hpricot 0.8.2 → 0.8.3

data/lib/hpricot/builder.rb

@@ -7,7 +7,7 @@ module Hpricot
   # XML unescape
   def self.uxs(str)
     str.to_s.
-        gsub(/\&(\w+);/) { [NamedCharacters[$1] || ??].pack("U*") }.
+        gsub(/\&(\w+);/) { [NamedCharacters[$1] || 63].pack("U*") }. # 63 = ?? (query char)
         gsub(/\&\#(\d+);/) { [$1.to_i].pack("U*") }
   end
 

data/lib/hpricot/elements.rb

@@ -9,7 +9,7 @@ module Hpricot
 #   Hpricot::Elements[ele1, ele2, ele3]
 #
 # Assuming that ele1, ele2 and ele3 contain element objects (Hpricot::Elem,
-# Hpricot::Doc, etc.)  
+# Hpricot::Doc, etc.)
 #
 # == Continuing Searches
 #
@@ -29,7 +29,7 @@ module Hpricot
 #   doc = Hpricot("That's my <b>spoon</b>, Tyler.")
 #   doc.at("b").swap("<i>fork</i>")
 #   doc.to_html
-#     #=> "That's my <i>fork</i>, Tyler." 
+#     #=> "That's my <i>fork</i>, Tyler."
 #
 # == Getting More Detailed
 #
@@ -50,7 +50,7 @@ module Hpricot
 # Most of the useful element methods are in the mixins Hpricot::Traverse
 # and Hpricot::Container::Trav.
   class Elements < Array
-  
+
     # Searches this list for any elements (or children of these elements) matching
     # the CSS or XPath expression +expr+.  Root is assumed to be the element scanned.
     #
@@ -65,7 +65,11 @@ module Hpricot
     #
     # See Hpricot::Container::Trav.at for more.
     def at(expr, &blk)
-      search(expr, &blk).first
+      if expr.kind_of? Fixnum
+        super
+      else
+        search(expr, &blk)[0]
+      end
     end
     alias_method :%, :at
 
@@ -138,7 +142,7 @@ module Hpricot
     def prepend(str = nil, &blk)
       each { |x| x.html(x.make(str, &blk) + x.children) }
     end
- 
+
     # Add some HTML just previous to each element in this list.
     # Pass in an HTML +str+, which is turned into Hpricot elements.
     def before(str = nil, &blk)
@@ -151,7 +155,7 @@ module Hpricot
       each { |x| x.parent.insert_after x.make(str, &blk), x }
     end
 
-    # Wraps each element in the list inside the element created by HTML +str+. 
+    # Wraps each element in the list inside the element created by HTML +str+.
     # If more than one element is found in the string, Hpricot locates the
     # deepest spot inside the first element.
     #
@@ -175,7 +179,7 @@ module Hpricot
     # Gets and sets attributes on all matched elements.
     #
     # Pass in a +key+ on its own and this method will return the string value
-    # assigned to that attribute for the first elements.  Or +nil+ if the 
+    # assigned to that attribute for the first elements.  Or +nil+ if the
     # attribute isn't found.
     #
     #   doc.search("a").attr("href")
@@ -185,11 +189,11 @@ module Hpricot
     # matched elements.
     #
     #   doc.search("p").attr("class", "basic")
-    # 
+    #
     # You may also use a Hash to set a series of attributes:
     #
     #   (doc/"a").attr(:class => "basic", :href => "http://hackety.org/")
-    # 
+    #
     # Lastly, a block can be used to rewrite an attribute based on the element
     # it belongs to.  The block will pass in an element.  Return from the block
     # the new value of the attribute.
@@ -203,8 +207,8 @@ module Hpricot
         each do |el|
           el.set_attribute(key, value || blk[el])
         end
-        return self      
-      end    
+        return self
+      end
       if key.is_a? Hash
         key.each { |k,v| self.attr(k,v) }
         return self
@@ -213,7 +217,7 @@ module Hpricot
       end
     end
     alias_method :set, :attr
-  
+
     # Adds the class to all matched elements.
     #
     #   (doc/"p").add_class("bacon")
@@ -237,7 +241,7 @@ module Hpricot
         next unless el.respond_to? :remove_attribute
         el.remove_attribute(name)
       end
-      self      
+      self
     end
 
