optionally decode html entities (defaults to true) to make behavior r…

…e html entities a bit more consistent

Rakefile

@@ -24,6 +24,7 @@ begin
     s.description = "Generic html element locators for integration testing"
     s.authors = ['Sven Fuchs']
     s.files =  FileList["[A-Z]*", "{lib,test,vendor}/**/*"]
+    s.add_dependency 'htmlentities'
   end
 rescue LoadError
   puts "Jeweler, or one of its dependencies, is not available. Install it with: sudo gem install technicalpickles-jeweler -s http://gems.github.com"

lib/locator.rb

@@ -1,12 +1,15 @@
 require 'core_ext/string/underscore'
 
 module Locator
-  autoload :Boolean, 'locator/boolean'
-  autoload :Dom,     'locator/dom'
-  autoload :Element, 'locator/element'
-  autoload :Matcher, 'locator/matcher'
-  autoload :Result,  'locator/result'
-  autoload :Xpath,   'locator/xpath'
+  autoload :Boolean,  'locator/boolean'
+  autoload :Dom,      'locator/dom'
+  autoload :Decoding, 'locator/decoding'
+  autoload :Element,  'locator/element'
+  autoload :Matcher,  'locator/matcher'
+  autoload :Result,   'locator/result'
+  autoload :Xpath,    'locator/xpath'
+
+  extend Decoding
 
   class ElementNotFound < StandardError
     def initialize(*args)

lib/locator/decoding.rb

@@ -0,0 +1,28 @@
+module Locator
+  module Decoding
+    def decode_entities?
+      @decode_entities ||= true
+    end
+
+    def decode_entities=(decode_entities)
+      @decode_entities = decode_entities
+    end
+
+    def decode_attributes(hash)
+      Hash[*hash.map { |name, value| [name, decode(value)] }.flatten]
+    end
+
+    def decode(value)
+      value.is_a?(String) ? html_entities.decode(value) : value
+    end
+
+    protected
+
+      def html_entities
+        html_entities ||= begin
+          require 'htmlentities'
+          HTMLEntities.new
+        end
+      end
+  end
+end

lib/locator/dom/nokogiri/element.rb

@@ -25,10 +25,6 @@ def css_path
           element.css_path.to_s
         end
 
-        def content
-          element.content
-        end
-
         def value
           case name
           when 'input'
@@ -45,6 +41,10 @@ def checked
           checked ? checked.value : nil
         end
 
+        def content
+          element.content
+        end
+
         def inner_html
           element.inner_html
         end

lib/locator/element.rb

@@ -52,6 +52,8 @@ def xpath(*args)
     protected
 
       def lookup(scope, selector, attributes = {})
+        attributes = Locator.decode_attributes(attributes) if Locator.decode_entities?
+
         scope = scope.respond_to?(:elements_by_xpath) ? scope : Locator::Dom.page(scope)
         xpath = xpath(attributes)
         xpath = ".#{xpath}" unless xpath[0, 1] == '.'

lib/locator/result.rb

@@ -8,6 +8,7 @@ class Result < Array
 
     class << self
       def matches?(name, value, selector)
+        value, selector = Locator.decode(value), Locator.decode(selector) if Locator.decode_entities?
         value = normalize_whitespace(value)
         case selector
         when Regexp

test/locator/dom/nokogiri_test.rb

@@ -13,4 +13,30 @@ class NokogiriTest < Test::Unit::TestCase
     dom = Locator::Dom::Nokogiri::Page.new('<a href="#">the link</a>')
     assert Element.new.locate(dom, 'the link')
   end
+
+  # test "Nokogiri and Umlauts" do
+  #   encoded = '<p>&ouml;</p>'
+  #   unencoded = '<p>ö</p>'
+  #
+  #   # this is ok
+  #   fragment = Nokogiri::HTML.fragment(encoded)
+  #   assert_equal encoded, fragment.to_s
+  #
+  #   # this seems weird ... wtf experiences ahead
+  #   fragment = Nokogiri::HTML.fragment(unencoded)
+  #   assert_equal encoded, fragment.to_s
+  #
+  #   options =
+  #     Nokogiri::XML::ParseOptions::RECOVER |
+  #     Nokogiri::XML::ParseOptions::NOERROR |
+  #     Nokogiri::XML::ParseOptions::NOWARNING
+  #
+  #   # this is ok
+  #   node = Nokogiri::HTML.parse(encoded, nil, nil, options)
+  #   assert_equal encoded, node.xpath('//p').first.to_s
+  #
+  #   # this is ok
+  #   node = Nokogiri::HTML.parse(unencoded, nil, 'utf-8', options)
+  #   assert_equal unencoded, node.xpath('//p').first.to_s
+  # end
 end

test/locator_test.rb

@@ -33,16 +33,6 @@ class LocatorTest < Test::Unit::TestCase
     assert_equal 'foo', element.attribute('id')
   end
 
-  test "locates an element by node name and selector (containing an umlaut)" do
-    html = '<span>Berlin</span><span>München</span>'
-    assert_equal 'München', locate(html, 'München').content
-  end
-
-  test "locates an element by node name and attribute (containing an umlaut)" do
-    html = '<input type="text" value="München">'
-    assert_equal html, locate(html, :value => 'München').to_s
-  end
-
   # within
 
   test "within scopes to a locator" do
@@ -97,4 +87,46 @@ class LocatorTest < Test::Unit::TestCase
     html = '<form></form><div><form><p id="foo"><p></form></div>'
     assert_nil locate(html, :div) { locate(html, :form, :class => 'bar')  { locate(html, :p) } }
   end
+
+  # locate with umlauts
+
+  test "locates an element by encoded selector from html containing an encoded umlaut" do
+    html = '<span>Berlin</span><span>M&uuml;nchen</span>'
+    assert_equal 'München', locate(html, 'M&uuml;nchen').content
+  end
+
+  test "locates an element by encoded selector from html containing an non-encoded umlaut" do
+    html = '<span>Berlin</span><span>München</span>'
+    assert_equal 'München', locate(html, 'M&uuml;nchen').content
+  end
+
+  test "locates an element by non-encoded selector from html containing an encoded umlaut" do
+    html = '<span>Berlin</span><span>M&uuml;nchen</span>'
+    assert_equal 'München', locate(html, 'München').content
+  end
+
+  test "locates an element by non-encoded selector from html containing a non-encoded umlaut" do
+    html = '<span>Berlin</span><span>München</span>'
+    assert_equal 'München', locate(html, 'München').content
+  end
+
+  test "locates an element by encoded attribute from html containing an encoded umlaut" do
+    html = '<input type="text" value="M&uuml;nchen">'
+    assert_equal 'München', locate(html, :value => 'M&uuml;nchen').value
+  end
+
+  test "locates an element by encoded attribute from html containing an non-encoded umlaut" do
+    html = '<input type="text" value="München">'
+    assert_equal 'München', locate(html, :value => 'M&uuml;nchen').value
+  end
+
+  test "locates an element by non-encoded attribute from html containing an encoded umlaut" do
+    html = '<input type="text" value="M&uuml;nchen">'
+    assert_equal 'München', locate(html, :value => 'München').value
+  end
+
+  test "locates an element by non-encoded attribute from html containing a non-encoded umlaut" do
+    html = '<input type="text" value="München">'
+    assert_equal 'München', locate(html, :value => 'München').value
+  end
 end

test/webrat.rb

@@ -0,0 +1,27 @@
+# encoding: utf-8
+require File.expand_path('../test_helper', __FILE__)
+require 'action_pack'
+require 'webrat'
+require 'rack/test'
+
+Webrat.configure do |config|
+  config.mode = :rack
+end
+
+class WebratBehaviorTest < Test::Unit::TestCase
+  include Webrat::Methods
+
+  def last_response
+    @last_response ||= Rack::Response.new
+  end
+
+  def with_html(html)
+    last_response.body = html
+  end
+
+  test "foo" do
+    with_html '<span>München</span>'
+    field = field_by_xpath(".//span")
+    p field.element.to_s
+  end
+end