From d20c58f74c3ead85a12277aa211ec9be96703fba Mon Sep 17 00:00:00 2001
From: moro <moronatural@gmail.com>
Date: Tue, 18 Nov 2008 10:43:43 +0900
Subject: [PATCH] Link#matches_text?() use @element.text as text

It decodes character references.
Userful for multibyte languages(eg. Japanese).

And also uses @element.inner_html to match with HTML (eg.image link)
---
 lib/webrat/core/link.rb       |   22 ++++++++++++++++------
 spec/webrat/core/link_spec.rb |   41 +++++++++++++++++++++++++++++++----------
 2 files changed, 47 insertions(+), 16 deletions(-)

diff --git a/lib/webrat/core/link.rb b/lib/webrat/core/link.rb
index 6980810..23da0f1 100644
--- a/lib/webrat/core/link.rb
+++ b/lib/webrat/core/link.rb
@@ -22,17 +22,15 @@ module Webrat
     end
     
     def matches_text?(link_text)
-      html = text.gsub('&#xA0;',' ').gsub('&nbsp;', ' ')
-      
       if link_text.is_a?(Regexp)
         matcher = link_text
       else
         matcher = /#{Regexp.escape(link_text.to_s)}/i
       end
-      
-      html =~ matcher || title =~ matcher
+
+      replace_nbsp(text) =~ matcher || replace_nbsp_ref(inner_html) =~ matcher || title =~ matcher
     end
-    
+
     def matches_id?(id_or_regexp)
       if id_or_regexp.is_a?(Regexp)
         (id =~ id_or_regexp) ? true : false
@@ -40,9 +38,13 @@ module Webrat
         (id == id_or_regexp) ? true : false
       end
     end
+
+    def inner_html
+      @element.inner_html
+    end
     
     def text
-      @element.inner_html
+      @element.text
     end
     
   protected
@@ -108,5 +110,13 @@ module Webrat
       end
     end
 
+  private
+    def replace_nbsp(str)
+      str.gsub([0xA0].pack('U'), ' ')
+    end
+
+    def replace_nbsp_ref(str)
+      str.gsub('&#xA0;',' ').gsub('&nbsp;', ' ')
+    end
   end
 end
diff --git a/spec/webrat/core/link_spec.rb b/spec/webrat/core/link_spec.rb
index bf20341..ef6ddf6 100755
--- a/spec/webrat/core/link_spec.rb
+++ b/spec/webrat/core/link_spec.rb
@@ -5,6 +5,7 @@ describe Webrat::Link do
 
   before do
     @session = mock(Webrat::TestSession)
+    @link_text_with_nbsp = 'Link' + [0xA0].pack("U") + 'Text'
   end
   
   it "should pass through relative urls" do
@@ -22,42 +23,62 @@ describe Webrat::Link do
   
   it "should matches_text? on regexp" do
     link = Webrat::Link.new(@session, nil)
-    link.should_receive(:text).and_return("Link Text")
+    link.should_receive(:text).and_return(@link_text_with_nbsp)
     link.matches_text?(/link/i).should == 0
   end
   
   it "should matches_text? on link_text" do
     link = Webrat::Link.new(@session, nil)
-    link.should_receive(:text).and_return("Link Text")
+    link.should_receive(:text).and_return(@link_text_with_nbsp)
     link.matches_text?("Link Text").should == 0
   end
   
   it "should matches_text? on substring" do
     link = Webrat::Link.new(@session, nil)
-    link.should_receive(:text).and_return("Link Text")
+    link.should_receive(:text).and_return(@link_text_with_nbsp)
     link.matches_text?("nk Te").should_not be_nil
   end
   
   it "should not matches_text? on link_text case insensitive" do
     link = Webrat::Link.new(@session, nil)
-    link.should_receive(:text).and_return("Link Text")
+    link.should_receive(:text).and_return(@link_text_with_nbsp)
+    link.should_receive(:inner_html).and_return('Link&nbsp;Text')
     link.should_receive(:title).and_return(nil)
     link.matches_text?("link_text").should == false
   end
   
-  it "should match text including &nbsp;" do
+  it "should match text not include &nbsp;" do
     link = Webrat::Link.new(@session, nil)
-    link.should_receive(:text).and_return("Link&nbsp;Text")
-    link.matches_text?("Link Text").should == 0
+    link.should_receive(:text).and_return('LinkText')
+    link.matches_text?("LinkText").should == 0
   end
   
   it "should not matches_text? on wrong text" do 
     link = Webrat::Link.new(@session, nil)
-    link.should_receive(:text).and_return("Some Other Link")
+    nbsp = [0xA0].pack("U")
+    link.should_receive(:text).and_return("Some"+nbsp+"Other"+nbsp+"Link")
+    link.should_receive(:inner_html).and_return("Some&nbsp;Other&nbsp;Link")
     link.should_receive(:title).and_return(nil)
     link.matches_text?("Link Text").should == false
   end
-  
+
+  it "should match text including character reference" do
+    no_ko_gi_ri = [0x30CE,0x30B3,0x30AE,0x30EA]
+    nokogiri_ja_kana = no_ko_gi_ri.pack("U*")
+    nokogiri_char_ref = no_ko_gi_ri.map{|c| "&#x%X;" % c }.join("")
+
+    link = Webrat::Link.new(@session, nil)
+    link.should_receive(:text).and_return(nokogiri_ja_kana)
+    link.matches_text?(nokogiri_ja_kana).should == 0
+  end
+
+  it "should match img link" do
+    link = Webrat::Link.new(@session, nil)
+    link.should_receive(:text).and_return('')
+    link.should_receive(:inner_html).and_return('<img src="logo.png" />')
+    link.matches_text?('logo.png').should == 10
+  end
+
   it "should matches_id? on exact matching id" do
     link = Webrat::Link.new(@session, nil)
     link.should_receive(:id).and_return("some_id")
@@ -76,4 +97,4 @@ describe Webrat::Link do
     link.matches_id?(/some/).should == true
   end
   
-end
\ No newline at end of file
+end
-- 
1.6.0