     # Removes a class from all matched elements.
@@ -247,7 +251,7 @@ module Hpricot
     # Or, to remove all classes:
     #
     #   (doc/"span").remove_class
-    # 
+    #
     def remove_class name = nil
       each do |el|
         next unless el.respond_to? :get_attribute
@@ -258,10 +262,10 @@ module Hpricot
           el.remove_attribute("class")
         end
       end
-      self      
+      self
     end
 
-    ATTR_RE = %r!\[ *(?:(@)([\w\(\)-]+)|([\w\(\)-]+\(\))) *([~\!\|\*$\^=]*) *'?"?([^'"]*)'?"? *\]!i
+    ATTR_RE = %r!\[ *(?:(@)([\w\(\)-]+)|([\w\(\)-]+\(\))) *([~\!\|\*$\^=]*) *'?"?([^'"]*)'?"? *\]!i # " (for emacs)
     BRACK_RE = %r!(\[) *([^\]]*) *\]+!i
     FUNC_RE = %r!(:)?([a-zA-Z0-9\*_-]*)\( *[\"']?([^ \)]*?)['\"]? *\)!
     CUST_RE = %r!(:)([a-zA-Z0-9\*_-]*)()!
@@ -301,7 +305,7 @@ module Hpricot
                     end
                 end
                 args << -1
-                nodes = Elements[*nodes.find_all do |x| 
+                nodes = Elements[*nodes.find_all do |x|
                                       args[-1] += 1
                                       x.send(meth, *args) ? truth : !truth
                                   end]
@@ -419,7 +423,7 @@ module Hpricot
     end
 
     filter ':nth-child' do |arg,i|
-      case arg 
+      case arg
       when 'even'; (parent.containers.index(self) + 1) % 2 == 0
       when 'odd';  (parent.containers.index(self) + 1) % 2 == 1
       else         self == (parent.containers[arg.to_i - 1])
@@ -429,7 +433,7 @@ module Hpricot
     filter ":last-child" do |i|
       self == parent.containers.last
     end
-    
+
     filter ":nth-last-child" do |arg,i|
       self == parent.containers[-1-arg.to_i]
     end
@@ -451,13 +455,13 @@ module Hpricot
     end
 
     filter :empty do |*a|
-      containers.length == 0
+      elem? && inner_html.length == 0
     end
 
     filter :root do |*a|
       self.is_a? Hpricot::Doc
     end
-    
+
     filter 'text' do |*a|
       self.text?
     end
@@ -469,13 +473,13 @@ module Hpricot
     filter :contains do |arg, ignore|
       html.include? arg
     end
-    
-    
+
+
 
     pred_procs =
       {'text()' => proc { |ele, *_| ele.inner_text.strip },
        '@'      => proc { |ele, attr, *_| ele.get_attribute(attr).to_s if ele.elem? }}
-    
+
     oper_procs =
       {'='      => proc { |a,b| a == b },
        '!='     => proc { |a,b| a != b },

data/lib/hpricot/htmlinfo.rb

@@ -484,7 +484,7 @@ module Hpricot
   end
 
   ElementInclusions =
-{"head"=>["link", "meta", "object", "script", "style"], "body"=>["del", "ins"]}
+{"head"=>["link", "meta", "object", "script", "style" , "noscript"], "body"=>["del", "ins"]}
   ElementInclusions.each do |k, v|
     v.each do |name|
       ElementContent[k][name.hash] = :allow

