scrappy 0.1.8 → 0.1.9

data/History.txt

@@ -1,3 +1,8 @@
+=== 0.1.9 2010-12-16
+
+* Corrected bug when filtering URI selectors
+* Added sc:suffix to uri builder
+
 === 0.1.8 2010-12-14
 
 * Added sc:sameas

data/lib/scrappy.rb

@@ -19,7 +19,7 @@ require 'scrappy/agent/agent'
 Namespace :sc, 'http://lab.gsi.dit.upm.es/scraping.rdf#'
 
 module Scrappy
-  VERSION = '0.1.8'
+  VERSION = '0.1.9'
 end
 
 # Require selectors

data/lib/scrappy/agent/extractor.rb

@@ -6,7 +6,7 @@ module Scrappy
       triples = []
       content = Nokogiri::HTML(html, nil, 'utf-8')
 
-      uri_selectors  = kb.find(nil, Node('rdf:type'), Node('sc:UriSelector')) + kb.find(nil, Node('rdf:type'), Node('sc:UriPatternSelector')).flatten.select do |uri_selector|
+      uri_selectors  = (kb.find(nil, Node('rdf:type'), Node('sc:UriSelector')) + kb.find(nil, Node('rdf:type'), Node('sc:UriPatternSelector'))).flatten.select do |uri_selector|
         class_name = uri_selector.rdf::type.first.to_s.split('#').last
         results = Kernel.const_get(class_name).filter uri_selector, {:content=>content, :uri=>uri}
         !results.empty?

data/lib/scrappy/selectors/new_uri.rb

@@ -8,7 +8,7 @@ module NewUriSelector
     end
 
     contents.map do |content|
-      new_uri = selector.sc::prefix.to_s + content.wikify
+      new_uri = selector.sc::prefix.first.to_s + content.wikify + selector.sc::suffix.first.to_s
       { :uri=>new_uri, :content=>doc[:content], :value=>new_uri }
     end
   end

data/scrappy.gemspec

@@ -2,11 +2,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{scrappy}
-  s.version = "0.1.8"
+  s.version = "0.1.9"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 1.2") if s.respond_to? :required_rubygems_version=
   s.authors = ["Jose Ignacio"]
-  s.date = %q{2010-12-14}
+  s.date = %q{2010-12-16}
   s.default_executable = %q{scrappy}
   s.description = %q{RDF web scraper}
   s.email = %q{joseignacio.fernandez@gmail.com}

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 1
-  - 8
-  version: 0.1.8
+  - 9
+  version: 0.1.9
 platform: ruby
 authors: 
 - Jose Ignacio
@@ -14,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-12-14 00:00:00 +01:00
+date: 2010-12-16 00:00:00 +01:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency