nokogiri 1.5.6.rc2-java → 1.5.6.rc3-java

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

@@ -72,6 +72,8 @@ public class XmlDomParserContext extends ParserContext {
         "http://apache.org/xml/features/nonvalidating/load-dtd-grammar";
     protected static final String FEATURE_INCLUDE_IGNORABLE_WHITESPACE =
         "http://apache.org/xml/features/dom/include-ignorable-whitespace";
+    protected static final String CONTINUE_AFTER_FATAL_ERROR =
+        "http://apache.org/xml/features/continue-after-fatal-error";
     protected static final String FEATURE_NOT_EXPAND_ENTITY =
         "http://apache.org/xml/features/dom/create-entity-ref-nodes";
     protected static final String FEATURE_VALIDATION = "http://xml.org/sax/features/validation";
@@ -110,17 +112,21 @@ public class XmlDomParserContext extends ParserContext {
             System.setProperty("org.apache.xerces.xni.parser.XMLParserConfiguration",
                     "org.apache.xerces.parsers.XIncludeParserConfiguration");
         }
-        
+
         parser = new NokogiriDomParser(options);
         parser.setErrorHandler(errorHandler);
 
         // Fix for Issue#586.  This limits entity expansion up to 100000 and nodes up to 3000.
-        setProperty(SECURITY_MANAGER, new org.apache.xerces.util.SecurityManager());        
-        
+        setProperty(SECURITY_MANAGER, new org.apache.xerces.util.SecurityManager());
+
         if (options.noBlanks) {
             setFeature(FEATURE_INCLUDE_IGNORABLE_WHITESPACE, false);
         }
 
+        if (options.recover) {
+            setFeature(CONTINUE_AFTER_FATAL_ERROR, true);
+        }
+
         if (options.dtdValid) {
             setFeature(FEATURE_VALIDATION, true);
         }

data/ext/nokogiri/xml_document.c

@@ -367,7 +367,7 @@ static VALUE new(int argc, VALUE *argv, VALUE klass)
  *
  *  For more information on why this probably is *not* a good thing in general,
  *  please direct your browser to
- *  http://tenderlovemaking.com/2009/04/23/namespaces-in-xml/
+ *  http://tenderlovemaking.com/2009/04/23/namespaces-in-xml.html
  */
 VALUE remove_namespaces_bang(VALUE self)
 {

data/ext/nokogiri/xml_node.c

@@ -84,7 +84,7 @@ static xmlNodePtr xmlReplaceNodeWrapper(xmlNodePtr pivot, xmlNodePtr new_node)
   }
 
   /* work around libxml2 issue: https://bugzilla.gnome.org/show_bug.cgi?id=615612 */
-  if (retval->type == XML_TEXT_NODE) {
+  if (retval && retval->type == XML_TEXT_NODE) {
     if (retval->prev && retval->prev->type == XML_TEXT_NODE) {
       retval = xmlTextMerge(retval->prev, retval);
     }
@@ -699,23 +699,40 @@ static VALUE set(VALUE self, VALUE property, VALUE value)
  *
  * Get the value for +attribute+
  */
-static VALUE get(VALUE self, VALUE attribute)
+static VALUE get(VALUE self, VALUE rattribute)
 {
   xmlNodePtr node;
-  xmlChar* propstr ;
-  VALUE rval ;
-  Data_Get_Struct(self, xmlNode, node);
+  xmlChar* value = 0;
+  VALUE rvalue ;
+  char* attribute = 0;
+  char *colon = 0, *attr_name = 0, *prefix = 0;
+  xmlNsPtr ns;
 
-  if(NIL_P(attribute)) return Qnil;
+  if (NIL_P(rattribute)) return Qnil;
 
-  propstr = xmlGetProp(node, (xmlChar *)StringValuePtr(attribute));
+  Data_Get_Struct(self, xmlNode, node);
+  attribute = strdup(StringValuePtr(rattribute));
+
+  colon = strchr(attribute, ':');
+  if (colon) {
+    (*colon) = 0 ; /* create two null-terminated strings of the prefix and attribute name */
+    prefix = attribute ;
+    attr_name = colon + 1 ;
+    ns = xmlSearchNs(node->doc, node, (const xmlChar *)(prefix));
+    if (ns) {
+      value = xmlGetNsProp(node, (xmlChar*)(attr_name), ns->href);
+    }
+  } else {
+    value = xmlGetNoNsProp(node, (xmlChar*)attribute);
+  }
 
-  if(!propstr) return Qnil;
+  free(attribute);
+  if (!value) return Qnil;
 
-  rval = NOKOGIRI_STR_NEW2(propstr);
+  rvalue = NOKOGIRI_STR_NEW2(value);
+  xmlFree(value);
 