data/lib/hpricot/tag.rb

@@ -43,7 +43,7 @@ module Hpricot
     def initialize e
       @element = e
     end
-    def [] k 
+    def [] k
       Hpricot.uxs((@element.raw_attributes || {})[k])
     end
     def []= k, v
@@ -195,7 +195,7 @@ module Hpricot
     def pathname; "procins()" end
     def raw_string; output("") end
     def output(out, opts = {})
-      out << 
+      out <<
         if_output(opts) do
           "<?#{target}" +
            (content ? " #{content}" : "") +

data/lib/hpricot/tags.rb

@@ -6,7 +6,7 @@ module Hpricot
   # Common sets of attributes.
   AttrCore = [:id, :class, :style, :title]
   AttrI18n = [:lang, 'xml:lang'.intern, :dir]
-  AttrEvents = [:onclick, :ondblclick, :onmousedown, :onmouseup, :onmouseover, :onmousemove, 
+  AttrEvents = [:onclick, :ondblclick, :onmousedown, :onmouseup, :onmouseover, :onmousemove,
       :onmouseout, :onkeypress, :onkeydown, :onkeyup]
   AttrFocus = [:accesskey, :tabindex, :onfocus, :onblur]
   AttrHAlign = [:align, :char, :charoff]
@@ -113,11 +113,11 @@ module Hpricot
     @tagset = XHTMLStrict.tagset.merge \
       :strike => Attrs,
       :center => Attrs,
-      :dir => Attrs + [:compact], 
+      :dir => Attrs + [:compact],
       :noframes => Attrs,
-      :basefont => [:id, :size, :color, :face], 
+      :basefont => [:id, :size, :color, :face],
       :u => Attrs,
-      :menu => Attrs + [:compact], 
+      :menu => Attrs + [:compact],
       :iframe => AttrCore + [:longdesc, :name, :src, :frameborder, :marginwidth, :marginheight, :scrolling, :align, :height, :width],
       :font => AttrCore + AttrI18n + [:size, :color, :face],
       :s => Attrs,

data/lib/hpricot/traverse.rb

@@ -106,16 +106,16 @@ module Hpricot
     # Find all preceding nodes.
     def preceding
       sibs = parent.children
-      si = sibs.index(self) 
-      return Elements[*sibs[0...si]] 
-    end 
- 
+      si = sibs.index(self)
+      return Elements[*sibs[0...si]]
+    end
+
     # Find all nodes which follow the current one.
     def following
-      sibs = parent.children 
-      si = sibs.index(self) + 1 
-      return Elements[*sibs[si...sibs.length]] 
-    end 
+      sibs = parent.children
+      si = sibs.index(self) + 1
+      return Elements[*sibs[si...sibs.length]]
+    end
 
     # Adds elements immediately after this element, contained in the +html+ string.
     def after(html = nil, &blk)
@@ -296,7 +296,7 @@ module Hpricot
               mt = after[%r!:[a-z0-9\\*_-]+!i, 0]
               oop = false
               if mt and not (mt == ":not" or Traverse.method_defined? "filter[#{mt}]")
-                after = $' 
+                after = $'
                 m[2] += mt
                 expr = after
               end
@@ -348,10 +348,10 @@ module Hpricot
     #
     # If _names_ are empty, it yields all elements.
     # If non-empty _names_ are given, it should be list of universal names.
-    # 
+    #
     # A nested element is yielded in depth first order as follows.
     #
-    #   t = Hpricot('<a id=0><b><a id=1 /></b><c id=2 /></a>') 
+    #   t = Hpricot('<a id=0><b><a id=1 /></b><c id=2 /></a>')
     #   t.traverse_element("a", "c") {|e| p e}
     #   # =>
     #   {elem <a id="0"> {elem <b> {emptyelem <a id="1">} </b>} {emptyelem <c id="2">} </a>}
@@ -363,7 +363,7 @@ module Hpricot
     #   t = Hpricot(<<'End')
     #   <html>
     #   <meta name="robots" content="index,nofollow">
-    #   <meta name="author" content="Who am I?">    
+    #   <meta name="author" content="Who am I?">
     #   </html>
     #   End
     #   t.traverse_element("{http://www.w3.org/1999/xhtml}meta") {|e| p e}
@@ -424,19 +424,19 @@ module Hpricot
 
