link_thumbnailer 3.2.0 → 3.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 9799bede3f23e3369ddafa4ddeb36c9436d7ee5d
-  data.tar.gz: 484c6c008967be9cd3d813bac81dffde02c18bd7
+  metadata.gz: f99697da11cef790d10eddde8f79ad1dd4a2e1e6
+  data.tar.gz: c113e7b8c50dda3e4ead33dde327b0beec15435d
 SHA512:
-  metadata.gz: 3c6a25faef22554a54a84a2ffbf0ee85babb1b088b5ed79d08e6550504f9e9853fa177cd730856cd1c3cb3d285dc1bf6a727cd89a85d410f00cf89023668d1de
-  data.tar.gz: 10216dc920ce54195179337fc4034ba14413c833369a2e566e2b7aa5d118a5c7599df4673564ff7a57bc3558be3fa8d826e0c8f35058aba64e9544a13bc1d09d
+  metadata.gz: f4f15babf9656fa90eae2cd10c9634011e766f02319e8d80ed03bb47149e30770420c6e17b3d4b91488022b7df4c1ed1cbc2f41b7d5b850f55d514708aed74fa
+  data.tar.gz: 1572e848bf83ae6eba3bef808474b2ca466b13d8b6ce467c73f0768cd0f57247083400a9a05823cd247a45fe5765438ae242bb2731804c2a6b2e64391896c554

data/.travis.yml

@@ -1,6 +1,4 @@
 bundler_args: --without development
 language: ruby
 rvm:
-  - 1.9.3
-  - 2.0.0
-  - 2.1.0
+  - 2.2.0

data/CHANGELOG.md

@@ -1,3 +1,10 @@
+# 3.2.1
+
+- Fixes #88
+- Override User-Agent header properly
+- Match xpath nodes if attribute content is present
+- Avoid nil urls in image parser
+
 # 3.2.0
 
 Makes scrapers configurable by allowing to set the scraping strategy:

data/lib/link_thumbnailer/image_parser.rb

@@ -6,7 +6,7 @@ module LinkThumbnailer
     attr_reader :images
 
     def initialize(urls)
-      @images = perform? ? ::ImageInfo.from(urls, max_concurrency: max_concurrency) : Array(urls).map(&method(:build_default_image))
+      @images = perform? ? ::ImageInfo.from(urls, max_concurrency: max_concurrency) : Array(urls).compact.map(&method(:build_default_image))
     end
 
     def size

data/lib/link_thumbnailer/processor.rb

@@ -39,7 +39,7 @@ module LinkThumbnailer
 
     def set_http_headers(headers = {})
       headers.each { |k, v| http.headers[k] = v }
-      http.headers['User-Agent']               = user_agent
+      http.override_headers['User-Agent']      = user_agent
       http.override_headers['Accept-Encoding'] = 'none'
     end
 

data/lib/link_thumbnailer/scrapers/base.rb

@@ -47,7 +47,7 @@ module LinkThumbnailer
         value     = options.fetch(:value, :content)
         attribute = options.fetch(:attribute, attribute_name)
 
-        document.xpath("//meta[translate(@#{key},'#{abc.upcase}','#{abc}') = '#{attribute}' and @#{value}]")
+        document.xpath("//meta[translate(@#{key},'#{abc.upcase}','#{abc}') = '#{attribute}' and string-length(@#{value}) > 0]")
       end
 
       def abc

data/lib/link_thumbnailer/scrapers/opengraph/image.rb

@@ -1,4 +1,5 @@
 require 'link_thumbnailer/scrapers/opengraph/base'
+require 'link_thumbnailer/uri'
 
 module LinkThumbnailer
   module Scrapers
@@ -20,7 +21,10 @@ module LinkThumbnailer
           end
 
           def model
-            nodes.map { |n| modelize(n, n.attributes['content'].to_s) }
+            nodes.map do |n|
+              uri =  LinkThumbnailer::URI.new(n.attributes['content'])
+              modelize(n, uri.to_s) if uri.valid?
+            end.compact
           end
 
           def modelize(node, text = nil)

data/lib/link_thumbnailer/uri.rb

@@ -0,0 +1,18 @@
+module LinkThumbnailer
+  class URI
+
+    attr_reader :attribute
+
+    def initialize(uri)
+      @attribute = uri.to_s
+    end
+
+    def valid?
+      !!(attribute =~ ::URI::regexp)
+    end
+
+    def to_s
+      attribute
+    end
+  end
+end

data/lib/link_thumbnailer/version.rb

@@ -1,3 +1,3 @@
 module LinkThumbnailer
-  VERSION = '3.2.0'
+  VERSION = '3.2.1'
 end

data/spec/processor_spec.rb

@@ -106,7 +106,7 @@ describe LinkThumbnailer::Processor do
   describe '#set_http_headers' do
 
     let(:user_agent)  { 'foo' }
-    let(:headers)     { instance.send(:http).headers }
+    let(:headers)     { instance.send(:http).override_headers }
     let(:action)      { instance.send(:set_http_headers) }
 
     before do

data/spec/uri_spec.rb

@@ -0,0 +1,42 @@
+require 'spec_helper'
+
+describe LinkThumbnailer::URI do
+
+  let(:uri)       { 'http://foo.com' }
+  let(:instance)  { described_class.new(uri) }
+
+  describe '#valid?' do
+
+    let(:action) { instance.send(:valid?) }
+
+    context 'when bad format' do
+
+      before do
+        instance.stub(:attribute).and_return("/invalid/path")
+      end
+
+      it { expect(action).to be_falsey }
+
+    end
+
+    context 'when valid format' do
+
+      before do
+        instance.stub(:attribute).and_return("http://foo.com")
+      end
+
+      it { expect(action).to be_truthy }
+
+    end
+
+  end
+
+  describe '#to_s' do
+
+    let(:action) { instance.to_s }
+
+    it { expect(action).to eq(uri) }
+
+  end
+
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: link_thumbnailer
 version: !ruby/object:Gem::Version
-  version: 3.2.0
+  version: 3.2.1
 platform: ruby
 authors:
 - Pierre-Louis Gottfrois
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-05-15 00:00:00.000000000 Z
+date: 2016-06-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -172,6 +172,7 @@ files:
 - lib/link_thumbnailer/scrapers/opengraph/title.rb
 - lib/link_thumbnailer/scrapers/opengraph/video.rb
 - lib/link_thumbnailer/scrapers/opengraph/videos.rb
+- lib/link_thumbnailer/uri.rb
 - lib/link_thumbnailer/version.rb
 - lib/link_thumbnailer/video_parser.rb
 - link_thumbnailer.gemspec
@@ -206,6 +207,7 @@ files:
 - spec/scrapers/base_spec.rb
 - spec/scrapers/opengraph/base_spec.rb
 - spec/spec_helper.rb
+- spec/uri_spec.rb
 - spec/video_parser_spec.rb
 homepage: https://github.com/gottfrois/link_thumbnailer
 licenses: []
@@ -263,4 +265,5 @@ test_files:
 - spec/scrapers/base_spec.rb
 - spec/scrapers/opengraph/base_spec.rb
 - spec/spec_helper.rb
+- spec/uri_spec.rb
 - spec/video_parser_spec.rb