-  xmlFree(propstr);
-  return rval ;
+  return rvalue ;
 }
 
 /*
@@ -892,7 +909,7 @@ static VALUE node_type(VALUE self)
  *
  * Set the content for this Node
  */
-static VALUE set_content(VALUE self, VALUE content)
+static VALUE native_content(VALUE self, VALUE content)
 {
   xmlNodePtr node, child, next ;
   Data_Get_Struct(self, xmlNode, node);
@@ -1288,7 +1305,7 @@ static VALUE in_context(VALUE self, VALUE _str, VALUE _options)
       child_iter = node;
       while (child_iter->parent)
         child_iter = child_iter->parent;
-	
+
       if (child_iter->type == XML_DOCUMENT_FRAG_NODE)
         node->doc->children = NULL;
     }
@@ -1458,6 +1475,7 @@ void init_xml_node()
   rb_define_method(klass, "create_external_subset", create_external_subset, 3);
   rb_define_method(klass, "pointer_id", pointer_id, 0);
   rb_define_method(klass, "line", line, 0);
+  rb_define_method(klass, "native_content=", native_content, 1);
 
   rb_define_private_method(klass, "process_xincludes", process_xincludes, 1);
   rb_define_private_method(klass, "in_context", in_context, 2);
@@ -1467,7 +1485,6 @@ void init_xml_node()
   rb_define_private_method(klass, "replace_node", replace, 1);
   rb_define_private_method(klass, "dump_html", dump_html, 0);
   rb_define_private_method(klass, "native_write_to", native_write_to, 4);
-  rb_define_private_method(klass, "native_content=", set_content, 1);
   rb_define_private_method(klass, "get", get, 1);
   rb_define_private_method(klass, "set", set, 2);
   rb_define_private_method(klass, "set_namespace", set_namespace, 1);

data/ext/nokogiri/xml_sax_parser.c

@@ -7,6 +7,7 @@ static ID id_start_document, id_end_document, id_start_element, id_end_element;
 static ID id_start_element_namespace, id_end_element_namespace;
 static ID id_comment, id_characters, id_xmldecl, id_error, id_warning;
 static ID id_cdata_block, id_cAttribute;
+static ID id_processing_instruction;
 
 #define STRING_OR_NULL(str) \
    (RTEST(str) ? StringValuePtr(str) : NULL)
@@ -236,6 +237,19 @@ static void cdata_block(void * ctx, const xmlChar * value, int len)
   rb_funcall(doc, id_cdata_block, 1, string);
 }
 
+static void processing_instruction(void * ctx, const xmlChar * name, const xmlChar * content)
+{
+  VALUE self = NOKOGIRI_SAX_SELF(ctx);
+  VALUE doc = rb_iv_get(self, "@document");
+
+  rb_funcall( doc,
+              id_processing_instruction,
+              2,
+              NOKOGIRI_STR_NEW2(name),
+              NOKOGIRI_STR_NEW2(content)
+  );
+}
+
 static void deallocate(xmlSAXHandlerPtr handler)
 {
   NOKOGIRI_DEBUG_START(handler);
@@ -260,6 +274,7 @@ static VALUE allocate(VALUE klass)
   handler->warning = warning_func;
   handler->error = error_func;
   handler->cdataBlock = cdata_block;
+  handler->processingInstruction = processing_instruction;
   handler->initialized = XML_SAX2_MAGIC;
 
   return Data_Wrap_Struct(klass, NULL, deallocate, handler);
@@ -290,4 +305,5 @@ void init_xml_sax_parser()
   id_cAttribute     = rb_intern("Attribute");
   id_start_element_namespace = rb_intern("start_element_namespace");
   id_end_element_namespace = rb_intern("end_element_namespace");
+  id_processing_instruction = rb_intern("processing_instruction");
 }

data/lib/nokogiri/version.rb

@@ -1,6 +1,6 @@
 module Nokogiri
   # The version of Nokogiri you are using
-  VERSION = '1.5.6.rc2'
+  VERSION = '1.5.6.rc3'
 
   class VersionInfo # :nodoc:
     def jruby?

data/lib/nokogiri/xml/document.rb

@@ -149,13 +149,15 @@ module Nokogiri
       # Non-prefixed default namespaces (as in "xmlns=") are not included
       # in the hash.
       #
-      # Note this is a very expensive operation in current implementation, as it
-      # traverses the entire graph, and also has to bring each node across the
-      # libxml bridge into a ruby object.
+      # Note that this method does an xpath lookup for nodes with
+      # namespaces, and as a result the order may be dependent on the
+      # implementation of the underlying XML library.
+      #
       def collect_namespaces
-        ns = {}
-        traverse { |j| ns.merge!(j.namespaces) }
-        ns
+        xpath("//namespace::*").inject({}) do |hash, ns|
+          hash[["xmlns",ns.prefix].compact.join(":")] = ns.href if ns.prefix != "xml"
+          hash
+        end
       end
 