     # Find all preceding sibling elements.   Like the other "sibling" methods, this weeds
     # out text and comment nodes.
-    def preceding_siblings() 
-      sibs = parent.containers 
-      si = sibs.index(self) 
-      return Elements[*sibs[0...si]] 
-    end 
- 
+    def preceding_siblings()
+      sibs = parent.containers
+      si = sibs.index(self)
+      return Elements[*sibs[0...si]]
+    end
+
     # Find sibling elements which follow the current one.   Like the other "sibling" methods, this weeds
     # out text and comment nodes.
-    def following_siblings() 
-      sibs = parent.containers 
-      si = sibs.index(self) + 1 
-      return Elements[*sibs[si...sibs.length]] 
-    end 
+    def following_siblings()
+      sibs = parent.containers
+      si = sibs.index(self) + 1
+      return Elements[*sibs[si...sibs.length]]
+    end
 
     # Puts together an array of neighboring sibling elements based on their proximity
     # to this element.
@@ -507,7 +507,7 @@ module Hpricot
     end
 
     # +find_element+ searches an element which universal name is specified by
-    # the arguments. 
+    # the arguments.
     # It returns nil if not found.
     def find_element(*names)
       traverse_element(*names) {|e| return e }
@@ -777,7 +777,7 @@ module Hpricot
           return author if !author.empty?
         rescue IndexError
         end
-      } 
+      }
 
       if channel = find_element('{http://purl.org/rss/1.0/}channel')
         channel.traverse_element('{http://purl.org/dc/elements/1.1/}creator') {|e|

data/lib/hpricot/xchar.rb

@@ -10,7 +10,7 @@ module Hpricot
 
   ####################################################################
   # XML Character converter, from Sam Ruby:
-  # (see http://intertwingly.net/stories/2005/09/28/xchar.rb). 
+  # (see http://intertwingly.net/stories/2005/09/28/xchar.rb).
   #
   module XChar # :nodoc:
 
@@ -59,7 +59,7 @@ module Hpricot
     # See http://www.w3.org/TR/REC-xml/#charsets for details.
     VALID = [
       0x9, 0xA, 0xD,
-      (0x20..0xD7FF), 
+      (0x20..0xD7FF),
       (0xE000..0xFFFD),
       (0x10000..0x10FFFF)
     ]
@@ -86,7 +86,7 @@ module Hpricot
     # XML unescape
     def uxs(str)
       str.to_s.
-          gsub(/\&\w+;/) { |x| (XChar::PREDEFINED_U[x] || ??).chr }.
+          gsub(/\&\w+;/) { |x| (XChar::PREDEFINED_U[x] || 63).chr }. # 63 = ?? (query char)
           gsub(/\&\#(\d+);/) { [$1.to_i].pack("U*") }
     end
   end

data/test/files/basic.xhtml

@@ -9,7 +9,7 @@
 </head>
 <body id='body1'>
     <p>Sample XHTML for <a id="link1" href="http://code.whytheluckystiff.net/mouseHole/">MouseHole 2</a>.</p>
-    <p class='ohmy'>Please filter <a id="link2" href="http://hobix.com/">me</a>!</p>
+    <p class='ohmy'>Please filter <span></span><a id="link2" href="http://hobix.com/">me</a>!</p>
     <p>The third paragraph</p>
     <p class="last final"><b>THE FINAL PARAGRAPH</b></p>
 </body>

data/test/test_parser.rb

@@ -354,6 +354,16 @@ class TestParser < Test::Unit::TestCase
     assert_equal "HAI", doc.at("body").inner_text
   end
 
