sanitize 1.2.2.dev.20101028 → 1.2.2.dev.20101118

data/HISTORY

@@ -11,9 +11,9 @@ Version 1.2.2 (git)
     have been whitelisted by other transformers. [Suggested by Nicholas Evans]
   * Added a :process_text_nodes config setting. If set to true, Sanitize will
     pass text nodes to transformers. The default is false. [Ardie Saeidi]
-  * Added a workaround for a bug in Nokogiri 1.4.2 and higher (issue #315) that
-    causes "</body></html>" to be appended to the CDATA inside unterminated
-    script and style elements.
+  * Bumped minimum Nokogiri version to 1.4.4 to avoid a bug in 1.4.2+ (issue
+    #315) that caused "</body></html>" to be appended to the CDATA inside
+    unterminated script and style elements.
 
 Version 1.2.1 (2010-04-20)
   * Added a :remove_contents config setting. If set to true, Sanitize will

data/lib/sanitize/version.rb

@@ -1,3 +1,3 @@
 class Sanitize
-  VERSION = '1.2.2.dev.20101028'
+  VERSION = '1.2.2.dev.20101118'
 end

data/lib/sanitize.rb

@@ -27,7 +27,6 @@ require 'sanitize/config'
 require 'sanitize/config/restricted'
 require 'sanitize/config/basic'
 require 'sanitize/config/relaxed'
-require 'sanitize/transformers/fix_fragment_cdata'
 
 class Sanitize
   attr_reader :config
@@ -91,14 +90,6 @@ class Sanitize
     # is generated at runtime by transformers, and is cleared before and after
     # a fragment is cleaned (so it applies only to a specific fragment).
     @whitelist_nodes = []
-
-    # Workaround for a fragment parsing bug in Nokogiri >= 1.4.2. The naïve
-    # version check is fine here; there are no side effects for unaffected
-    # versions except slightly worse performance, and I plan to remove this hack
-    # as soon as Nokogiri fixes the bug on their end.
-    if Nokogiri::VERSION > '1.4.1'
-      @config[:transformers] << Transformers::FIX_FRAGMENT_CDATA
-    end
   end
 
   # Returns a sanitized copy of _html_.

metadata

@@ -7,8 +7,8 @@ version: !ruby/object:Gem::Version
   - 2
   - 2
   - dev
-  - 20101028
-  version: 1.2.2.dev.20101028
+  - 20101118
+  version: 1.2.2.dev.20101118
 platform: ruby
 authors: 
 - Ryan Grove
@@ -16,7 +16,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-10-28 00:00:00 -07:00
+date: 2010-11-18 00:00:00 -08:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -30,8 +30,8 @@ dependencies:
         segments: 
         - 1
         - 4
-        - 1
-        version: 1.4.1
+        - 4
+        version: 1.4.4
   type: :runtime
   version_requirements: *id001
 - !ruby/object:Gem::Dependency 
@@ -80,7 +80,6 @@ files:
 - lib/sanitize/config/relaxed.rb
 - lib/sanitize/config/restricted.rb
 - lib/sanitize/config.rb
-- lib/sanitize/transformers/fix_fragment_cdata.rb
 - lib/sanitize/version.rb
 - lib/sanitize.rb
 has_rdoc: true

data/lib/sanitize/transformers/fix_fragment_cdata.rb

@@ -1,27 +0,0 @@
-class Sanitize; module Transformers
-
-  # Nokogiri 1.4.2 and higher contain a fragment parsing bug that causes the
-  # string "</body></html>" to be appended to the CDATA inside an unterminated
-  # <script> or <style> element. This transformer works around this bug by
-  # finding affected elements and removing the spurious text.
-  #
-  # See http://github.com/tenderlove/nokogiri/issues#issue/315
-  FIX_FRAGMENT_CDATA = lambda do |env|
-    node_name = env[:node_name]
-
-    if node_name == 'script' || node_name == 'style'
-      node = env[:node]
-
-      unless node.children.empty?
-        last_child = node.children.last
-
-        if last_child.text? && last_child.content =~ %r|</body></html>$|
-          last_child.content = last_child.content.chomp('</body></html>')
-        end
-      end
-    end
-
-    nil
-  end
-
-end; end