thumbnail_scraper 0.0.2 → 1.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 85ec2104d8ba31fd892e4128ebd3d997c916a811
+  data.tar.gz: dd3187d661a612bc667306dc3d9bb181feb65905
+SHA512:
+  metadata.gz: 47f9a35c9d88abd11dea525c92c6476ffb97f76de3da38ab29f5b6e8c1dd560bbc84b82e4056ee8df888607c56199265a541a7170f4be8d1255e3b3e8554d747
+  data.tar.gz: 50d9eb695180716d5cd098e016cbd4275c54db2fe9dfc60015ad10d719501c0d75b5c42958209fe294bc73546915120e6697a8633a6c0b37680911f5263d9133

data/.rspec

@@ -0,0 +1 @@
+--colour

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    thumbnail_scraper (0.0.1)
+    thumbnail_scraper (0.0.2)
       fastimage
       httpclient
       nokogiri
@@ -12,9 +12,11 @@ GEM
     addressable (2.2.8)
     crack (0.3.1)
     diff-lcs (1.1.3)
-    fastimage (1.2.13)
-    httpclient (2.2.5)
-    nokogiri (1.5.3)
+    fastimage (1.5.2)
+    httpclient (2.3.4.1)
+    mini_portile (0.5.3)
+    nokogiri (1.6.1)
+      mini_portile (~> 0.5.0)
     rspec (2.11.0)
       rspec-core (~> 2.11.0)
       rspec-expectations (~> 2.11.0)

data/README.md

@@ -10,11 +10,11 @@ require 'thumbnail_scraper'
 
 include ThumbnailScraper
 scraper = ThumbnailScraper.new
-image = scraper.image_to_thumbnail_url("http://www.monibuds.com/")
+image = scraper.image_to_thumbnail_for_url("http://www.example.com/")
 thumbnail_url = image.url
 ```
 
-ThumbnailScraper#image_to_thumbnail_url method returns Image object, which contains its size and url.
+ThumbnailScraper#image_to_thumbnail_for_url method returns Image object, which contains its size and url.
 
 Suggested usage
 ---------------
@@ -25,13 +25,13 @@ We encourage you to use it with delayed_job as jobs queue and dragonfly as image
 require 'thumbnail_scraper'
 
 module Jobs
-  class ScrapThumbnailJob < Struct.new(:page)
+  class ScrapeThumbnailJob < Struct.new(:page)
     def perform
       scraper = ::ThumbnailScraper::ThumbnailScraper.new
       image = scraper.image_to_thumbnail_for_url(page.url)
       page.thumbnail_url = image.url.to_s
       page.save!
-    end 
-  end 
+    end
+  end
 end
 ```

data/lib/thumbnail_scraper.rb

@@ -35,11 +35,7 @@ module ThumbnailScraper
     end
 
     def select_valid_images(images)
-      images.select{|image| image_is_valid?(image)}
-    end
-
-    def image_is_valid?(image)
-      image.width >= 50 && image.height >= 50 && image.width.to_f / image.height.to_f <= 3 && image.height.to_f / image.width.to_f <= 3
+      images.select(&:exists?).select(&:valid?)
     end
   end
 end

data/lib/thumbnail_scraper/image.rb

@@ -9,6 +9,10 @@ module ThumbnailScraper
       self.url = url
     end
 
+    def exists?
+      size ? true : false
+    end
+
     def size
       @size ||= ::FastImage.size(url.to_s)
     end
@@ -24,5 +28,9 @@ module ThumbnailScraper
     def area
       width * height
     end
+
+    def valid?
+      width >= 50 && height >= 50 && width.to_f / height.to_f <= 3 && height.to_f / width.to_f <= 3
+    end
   end
 end

data/lib/thumbnail_scraper/version.rb

@@ -1,3 +1,3 @@
 module ThumbnailScraper
-  VERSION = "0.0.2"
+  VERSION = "1.0.0"
 end

data/lib/thumbnail_scraper/webpage.rb

@@ -59,7 +59,13 @@ module ThumbnailScraper
 
     def attached_images_urls
       elements = document.xpath("//img/@src")
-      elements.map{|element| image_url(element.value)}
+      elements.map do |element|
+        if element.value.start_with?("http://") || element.value.start_with?("https://")
+          image_url(element.value)
+        else
+          image_url("#{url.scheme}://#{url.host}#{element.value.gsub(/^\/?/,'/')}")
+        end
+      end
     end
   end
 end

data/spec/units/thumbnail_scraper_spec.rb

@@ -108,13 +108,13 @@ module ThumbnailScraper
       before :each do
         @image1 = mock("image")
         @image2 = mock("image")
-        @thumbnail_scraper.stub!(:image_is_valid?).with(@image1).and_return(true)
-        @thumbnail_scraper.stub!(:image_is_valid?).with(@image2).and_return(false)
+        @image1.stub!(:exists?).and_return(true)
+        @image2.stub!(:exists?).and_return(true)
+        @image1.stub!(:valid?).and_return(true)
+        @image2.stub!(:valid?).and_return(false)
       end
 
       it "should use #image_is_valid? to validate image" do
-        @thumbnail_scraper.stub!(:image_is_valid?).with(@image1).and_return(true)
-        @thumbnail_scraper.stub!(:image_is_valid?).with(@image2).and_return(false)
         @thumbnail_scraper.select_valid_images([@image1, @image2])
       end
 