       # Get the list of decorators given +key+

data/lib/nokogiri/xml/document_fragment.rb

@@ -13,7 +13,8 @@ module Nokogiri
         children = if ctx
                      # Fix for issue#490
                      if Nokogiri.jruby?
-                       ctx.parse("<root>#{tags}</root>").xpath("/root/node()")
+                       # fix for issue #770
+                       ctx.parse("<root #{namespace_declarations(ctx)}>#{tags}</root>").children
                      else
                        ctx.parse(tags)
                      end
@@ -93,6 +94,14 @@ module Nokogiri
 
       private
 
+      # fix for issue 770
+      def namespace_declarations ctx
+        ctx.namespace_scopes.map do |namespace|
+          prefix = namespace.prefix.nil? ? "" : ":#{namespace.prefix}"
+          %Q{xmlns#{prefix}="#{namespace.href}"}
+        end.join ' '
+      end
+
       def coerce data
         return super unless String === data
 

data/lib/nokogiri/xml/node.rb

@@ -251,14 +251,13 @@ module Nokogiri
       ###
       # Get the attribute value for the attribute +name+
       def [] name
-        return nil unless key?(name.to_s)
         get(name.to_s)
       end
 
       ###
       # Set the attribute value for the attribute +name+ to +value+
       def []= name, value
-        set name.to_s, value
+        set name.to_s, value.to_s
       end
 
       ###
@@ -377,17 +376,22 @@ module Nokogiri
       #
       # Also see related method +swap+.
       def replace node_or_tags
+        # We cannot replace a text node directly, otherwise libxml will return
+        # an internal error at parser.c:13031, I don't know exactly why
+        # libxml is trying to find a parent node that is an element or document
+        # so I can't tell if this is bug in libxml or not. issue #775.
+        if text?
+          replacee = Nokogiri::XML::Node.new 'dummy', document
+          add_previous_sibling_node replacee
+          unlink
+          return replacee.replace node_or_tags
+        end
+
         node_or_tags = coerce(node_or_tags)
+
         if node_or_tags.is_a?(XML::NodeSet)
-          if text?
-            replacee = Nokogiri::XML::Node.new 'dummy', document
-            add_previous_sibling_node replacee
-            unlink
-          else
-            replacee = self
-          end
-          node_or_tags.each { |n| replacee.add_previous_sibling n }
-          replacee.unlink
+          node_or_tags.each { |n| add_previous_sibling n }
+          unlink
         else
           replace_node node_or_tags
         end

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

@@ -158,6 +158,13 @@ module Nokogiri
         # +string+ contains the cdata content
         def cdata_block string
         end
+
+        ###
+        # Called when processing instructions are found
+        # +name+ is the target of the instruction
+        # +content+ is the value of the instruction
+        def processing_instruction name, content
+        end
       end
     end
   end

data/lib/nokogiri/xml/xpath_context.rb

@@ -6,7 +6,7 @@ module Nokogiri
       # Register namespaces in +namespaces+
       def register_namespaces(namespaces)
         namespaces.each do |k, v|
-          k = k.gsub(/.*:/,'') # strip off 'xmlns:' or 'xml:'
+          k = k.to_s.gsub(/.*:/,'') # strip off 'xmlns:' or 'xml:'
           register_ns(k, v)
         end
       end

data/test/helper.rb

@@ -78,6 +78,7 @@ module Nokogiri
         attr_reader :data, :comments, :cdata_blocks, :start_elements_namespace
         attr_reader :errors, :warnings, :end_elements_namespace
         attr_reader :xmldecls
+        attr_reader :processing_instructions
 
         def xmldecl version, encoding, standalone
           @xmldecls = [version, encoding, standalone].compact
@@ -141,6 +142,11 @@ module Nokogiri
           @cdata_blocks += [string]
           super
         end
+
+        def processing_instruction name, content
+          @processing_instructions ||= []
+          @processing_instructions << [name, content]
+        end
       end
     end
   end

data/test/html/test_document_fragment.rb

@@ -24,6 +24,11 @@ module Nokogiri
         end
       end
 
+      def test_colons_are_not_removed
+        doc = Nokogiri::HTML::DocumentFragment.parse("<span>3:30pm</span>")
+        assert_match /3:30/, doc.to_s
+      end
+
       def test_parse_encoding
         fragment = "<div>hello world</div>"
         f = Nokogiri::HTML::DocumentFragment.parse fragment, 'ISO-8859-1'

data/test/xml/sax/test_parser.rb

@@ -173,7 +173,12 @@ module Nokogiri
             end
           end
 
-          assert_equal doc.errors.length, @parser.document.errors.length
+          # when using JRuby Nokogiri, more errors will be generated as the DOM
+          # parser continue to parse an ill formed document, while the sax parser
+          # will stop at the first error
+          unless Nokogiri.jruby?
+            assert_equal doc.errors.length, @parser.document.errors.length
+          end
         end
 
         def test_parse_with_memory_argument
@@ -313,6 +318,15 @@ module Nokogiri
                        @parser.document.start_elements
         end
 
+        def test_processing_instruction
+          @parser.parse_memory(<<-eoxml)
+            <?xml-stylesheet href="a.xsl" type="text/xsl"?>
+            <?xml version="1.0"?>
+          eoxml
+          assert_equal [['xml-stylesheet', 'href="a.xsl" type="text/xsl"']],
+                       @parser.document.processing_instructions
+        end
+
         if Nokogiri.uses_libxml? # JRuby SAXParser only parses well-formed XML documents
           def test_parse_document
             @parser.parse_memory(<<-eoxml)

data/test/xml/test_builder.rb

@@ -209,6 +209,25 @@ module Nokogiri
         assert_equal ["bbb","ccc"], builder.doc.at_css("aaa").children.collect(&:name)
       end
 
+      def test_raw_xml_append_with_namespaces
+        doc = Nokogiri::XML::Builder.new do |xml|
+          xml.root("xmlns:foo" => "x", "xmlns" => "y") do
+            xml << '<Element foo:bar="bazz"/>'
+          end
+        end.doc
+
+        el = doc.at 'Element'
+        assert_not_nil el
+
+        assert_equal 'y', el.namespace.href
+        assert_nil el.namespace.prefix
+
+        attr = el.attributes["bar"]
+        assert_not_nil attr
+        assert_not_nil attr.namespace
+        assert_equal "foo", attr.namespace.prefix
+      end
+
       def test_cdata
         builder = Nokogiri::XML::Builder.new do
           root {

data/test/xml/test_document.rb

@@ -16,11 +16,27 @@ module Nokogiri
         @xml = Nokogiri::XML.parse(File.read(XML_FILE), XML_FILE)
       end
 
+      def test_document_with_initial_space
+        doc = Nokogiri::XML(" <?xml version='1.0' encoding='utf-8' ?><first \>")
+        assert_equal 2, doc.children.size
+      end
+
       def test_root_set_to_nil
         @xml.root = nil
         assert_equal nil, @xml.root
       end
 
+      def test_ignore_unknown_namespace
+        doc = Nokogiri::XML(<<-eoxml)
+        <xml>
+          <unknown:foo xmlns='hello' />
+          <bar />
+        </xml>
+        eoxml
+        refute doc.xpath('//foo').first.namespace # assert that the namespace is nil
+        refute_empty doc.xpath('//bar'), "bar wasn't found in the document" # bar should be part of the doc
+      end
+
       def test_collect_namespaces
         doc = Nokogiri::XML(<<-eoxml)
         <xml>
@@ -716,26 +732,43 @@ module Nokogiri
         assert @xml.children.respond_to?(:awesome!)
       end
 
-      def test_java_integration
-        if Nokogiri.jruby?
+      if Nokogiri.jruby?
+        def wrap_java_document
           require 'java'
           factory = javax.xml.parsers.DocumentBuilderFactory.newInstance
           builder = factory.newDocumentBuilder
           document = builder.newDocument
           root = document.createElement("foo")
           document.appendChild(root)
-          noko_doc = Nokogiri::XML::Document.wrap(document)
-          assert_equal 'foo', noko_doc.root.name
+          Nokogiri::XML::Document.wrap(document)
+        end
+      end
 
-          noko_doc = Nokogiri::XML(<<eoxml)
+      def test_java_integration
+        skip("Ruby doesn't have the wrap method") unless Nokogiri.jruby?
+        noko_doc = wrap_java_document
+        assert_equal 'foo', noko_doc.root.name
+
+        noko_doc = Nokogiri::XML(<<eoxml)
 <foo xmlns='hello'>
   <bar xmlns:foo='world' />
 </foo>
 eoxml
-          dom = noko_doc.to_java
-          assert dom.kind_of? org.w3c.dom.Document
-          assert_equal 'foo', dom.getDocumentElement().getTagName()
-        end
+        dom = noko_doc.to_java
+        assert dom.kind_of? org.w3c.dom.Document
+        assert_equal 'foo', dom.getDocumentElement().getTagName()
+      end
+
+      def test_add_child
+        skip("Ruby doesn't have the wrap method") unless Nokogiri.jruby?
+        doc = wrap_java_document
+        doc.root.add_child "<bar />"
+      end
+
+      def test_can_be_closed
+        f = File.open XML_FILE
+        Nokogiri::XML f
+        f.close
       end
     end
   end