hpricot 0.6 → 0.6.161

data/ext/hpricot_scan/hpricot_scan.java.rl

@@ -6,7 +6,9 @@ import org.jruby.RubyClass;
 import org.jruby.RubyHash;
 import org.jruby.RubyModule;
 import org.jruby.RubyNumeric;
+import org.jruby.RubyObjectAdapter;
 import org.jruby.RubyString;
+import org.jruby.javasupport.JavaEmbedUtils;
 import org.jruby.runtime.Block;
 import org.jruby.runtime.CallbackFactory;
 import org.jruby.runtime.builtin.IRubyObject;
@@ -15,6 +17,7 @@ import org.jruby.runtime.load.BasicLibraryService;
 
 public class HpricotScanService implements BasicLibraryService {
        public static String NO_WAY_SERIOUSLY="*** This should not happen, please send a bug report with the HTML you're parsing to why@whytheluckystiff.net.  So sorry!";
+       private static RubyObjectAdapter rubyApi;
 
        public void ELE(IRubyObject N) {
          if (tokend > tokstart || text) {
@@ -239,8 +242,8 @@ IRubyObject hpricot_scan(IRubyObject recv, IRubyObject port) {
   }
 
   buffer_size = BUFSIZE;
-  if (recv.getInstanceVariable("@buffer_size") != null) {
-    bufsize = recv.getInstanceVariable("@buffer_size");
+  if (rubyApi.getInstanceVariable(recv, "@buffer_size") != null) {
+    bufsize = rubyApi.getInstanceVariable(recv, "@buffer_size");
     if (!bufsize.isNil()) {
       buffer_size = RubyNumeric.fix2int(bufsize);
     }
@@ -359,5 +362,6 @@ public static void Init_hpricot_scan(Ruby runtime) {
   CallbackFactory fact = runtime.callbackFactory(HpricotScanService.class);
   mHpricot.getMetaClass().defineMethod("scan",fact.getSingletonMethod("__hpricot_scan",IRubyObject.class));
   mHpricot.defineClassUnder("ParseError",runtime.getClass("Exception"),runtime.getClass("Exception").getAllocator());
+  rubyApi = JavaEmbedUtils.newObjectAdapter();
 }
 }

data/ext/hpricot_scan/hpricot_scan.rl

@@ -8,6 +8,12 @@
  */
 #include <ruby.h>
 
+#ifndef RARRAY_LEN
+#define RARRAY_LEN(arr)  RARRAY(arr)->len
+#define RSTRING_LEN(str) RSTRING(str)->len
+#define RSTRING_PTR(str) RSTRING(str)->ptr
+#endif
+
 #define NO_WAY_SERIOUSLY "*** This should not happen, please send a bug report with the HTML you're parsing to why@whytheluckystiff.net.  So sorry!"
 
 static VALUE sym_xmldecl, sym_doctype, sym_procins, sym_stag, sym_etag, sym_emptytag, sym_comment,
@@ -169,7 +175,7 @@ VALUE hpricot_scan(VALUE self, VALUE port)
       /* We've used up the entire buffer storing an already-parsed token
        * prefix that must be preserved.  Likely caused by super-long attributes.
        * See ticket #13. */
-      rb_raise(rb_eHpricotParseError, "ran out of buffer space on element <%s>, starting on line %d.", RSTRING(tag)->ptr, curline);
+      rb_raise(rb_eHpricotParseError, "ran out of buffer space on element <%s>, starting on line %d.", RSTRING_PTR(tag), curline);
     }
 
     if ( rb_respond_to( port, s_read ) )
@@ -182,8 +188,8 @@ VALUE hpricot_scan(VALUE self, VALUE port)
     }
 
     StringValue(str);
-    memcpy( p, RSTRING(str)->ptr, RSTRING(str)->len );
-    len = RSTRING(str)->len;
+    memcpy( p, RSTRING_PTR(str), RSTRING_LEN(str) );
+    len = RSTRING_LEN(str);
     nread += len;
 
     /* If this is the last buffer, tack on an EOF. */
@@ -199,7 +205,7 @@ VALUE hpricot_scan(VALUE self, VALUE port)
       free(buf);
       if ( !NIL_P(tag) )
       {
-        rb_raise(rb_eHpricotParseError, "parse error on element <%s>, starting on line %d.\n" NO_WAY_SERIOUSLY, RSTRING(tag)->ptr, curline);
+        rb_raise(rb_eHpricotParseError, "parse error on element <%s>, starting on line %d.\n" NO_WAY_SERIOUSLY, RSTRING_PTR(tag), curline);
       }
       else
       {

data/lib/hpricot/builder.rb

@@ -1,8 +1,23 @@
 require 'hpricot/tags'
-require 'hpricot/xchar'
+require 'fast_xs'
 require 'hpricot/blankslate'
 
 module Hpricot
+  PREDEFINED = {
+    34 => '"', # quotation mark
+    38 => '&',  # ampersand
+    60 => '<',   # left angle bracket
+    62 => '>'    # right angle bracket
+  }
+  PREDEFINED_U = PREDEFINED.inject({}) { |hsh, (k, v)| hsh[v] = k; hsh }
+
+  # XML unescape
+  def self.uxs(str)
+    str.to_s.
+        gsub(/\&\w+;/) { |x| (PREDEFINED_U[x] || ??).chr }.
+        gsub(/\&\#(\d+);/) { [$1.to_i].pack("U*") }
+  end
+
   def self.build(ele = Doc.new, assigns = {}, &blk)
     ele.extend Builder
     assigns.each do |k, v|
@@ -32,7 +47,7 @@ module Hpricot
 
     # Write a +string+ to the HTML stream, making sure to escape it.
     def text!(string)
-      @children << Text.new(Hpricot.xs(string))
+      @children << Text.new(string.fast_xs)
     end
 
     # Write a +string+ to the HTML stream without escaping it.
@@ -75,16 +90,16 @@ module Hpricot
       # turn arguments into children or attributes
       childs = []
       attrs = args.grep(Hash)
-      childs.concat((args - attrs).map do |x|
+      childs.concat((args - attrs).flatten.map do |x|
         if x.respond_to? :to_html
           Hpricot.make(x.to_html)
         elsif x
-          Text.new(Hpricot.xs(x))
+          Text.new(x.fast_xs)
         end
       end.flatten)
       attrs = attrs.inject({}) do |hsh, ath|
         ath.each do |k, v|
-          hsh[k] = Hpricot.xs(v.to_s) if v
+          hsh[k] = v.to_s.fast_xs if v
         end
         hsh
       end

data/lib/hpricot/elements.rb

@@ -130,25 +130,25 @@ module Hpricot
     # Add to the end of the contents inside each element in this list.
     # Pass in an HTML +str+, which is turned into Hpricot elements.
     def append(str = nil, &blk)
-      each { |x| x.html(x.children + Hpricot.make(str, &blk)) }
+      each { |x| x.html(x.children + x.make(str, &blk)) }
     end
 
     # Add to the start of the contents inside each element in this list.
     # Pass in an HTML +str+, which is turned into Hpricot elements.
     def prepend(str = nil, &blk)
-      each { |x| x.html(Hpricot.make(str, &blk) + x.children) }
+      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)
-      each { |x| x.parent.insert_before Hpricot.make(str, &blk), x }
+      each { |x| x.parent.insert_before x.make(str, &blk), x }
     end
 
     # Just after each element in this list, add some HTML.
     # Pass in an HTML +str+, which is turned into Hpricot elements.
     def after(str = nil, &blk)
-      each { |x| x.parent.insert_after Hpricot.make(str, &blk), x }
+      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+. 
@@ -161,7 +161,7 @@ module Hpricot
     # This code wraps every link on the page inside a +div.link+ and a +div.link_inner+ nest.
     def wrap(str = nil, &blk)
       each do |x|
-        wrap = Hpricot.make(str, &blk)
+        wrap = x.make(str, &blk)
         nest = wrap.detect { |w| w.respond_to? :children }
         unless nest
           raise Exception, "No wrapping element found."
@@ -261,7 +261,7 @@ module Hpricot
       self      
     end
 