@@ -123,9 +123,10 @@ module ThumbnailScraper
       end
     end
 
-    describe "#image_is_valid?" do
+    describe "Image#valid?" do
       before :each do
-        @image = mock("image")
+        @image = Image.new('http://example.com/image.png')
+        @image.stub!(:exists?).and_return(true)
         @image.stub!(:width).and_return(100)
         @image.stub!(:height).and_return(100)
       end
@@ -136,7 +137,7 @@ module ThumbnailScraper
         end
 
         it "should be false" do
-          @thumbnail_scraper.image_is_valid?(@image).should be_false
+          @image.valid?.should be_false
         end
       end
 
@@ -146,7 +147,7 @@ module ThumbnailScraper
         end
 
         it "should be false" do
-          @thumbnail_scraper.image_is_valid?(@image).should be_false
+          @image.valid?.should be_false
         end
       end
 
@@ -157,7 +158,7 @@ module ThumbnailScraper
         end
 
         it "should be false" do
-          @thumbnail_scraper.image_is_valid?(@image).should be_false
+          @image.valid?.should be_false
         end
       end
 
@@ -168,13 +169,13 @@ module ThumbnailScraper
         end
 
         it "should be false" do
-          @thumbnail_scraper.image_is_valid?(@image).should be_false
+          @image.valid?.should be_false
         end
       end
 
       context "otherwise" do
         it "should be true" do
-          @thumbnail_scraper.image_is_valid?(@image).should be_true
+          @image.valid?.should be_true
         end
       end
     end

data/spec/units/webpage_spec.rb

@@ -160,9 +160,9 @@ module ThumbnailScraper
       context "with images in body" do
         before :each do
           @webpage.body = File.read(asset_file_path("images.html"))
-          @webpage.stub!(:image_url).with("/images/subcatalog/smallest.jpg").and_return(:smallest_url)
-          @webpage.stub!(:image_url).with("/images/kitty.jpg").and_return(:kitty_url)
-          @webpage.stub!(:image_url).with("images/biggest.jpg").and_return(:biggest_url)
+          @webpage.stub!(:image_url).with("http://www.example.com/images/subcatalog/smallest.jpg").and_return(:smallest_url)
+          @webpage.stub!(:image_url).with("http://www.example.com/images/kitty.jpg").and_return(:kitty_url)
+          @webpage.stub!(:image_url).with("http://www.example.com/images/biggest.jpg").and_return(:biggest_url)
         end
 
         it "should return all urls of found images" do

data/thumbnail_scraper.gemspec

@@ -3,9 +3,9 @@ require File.expand_path('../lib/thumbnail_scraper/version', __FILE__)
 
 Gem::Specification.new do |gem|
   gem.authors       = ["Jeppe Liisberg", "Jan Filipowski"]
-  gem.email         = ["jachuf@gmail.com"]
-  gem.description   = %q{detect, fetch and generate a thumbnail for any url and store it on s3}
-  gem.summary       = %q{detect, fetch and generate a thumbnail for any url and store it on s3}
+  gem.email         = ["jeppe@liisberg.net"]
+  gem.description   = %q{detect and fetch an image suitable as a thumbnail for any url}
+  gem.summary       = %q{detect and fetch an image suitable as a thumbnail for any url}
   gem.homepage      = ""
 
   gem.add_development_dependency "rspec"

metadata

@@ -1,8 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: thumbnail_scraper
 version: !ruby/object:Gem::Version
-  version: 0.0.2
-  prerelease: 
+  version: 1.0.0
 platform: ruby
 authors:
 - Jeppe Liisberg
@@ -10,96 +9,87 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-08-06 00:00:00.000000000 Z
+date: 2014-05-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
   name: webmock
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
   name: httpclient
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
   name: fastimage
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
   name: nokogiri
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-description: detect, fetch and generate a thumbnail for any url and store it on s3
+description: detect and fetch an image suitable as a thumbnail for any url
 email:
-- jachuf@gmail.com
+- jeppe@liisberg.net
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- .gitignore
+- ".gitignore"
+- ".rspec"
 - Gemfile
 - Gemfile.lock
 - LICENSE
@@ -127,28 +117,27 @@ files:
 - thumbnail_scraper.gemspec
 homepage: ''
 licenses: []
+metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.24
+rubygems_version: 2.2.2
 signing_key: 
-specification_version: 3
-summary: detect, fetch and generate a thumbnail for any url and store it on s3
+specification_version: 4
+summary: detect and fetch an image suitable as a thumbnail for any url
 test_files:
 - spec/sample_pages/images.html
 - spec/sample_pages/images/biggest.jpg