nokogiri 1.5.6.rc1-java → 1.5.6.rc2-java

data/ext/java/nokogiri/internals/NokogiriHandler.java

@@ -62,12 +62,10 @@ import org.xml.sax.ext.DefaultHandler2;
  * @author Yoko Harada <yokolet@gmail.com>
  */
 public class NokogiriHandler extends DefaultHandler2 implements XmlDeclHandler {
-
-    boolean inCDATA = false;
-
-    private Ruby ruby;
-    private RubyClass attrClass;
-    private IRubyObject object;
+    private StringBuffer buffer;
+    private final Ruby ruby;
+    private final RubyClass attrClass;
+    private final IRubyObject object;
 
     /**
      * Stores parse errors with the most-recent error last.
@@ -235,8 +233,11 @@ public class NokogiriHandler extends DefaultHandler2 implements XmlDeclHandler {
 
     @Override
     public void characters(char[] ch, int start, int length) throws SAXException {
-        String target = inCDATA ? "cdata_block" : "characters";
-        call(target, ruby.newString(new String(ch, start, length)));
+        if (buffer != null) {
+          buffer.append(new String(ch, start, length));
+        } else {
+          call("characters", ruby.newString(new String(ch, start, length)));
+        }
     }
 
     @Override
@@ -246,12 +247,13 @@ public class NokogiriHandler extends DefaultHandler2 implements XmlDeclHandler {
 
     @Override
     public void startCDATA() throws SAXException {
-        inCDATA = true;
+        buffer = new StringBuffer();
     }
 
     @Override
     public void endCDATA() throws SAXException {
-        inCDATA = false;
+        call("cdata_block", ruby.newString(buffer.toString()));
+        buffer = null;
     }
 
     @Override

data/ext/java/nokogiri/internals/NokogiriHelpers.java

@@ -63,6 +63,7 @@ import nokogiri.XmlNode;
 import nokogiri.XmlProcessingInstruction;
 import nokogiri.XmlText;
 
+import org.jcodings.specific.UTF8Encoding;
 import org.jruby.Ruby;
 import org.jruby.RubyArray;
 import org.jruby.RubyClass;
@@ -179,7 +180,7 @@ public class NokogiriHelpers {
 
     public static IRubyObject stringOrNil(Ruby runtime, String s) {
         if (s == null) return runtime.getNil();
-        return RubyString.newString(runtime, s);
+        return convertJavaStringToRuby(runtime, s);
     }
     
     public static IRubyObject stringOrNil(Ruby runtime, byte[] bytes) {
@@ -189,7 +190,16 @@ public class NokogiriHelpers {
     
     public static IRubyObject stringOrBlank(Ruby runtime, String s) {
         if (s == null) return runtime.newString();
-        return RubyString.newString(runtime, s);
+        return convertJavaStringToRuby(runtime, s);
+    }
+    
+    private static IRubyObject convertJavaStringToRuby(Ruby runtime, String str) {
+        if (runtime.is1_9()) {
+            ByteList bytes = new ByteList(str.getBytes(RubyEncoding.UTF8), UTF8Encoding.INSTANCE);
+            return RubyString.newString(runtime, bytes);
+        } else {
+            return RubyString.newString(runtime, str);
+        }
     }
 
     /**

data/ext/java/nokogiri/internals/XmlDomParserContext.java

@@ -111,7 +111,7 @@ public class XmlDomParserContext extends ParserContext {
                     "org.apache.xerces.parsers.XIncludeParserConfiguration");
         }
         
-        parser = new XmlDomParser(options);
+        parser = new NokogiriDomParser(options);
         parser.setErrorHandler(errorHandler);
 
         // Fix for Issue#586.  This limits entity expansion up to 100000 and nodes up to 3000.

data/ext/nokogiri/extconf.rb

@@ -27,6 +27,7 @@ end
 
 if RbConfig::MAKEFILE_CONFIG['CC'] =~ /mingw/
   $CFLAGS << " -DIN_LIBXML"
+  $LIBS << " -lz" # TODO why is this necessary?
 end
 
 if RbConfig::MAKEFILE_CONFIG['CC'] =~ /gcc/

data/ext/nokogiri/xslt_stylesheet.c

@@ -109,6 +109,10 @@ static VALUE serialize(VALUE self, VALUE xmlobj)
     return rval ;
 }
 
+static void swallow_superfluous_xml_errors(void * userdata, xmlErrorPtr error, ...)
+{
+}
+
 /*
  *  call-seq:
  *    transform(document, params = [])
@@ -126,12 +130,13 @@ static VALUE serialize(VALUE self, VALUE xmlobj)
  */
 static VALUE transform(int argc, VALUE* argv, VALUE self)
 {
-    VALUE xmldoc, paramobj ;
+    VALUE xmldoc, paramobj, errstr, exception ;
     xmlDocPtr xml ;
     xmlDocPtr result ;
     nokogiriXsltStylesheetTuple *wrapper;
     const char** params ;
     long param_len, j ;
+    int parse_error_occurred ;
 
     rb_scan_args(argc, argv, "11", &xmldoc, &paramobj);
     if (NIL_P(paramobj)) { paramobj = rb_ary_new2(0L) ; }
@@ -158,10 +163,22 @@ static VALUE transform(int argc, VALUE* argv, VALUE self)
     }
     params[param_len] = 0 ;
 
+    errstr = rb_str_new(0, 0);
+    xsltSetGenericErrorFunc((void *)errstr, xslt_generic_error_handler);
+    xmlSetGenericErrorFunc(NULL, (xmlGenericErrorFunc)&swallow_superfluous_xml_errors);
+
     result = xsltApplyStylesheet(wrapper->ss, xml, params);
     free(params);
 
-    if (!result) rb_raise(rb_eRuntimeError, "could not perform xslt transform on document");
+    xsltSetGenericErrorFunc(NULL, NULL);
+    xmlSetGenericErrorFunc(NULL, NULL);
+
+    parse_error_occurred = (Qfalse == rb_funcall(errstr, rb_intern("empty?"), 0));
+
+    if (parse_error_occurred) {
+      exception = rb_exc_new3(rb_eRuntimeError, errstr);
+      rb_exc_raise(exception);
+    }
 
     return Nokogiri_wrap_xml_document((VALUE)0, result) ;
 }

data/lib/nokogiri/version.rb

@@ -1,6 +1,6 @@
 module Nokogiri
   # The version of Nokogiri you are using
-  VERSION = '1.5.6.rc1'
+  VERSION = '1.5.6.rc2'
 
   class VersionInfo # :nodoc:
     def jruby?

data/lib/nokogiri/xml/node.rb

@@ -299,20 +299,7 @@ module Nokogiri
       def add_previous_sibling node_or_tags
         raise ArgumentError.new("A document may not have multiple root nodes.") if parent.is_a?(XML::Document) && !node_or_tags.is_a?(XML::ProcessingInstruction)
 
-        node_or_tags = coerce(node_or_tags)
-        if node_or_tags.is_a?(XML::NodeSet)
-          if text?
-            pivot = Nokogiri::XML::Node.new 'dummy', document
-            add_previous_sibling_node pivot
-          else
-            pivot = self
-          end
-          node_or_tags.each { |n| pivot.send :add_previous_sibling_node, n }
-          pivot.unlink if text?
-        else
-          add_previous_sibling_node node_or_tags
-        end
-        node_or_tags
+        add_sibling :previous, node_or_tags
       end
 
       ###
@@ -325,20 +312,7 @@ module Nokogiri
       def add_next_sibling node_or_tags
         raise ArgumentError.new("A document may not have multiple root nodes.") if parent.is_a?(XML::Document)
         
-        node_or_tags = coerce(node_or_tags)
-        if node_or_tags.is_a?(XML::NodeSet)
-          if text?
-            pivot = Nokogiri::XML::Node.new 'dummy', document
-            add_next_sibling_node pivot
-          else
-            pivot = self
-          end
-          node_or_tags.reverse_each { |n| pivot.send :add_next_sibling_node, n }
-          pivot.unlink if text?
-        else
-          add_next_sibling_node node_or_tags
-        end
-        node_or_tags
+        add_sibling :next, node_or_tags
       end
 
       ####
@@ -761,12 +735,7 @@ module Nokogiri
       # See Node#write_to for a list of +options+.  For formatted output,
       # use Node#to_xhtml instead.
       def to_html options = {}
-        # FIXME: this is a hack around broken libxml versions
-        return dump_html if Nokogiri.uses_libxml? && %w[2 6] === LIBXML_VERSION.split('.')[0..1]
-
-        options[:save_with] |= SaveOptions::DEFAULT_HTML if options[:save_with]
-        options[:save_with] = SaveOptions::DEFAULT_HTML unless options[:save_with]
-        serialize(options)
+        to_format SaveOptions::DEFAULT_HTML, options
       end
 
       ###
@@ -787,12 +756,7 @@ module Nokogiri
       #
       # See Node#write_to for a list of +options+
       def to_xhtml options = {}
-        # FIXME: this is a hack around broken libxml versions
-        return dump_html if Nokogiri.uses_libxml? && %w[2 6] === LIBXML_VERSION.split('.')[0..1]
-
-        options[:save_with] |= SaveOptions::DEFAULT_XHTML if options[:save_with]
-        options[:save_with] = SaveOptions::DEFAULT_XHTML unless options[:save_with]
-        serialize(options)
+        to_format SaveOptions::DEFAULT_XHTML, options
       end
 
       ###
@@ -835,11 +799,7 @@ module Nokogiri
       #
       # See Node#write_to for a list of +options+
       def write_html_to io, options = {}
-        # FIXME: this is a hack around broken libxml versions
-        return (io << dump_html) if Nokogiri.uses_libxml? && %w[2 6] === LIBXML_VERSION.split('.')[0..1]
-
-        options[:save_with] ||= SaveOptions::DEFAULT_HTML
-        write_to io, options
+        write_format_to SaveOptions::DEFAULT_HTML, io, options
       end
 
       ###
@@ -847,11 +807,7 @@ module Nokogiri
       #
       # See Node#write_to for a list of +options+
       def write_xhtml_to io, options = {}
-        # FIXME: this is a hack around broken libxml versions
-        return (io << dump_html) if Nokogiri.uses_libxml? && %w[2 6] === LIBXML_VERSION.split('.')[0..1]
-
-        options[:save_with] ||= SaveOptions::DEFAULT_XHTML
-        write_to io, options
+        write_format_to SaveOptions::DEFAULT_XHTML, io, options
       end
 
       ###
@@ -898,6 +854,43 @@ module Nokogiri
 
       private
 
+      def add_sibling next_or_previous, node_or_tags
+        impl = (next_or_previous == :next) ? :add_next_sibling_node : :add_previous_sibling_node
+        iter = (next_or_previous == :next) ? :reverse_each          : :each
+
+        node_or_tags = coerce node_or_tags
+        if node_or_tags.is_a?(XML::NodeSet)
+          if text?
+            pivot = Nokogiri::XML::Node.new 'dummy', document
+            send impl, pivot
+          else
+            pivot = self
+          end
+          node_or_tags.send(iter) { |n| pivot.send impl, n }
+          pivot.unlink if text?
+        else
+          send impl, node_or_tags
+        end
+        node_or_tags
+      end
+
+      def to_format save_option, options
+        # FIXME: this is a hack around broken libxml versions
+        return dump_html if Nokogiri.uses_libxml? && %w[2 6] === LIBXML_VERSION.split('.')[0..1]
+
+        options[:save_with] |= save_option if options[:save_with]
+        options[:save_with] = save_option unless options[:save_with]
+        serialize(options)
+      end
+
+      def write_format_to save_option, io, options
+        # FIXME: this is a hack around broken libxml versions
+        return (io << dump_html) if Nokogiri.uses_libxml? && %w[2 6] === LIBXML_VERSION.split('.')[0..1]
+
+        options[:save_with] ||= save_option
+        write_to io, options
+      end
+
       def extract_params params # :nodoc:
         # Pop off our custom function handler if it exists
         handler = params.find { |param|

data/lib/nokogiri/xml/sax/parser.rb

@@ -68,6 +68,7 @@ module Nokogiri
 
         # Create a new Parser with +doc+ and +encoding+
         def initialize doc = Nokogiri::XML::SAX::Document.new, encoding = 'UTF-8'
+          check_encoding(encoding)
           @encoding = encoding
           @document = doc
           @warned   = false
@@ -87,6 +88,7 @@ module Nokogiri
         ###
         # Parse given +io+
         def parse_io io, encoding = 'ASCII'
+          check_encoding(encoding)
           @encoding = encoding
           ctx = ParserContext.io(io, ENCODINGS[encoding])
           yield ctx if block_given?
@@ -109,6 +111,11 @@ module Nokogiri
           yield ctx if block_given?
           ctx.parse_with self
         end
+
+        private
+        def check_encoding(encoding)
+          raise ArgumentError.new("'#{encoding}' is not a valid encoding") unless ENCODINGS[encoding]
+        end
       end
     end
   end

data/lib/nokogiri/xslt.rb

@@ -24,7 +24,7 @@ module Nokogiri
       def parse string, modules = {}
         modules.each do |url, klass|
           XSLT.register url, klass
-	end
+        end
 
         if Nokogiri.jruby?
           Stylesheet.parse_stylesheet_doc(XML.parse(string), string)

data/tasks/cross_compile.rb

@@ -4,10 +4,10 @@ HOST = Rake::ExtensionCompiler.mingw_host
 
 require 'mini_portile'
 $recipes = {}
-$recipes[:zlib]     = MiniPortile.new "zlib", "1.2.5"
+$recipes[:zlib]     = MiniPortile.new "zlib",     "1.2.7"
 $recipes[:libiconv] = MiniPortile.new "libiconv", "1.13.1"
-$recipes[:libxml2]  = MiniPortile.new "libxml2", "2.7.7"
-$recipes[:libxslt]  = MiniPortile.new "libxslt", "1.1.26"
+$recipes[:libxml2]  = MiniPortile.new "libxml2",  "2.7.7"
+$recipes[:libxslt]  = MiniPortile.new "libxslt",  "1.1.26"
 $recipes.each { |_, recipe| recipe.host = HOST }
 
 file "lib/nokogiri/nokogiri.rb" do

data/test/html/test_document.rb

@@ -369,6 +369,29 @@ eohtml
         assert_equal('Hello world!', node.inner_text.strip)
       end
 