+  # http://github.com/hpricot/hpricot/issues#issue/28
+  def test_invalid_inner_text
+    assert_equal "A", Hpricot('A&B;').inner_text[0...1]
+  end
+
+  # http://github.com/hpricot/hpricot/issues#issue/25
+  def test_encoding_compatibility_error
+    Hpricot("<p>\xC3\x9Cber</p><p>M&sup3;</p>").inner_text
+  end
+
   # Reported by Jonathan Nichols on the Hpricot list (24 May 2007)
   def test_self_closed_form
     doc = Hpricot(<<-edoc)
@@ -370,8 +380,8 @@ class TestParser < Test::Unit::TestCase
     @basic = Hpricot.parse(TestFiles::BASIC)
     assert_equal 0, (@basic/"title:parent").size
     assert_equal 3, (@basic/"p:parent").size
-    assert_equal 1, (@basic/"title:empty").size
-    assert_equal 1, (@basic/"p:empty").size
+    assert_equal 3, (@basic/"link:empty").size
+    assert_equal 1, (@basic/"span:empty").size
   end
 
   def test_keep_cdata
@@ -425,4 +435,23 @@ class TestParser < Test::Unit::TestCase
       assert_equal "This is STDOUT", (xml/:peon/:stdout).inner_text
       assert_equal "This is STDERR", (xml/:peon/:stderr).inner_text
   end
+
+  def test_parsing_html_with_noscript
+   doc = Hpricot(<<-edoc)
+    <html>
+    <head>
+    <noscript>
+    <meta http-equiv="refresh" content="0; url=http://www.yoursite.com/noscripts.html"/>
+    </noscript>
+    <meta name="verification" content="7ff5e90iormq5niy6x98j75" />
+    </head>
+    <body>
+    <h1>Testing</h1>
+    </body>
+    </html>
+
+   edoc
+    assert_equal "7ff5e90iormq5niy6x98j75", doc.at("/html/head/meta[@name='verification']")['content']
+ end
+
 end

metadata

@@ -1,7 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: hpricot
 version: !ruby/object:Gem::Version 
-  version: 0.8.2
+  hash: 57
+  prerelease: false
+  segments: 
+  - 0
+  - 8
+  - 3
+  version: 0.8.3
 platform: ruby
 authors: 
 - why the lucky stiff
@@ -9,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-11-05 00:00:00 -06:00
+date: 2010-11-03 00:00:00 -05:00
 default_executable: 
 dependencies: []
 
@@ -21,13 +27,13 @@ extensions:
 - ext/fast_xs/extconf.rb
 - ext/hpricot_scan/extconf.rb
 extra_rdoc_files: 
-- README
+- README.md
 - CHANGELOG
 - COPYING
 files: 
 - CHANGELOG
 - COPYING
-- README
+- README.md
 - Rakefile
 - test/files/basic.xhtml
 - test/files/boingboing.html
@@ -47,6 +53,7 @@ files:
 - test/test_paths.rb
 - test/test_preserved.rb
 - test/test_xml.rb
+- extras/hpricot.png
 - lib/hpricot/blankslate.rb
 - lib/hpricot/builder.rb
 - lib/hpricot/elements.rb
@@ -59,7 +66,6 @@ files:
 - lib/hpricot/traverse.rb
 - lib/hpricot/xchar.rb
 - lib/hpricot.rb
-- extras/mingw-rbconfig.rb
 - ext/hpricot_scan/hpricot_scan.h
 - ext/fast_xs/FastXsService.java
 - ext/hpricot_scan/HpricotCss.java
@@ -84,26 +90,32 @@ rdoc_options:
 - --title
 - The Hpricot Reference
 - --main
-- README
+- README.md
 - --inline-source
 require_paths: 
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
       version: "0"
-  version: 
 required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
       version: "0"
-  version: 
 requirements: []
 
 rubyforge_project: hobix
-rubygems_version: 1.3.5
+rubygems_version: 1.3.7
 signing_key: 
 specification_version: 3
 summary: a swift, liberal HTML parser with a fantastic library