-    ATTR_RE = %r!\[ *(?:(@)([\w\(\)-]+)|([\w\(\)-]+\(\))) *([~\!\|\*$\^=]*) *'?"?([^\]'"]*)'?"? *\]!i
+    ATTR_RE = %r!\[ *(?:(@)([\w\(\)-]+)|([\w\(\)-]+\(\))) *([~\!\|\*$\^=]*) *'?"?([^'"]*)'?"? *\]!i
     BRACK_RE = %r!(\[) *([^\]]*) *\]+!i
     FUNC_RE = %r!(:)?([a-zA-Z0-9\*_-]*)\( *[\"']?([^ \)]*?)['\"]? *\)!
     CUST_RE = %r!(:)([a-zA-Z0-9\*_-]*)()!

data/lib/hpricot/parse.rb

@@ -12,13 +12,14 @@ module Hpricot
   # Hpricot.parse parses <i>input</i> and return a document tree.
   # represented by Hpricot::Doc.
   def Hpricot.parse(input = nil, opts = {}, &blk)
-    Doc.new(make(input, opts, &blk))
+    Doc.new(make(input, opts, &blk), opts)
   end
 
   # Hpricot::XML parses <i>input</i>, disregarding all the HTML rules
   # and returning a document tree.
-  def Hpricot.XML(input, opts = {})
-    Doc.new(make(input, opts.merge(:xml => true)))
+  def Hpricot.XML(input = nil, opts = {}, &blk)
+    opts.merge! :xml => true
+    Doc.new(make(input, opts, &blk), opts)
   end
 
   # :stopdoc:

data/lib/hpricot/tag.rb

@@ -3,8 +3,9 @@ module Hpricot
 
   class Doc
     attr_accessor :children
-    def initialize(children = [])
+    def initialize(children = [], options = {})
       @children = children ? children.each { |c| c.parent = self }  : []
+      @options = options
     end
     def output(out, opts = {})
       @children.each do |n|
@@ -12,6 +13,9 @@ module Hpricot
       end
       out
     end
+    def make(input = nil, &blk)
+      Hpricot.make(input, @options, &blk)
+    end
     def altered!; end
   end
 
@@ -100,7 +104,7 @@ module Hpricot
       if @raw_attributes
         @raw_attributes.map do |aname, aval|
           " #{aname}" +
-            (aval ? "=\"#{aval}\"" : "")
+            (aval ? "=#{html_quote aval}" : "")
         end.join
       end
     end

data/lib/hpricot/traverse.rb

@@ -20,6 +20,16 @@ module Hpricot
     # Is this object a stranded end tag?
     def bogusetag?() BogusETag::Trav === self end
 
+    # Parses an HTML string, making an HTML fragment based on
+    # the options used to create the container document.
+    def make(input = nil, &blk)
+      if parent and parent.respond_to? :make
+        parent.make(input, &blk)
+      else
+        Hpricot.make(input, &blk)
+      end
+    end
+
     # Builds an HTML string from this node and its contents.
     # If you need to write to a stream, try calling <tt>output(io)</tt>
     # as a method on this object.
@@ -109,12 +119,12 @@ module Hpricot
 
     # Adds elements immediately after this element, contained in the +html+ string.
     def after(html = nil, &blk)
-      parent.insert_after(Hpricot.make(html, &blk), self)
+      parent.insert_after(make(html, &blk), self)
     end
 
     # Adds elements immediately before this element, contained in the +html+ string.
     def before(html = nil, &blk)
-      parent.insert_before(Hpricot.make(html, &blk), self)
+      parent.insert_before(make(html, &blk), self)
     end
 
 
@@ -122,7 +132,7 @@ module Hpricot
     # in the +html+ string.
     def swap(html = nil, &blk)
       parent.altered!
-      parent.replace_child(self, Hpricot.make(html, &blk))
+      parent.replace_child(self, make(html, &blk))
     end
 
     def get_subnode(*indexes)
@@ -158,7 +168,7 @@ module Hpricot
         when Array
           self.children = inner
         else
-          self.children = Hpricot.make(inner, &blk)
+          self.children = make(inner, &blk)
         end
         reparent self.children
       else
@@ -513,8 +523,9 @@ module Hpricot
 