+      def test_doc_type
+        html = Nokogiri::HTML(<<-eohtml)
+        <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
+          <html xmlns="http://www.w3.org/1999/xhtml">
+            <body>
+              <p>Rainbow Dash</p>
+            </body>
+          </html>
+        eohtml
+        assert_equal "html", html.internal_subset.name
+        assert_equal "-//W3C//DTD XHTML 1.1//EN", html.internal_subset.external_id
+        assert_equal "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd", html.internal_subset.system_id
+        assert_equal "<!DOCTYPE html PUBLIC \"-//W3C//DTD XHTML 1.1//EN\" \"http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd\">", html.to_s[0,97]
+      end
+      
+      def test_content_size
+        html = Nokogiri::HTML('<div>
+</div>')
+        assert_equal 1, html.content.size
+        assert_equal 1, html.content.split("").size
+        assert_equal "\n", html.content
+      end
+
       def test_find_by_xpath
         found = @html.xpath('//div/a')
         assert_equal 3, found.length

data/test/test_xslt_transforms.rb

@@ -189,8 +189,37 @@ encoding="iso-8859-1" indent="yes"/>
           Nokogiri::XSLT.quote_params(params.to_a.flatten)))
       check_params result_doc, params
     end
+
+    def test_xslt_paramaters
+      xslt_str = <<-EOX
+  <xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" >
+    <xsl:template match="/">
+      <xsl:value-of select="$foo" />
+    </xsl:template>
+  </xsl:stylesheet>
+      EOX
+
+      xslt = Nokogiri::XSLT(xslt_str)
+      doc = Nokogiri::XML("<root />")
+      assert_match %r{bar}, xslt.transform(doc, Nokogiri::XSLT.quote_params('foo' => 'bar')).to_s
+    end
+
+    def test_xslt_transform_error
+      xslt_str = <<-EOX
+  <xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" >
+    <xsl:template match="/">
+      <xsl:value-of select="$foo" />
+    </xsl:template>
+  </xsl:stylesheet>
+      EOX
+
+      xslt = Nokogiri::XSLT(xslt_str)
+      doc = Nokogiri::XML("<root />")
+      assert_raises(RuntimeError) { xslt.transform(doc) }
+    end
   end
 
+
   def test_xslt_parse_error
     xslt_str = <<-EOX
 <xsl:stylesheet version="1.0"
@@ -210,6 +239,7 @@ encoding="iso-8859-1" indent="yes"/>
     assert_raises(RuntimeError) { Nokogiri::XSLT.parse(xslt_str) }
   end
 
+
   def test_passing_a_non_document_to_transform
     xsl = Nokogiri::XSLT('<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"></xsl:stylesheet>')
     assert_raises(ArgumentError) { xsl.transform("<div></div>") }

data/test/xml/sax/test_parser.rb

@@ -250,6 +250,11 @@ module Nokogiri
           assert_raises(ArgumentError) { @parser.parse_memory(nil) }
         end
 
+        def test_bad_encoding_args
+          assert_raises(ArgumentError) { XML::SAX::Parser.new(Doc.new, 'not an encoding') }
+          assert_raises(ArgumentError) { @parser.parse_io(StringIO.new('<root/>'), 'not an encoding')}
+        end
+
         def test_ctag
           @parser.parse_memory(<<-eoxml)
             <p id="asdfasdf">

data/test/xml/test_node.rb

@@ -1031,7 +1031,15 @@ EOXML
         subject = Nokogiri::XML::Node.new 'foo', document
         ns = subject.add_namespace nil, 'bar'
         subject.namespace = ns
-        assert_match subject.to_xml, /xmlns="bar"/
+        assert_match(/xmlns="bar"/, subject.to_xml)
+      end
+
+      def test_text_node_colon
+        document = Nokogiri::XML::Document.new
+        root = Nokogiri::XML::Node.new 'foo', document
+        document.root = root
+        root << "<a>hello:with_colon</a>"
+        assert_match(/hello:with_colon/, document.to_xml)
       end
     end
   end