_url => 'golden.jpg', :price => '4.56', :ingredient_urls => ['chicken.html', 'delight.html']
+
+  it 'should return empty array if object is not found' do
+    klass_or_object.where(:name => 'does not exist').should be_empty
+  end
+
+  it 'should throw exception if attribute is not defined' do
+    lambda {
+      klass_or_object.where(:some_attribute => 'chicken golden delight')
+    }.should raise_error(Scrapify::AttributeDoesNotExist)
+  end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: scrapify
 version: !ruby/object:Gem::Version
-  version: 0.0.7
+  version: 0.0.8
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-06-28 00:00:00.000000000Z
+date: 2012-07-11 00:00:00.000000000Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &70233941360460 !ruby/object:Gem::Requirement
+  requirement: &70278707139780 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -21,10 +21,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70233941360460
+  version_requirements: *70278707139780
 - !ruby/object:Gem::Dependency
   name: mocha
-  requirement: &70233941359980 !ruby/object:Gem::Requirement
+  requirement: &70278707139300 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -32,10 +32,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70233941359980
+  version_requirements: *70278707139300
 - !ruby/object:Gem::Dependency
   name: fakeweb
-  requirement: &70233941359460 !ruby/object:Gem::Requirement
+  requirement: &70278707138800 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -43,10 +43,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70233941359460
+  version_requirements: *70278707138800
 - !ruby/object:Gem::Dependency
   name: nokogiri
-  requirement: &70233941358800 !ruby/object:Gem::Requirement
+  requirement: &70278707138260 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -54,10 +54,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70233941358800
+  version_requirements: *70278707138260
 - !ruby/object:Gem::Dependency
   name: activesupport
-  requirement: &70233941358220 !ruby/object:Gem::Requirement
+  requirement: &70278707137740 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -65,10 +65,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70233941358220
+  version_requirements: *70278707137740
 - !ruby/object:Gem::Dependency
   name: json
-  requirement: &70233941351380 !ruby/object:Gem::Requirement
+  requirement: &70278707137260 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -76,7 +76,7 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70233941351380
+  version_requirements: *70278707137260
 description: ScrApify scraps static html sites to RESTlike APIs
 email:
 - sathish316@gmail.com
@@ -92,6 +92,7 @@ files:
 - lib/jsonify.rb
 - lib/scrapify.rb
 - lib/scrapify/base.rb
+- lib/scrapify/exceptions.rb
 - lib/scrapify/scraper.rb
 - lib/scrapify/version.rb
 - scrapify.gemspec
@@ -99,6 +100,7 @@ files:
 - spec/pizza.rb
 - spec/scraper_spec.rb
 - spec/scrapify_spec.rb
+- spec/shared/finder.rb
 - spec/shared/scrapify.rb
 - spec/spec_helper.rb
 - spec/test_models.rb
@@ -131,6 +133,7 @@ test_files:
 - spec/pizza.rb
 - spec/scraper_spec.rb
 - spec/scrapify_spec.rb
+- spec/shared/finder.rb
 - spec/shared/scrapify.rb
 - spec/spec_helper.rb
 - spec/test_models.rb