     def get_elements_by_tag_name(*a)
       list = Elements[]
+      a.delete("*")
       traverse_element(*a.map { |tag| [tag, "{http://www.w3.org/1999/xhtml}#{tag}"] }.flatten) do |e|
-          list << e
+        list << e if e.elem?
       end
       list
     end
@@ -806,7 +817,7 @@ module Hpricot
     def set_attribute(name, val)
       altered!
       self.raw_attributes ||= {}
-      self.raw_attributes[name.to_s] = Hpricot.xs(val)
+      self.raw_attributes[name.to_s] = val.fast_xs
     end
     alias_method :[]=, :set_attribute
     def remove_attribute(name)

data/test/test_alter.rb

@@ -58,6 +58,18 @@ class TestAlter < Test::Unit::TestCase
     assert_changed(@basic, "p[@class]", all_c2) { |p| p['class'].nil? }
   end
 
+  def test_xml_casing
+    doc = Hpricot.XML("<root><wildCat>text</wildCat></root>")
+    (doc/:root/:wildCat).after("<beanPole>gravity</beanPole>")
+    assert_equal doc.to_s, "<root><wildCat>text</wildCat><beanPole>gravity</beanPole></root>"
+
+    frag = Hpricot.XML do
+      b { i "A bit of HTML" }
+    end
+    (frag/:b).after("<beanPole>gravity</beanPole>")
+    assert_equal frag.to_s, "<b><i>A bit of HTML</i></b><beanPole>gravity</beanPole>"
+  end
+
   def assert_changed original, selector, set, &block
     assert set.all?(&block)
     assert Hpricot(original.to_html).search(selector).all?(&block)

data/test/test_builder.rb

@@ -21,4 +21,17 @@ class TestBuilder < Test::Unit::TestCase
     assert_equal "<b>&#8364;&#8226;</b>", doc.to_html
     assert_equal "\342\202\254\342\200\242", doc.at("text()").to_s
   end
+
+  def test_escaping_attrs
+    text = "<span style='font-family:\"MS Mincho\"'>Some text</span>"
+    assert_equal "<span style=\"font-family:\\\"MS Mincho\\\"\">Some text</span>",
+      Hpricot(text).to_html
+  end
+
+  def test_korean_utf8_entities
+    # a = '한글'
+    a = "\xed\x95\x9c\xea\xb8\x80"
+    doc = Hpricot() { b a }
+    assert_equal "<b>&#54620;&#44544;</b>", doc.to_html
+  end
 end

data/test/test_parser.rb

@@ -47,6 +47,13 @@ class TestParser < Test::Unit::TestCase
     assert_equal 'link1', @basic.get_elements_by_tag_name('a')[0].get_attribute('id')
     assert_equal 'link1', @basic.get_elements_by_tag_name('body')[0].get_element_by_id('link1').get_attribute('id')
   end
+  
+  def test_get_elements_by_tag_name_star
+    simple = Hpricot.parse("<div><p id='first'>First</p><p id='second'>Second</p></div>")
+    assert_equal 3, simple.get_elements_by_tag_name("*").size
+    assert_equal 1, simple.get_elements_by_tag_name("div").size
+    assert_equal 2, simple.get_elements_by_tag_name("p").size
+  end
 
   def test_output_basic
     @basic = Hpricot.parse(TestFiles::BASIC)

data/test/test_paths.rb

@@ -13,4 +13,13 @@ class TestParser < Test::Unit::TestCase
       assert_equal ele, @basic.at(ele.xpath)
     end
   end
+  def test_attr_brackets
+    doc = Hpricot('<input name="vendor[porkpies]"/>')
+    assert_equal 1, (doc/'input[@name^="vendor[porkpies]"]').length
+    assert_equal 1, (doc/'input[@name="vendor[porkpies]"]').length
+    assert_equal 0, (doc/'input[@name$="]]]]]"]').length
+
+    doc = Hpricot('<input name="vendor[porkpies][meaty]"/>')
+    assert_equal 1, (doc/'input[@name^="vendor[porkpies][meaty]"]').length
+  end
 end

metadata

@@ -1,81 +1,80 @@
 --- !ruby/object:Gem::Specification 
-rubygems_version: 0.9.0
-specification_version: 1
 name: hpricot
 version: !ruby/object:Gem::Version 
-  version: "0.6"
-date: 2007-06-15 00:00:00 -07:00
-summary: a swift, liberal HTML parser with a fantastic library
-require_paths: 
-- lib/i686-linux
-- lib
-email: why@ruby-lang.org
-homepage: http://code.whytheluckystiff.net/hpricot/
-rubyforge_project: 
-description: a swift, liberal HTML parser with a fantastic library
-autorequire: 
-default_executable: 
-bindir: bin
-has_rdoc: true
-required_ruby_version: !ruby/object:Gem::Version::Requirement 
-  requirements: 
-  - - ">"
-    - !ruby/object:Gem::Version 
-      version: 0.0.0
-  version: 
+  version: 0.6.161
 platform: ruby
-signing_key: 
-cert_chain: 
-post_install_message: 
 authors: 
 - why the lucky stiff
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2008-03-19 00:00:00 -05:00
+default_executable: 
+dependencies: []
+
+description: a swift, liberal HTML parser with a fantastic library
+email: why@ruby-lang.org
+executables: []
+
+extensions: 
+- ext/fast_xs/extconf.rb
+- ext/hpricot_scan/extconf.rb
+extra_rdoc_files: 
+- README
+- CHANGELOG
+- COPYING
 files: 
 - CHANGELOG
 - COPYING
 - README
 - Rakefile
 - test/files
-- test/test_preserved.rb
-- test/test_paths.rb
-- test/load_files.rb
-- test/test_xml.rb
-- test/test_parser.rb
-- test/test_alter.rb
-- test/test_builder.rb
-- test/files/why.xml
-- test/files/boingboing.html
-- test/files/uswebgen.html
-- test/files/immob.html
-- test/files/week9.html
-- test/files/utf8.html
 - test/files/basic.xhtml
+- test/files/boingboing.html
 - test/files/cy0.html
-- test/files/tenderlove.html
+- test/files/immob.html
 - test/files/pace_application.html
+- test/files/tenderlove.html
+- test/files/uswebgen.html
+- test/files/utf8.html
+- test/files/week9.html
+- test/files/why.xml
+- test/load_files.rb
+- test/test_alter.rb
+- test/test_builder.rb
+- test/test_parser.rb
+- test/test_paths.rb
+- test/test_preserved.rb
+- test/test_xml.rb
 - lib/hpricot
-- lib/hpricot.rb
-- lib/i686-linux
+- lib/hpricot/blankslate.rb
 - lib/hpricot/builder.rb
+- lib/hpricot/elements.rb
 - lib/hpricot/htmlinfo.rb
-- lib/hpricot/xchar.rb
 - lib/hpricot/inspect.rb
 - lib/hpricot/modules.rb
 - lib/hpricot/parse.rb
 - lib/hpricot/tag.rb
-- lib/hpricot/traverse.rb
-- lib/hpricot/elements.rb
 - lib/hpricot/tags.rb
-- lib/hpricot/blankslate.rb
+- 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/hpricot_scan.java.java
 - ext/hpricot_scan/HpricotScanService.java
+- ext/fast_xs/fast_xs.c
 - ext/hpricot_scan/hpricot_scan.c
+- ext/fast_xs/extconf.rb
 - ext/hpricot_scan/extconf.rb
 - ext/hpricot_scan/hpricot_common.rl
-- ext/hpricot_scan/hpricot_scan.rl
 - ext/hpricot_scan/hpricot_scan.java.rl
-test_files: []
-
+- ext/hpricot_scan/hpricot_scan.rl
+has_rdoc: true
+homepage: http://code.whytheluckystiff.net/hpricot/
+post_install_message: 
 rdoc_options: 
 - --quiet
 - --title
@@ -83,15 +82,27 @@ rdoc_options:
 - --main
 - README
 - --inline-source
-extra_rdoc_files: 
-- README
-- CHANGELOG
-- COPYING
-executables: []
-
-extensions: 
-- ext/hpricot_scan/extconf.rb
+require_paths: 
+- lib/universal-darwin9.0
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
 requirements: []
 
-dependencies: []
+rubyforge_project: 
+rubygems_version: 1.0.1
+signing_key: 
+specification_version: 2
+summary: a swift, liberal HTML parser with a fantastic library
+test_files: []