change: CData.new no longer accepts nil as the content argument

An upcoming libxml2 change will force us to make a behavioral change
of some kind, so I'm choosing to bring CData in line with the other
character data classes (Comment and Text).

The alternative would have been to continue accepting nil; but then
the constructed node would contain a blank string (currently in this
case Node#content returns nil). I think it's preferable to force
callers to be explicit here.

See upstream commit GNOME/libxml2@9991fae4

Author: flavorjones

CHANGELOG.md

@@ -18,6 +18,11 @@ Nokogiri follows [Semantic Versioning](https://semver.org/), please see the [REA
 * `Node#clone`, `NodeSet#clone`, and `*::Document#clone` all properly copy the metaclass of the original as expected. Previously, `#clone` had been aliased to `#dup` for these classes (since v1.3.0 in 2009). [#316, #3117] @flavorjones
 
 
+### Changed
+
+* [CRuby] `Nokogiri::XML::CData.new` no longer accepts `nil` as the content argument, making `CData` behave like other character data classes (like `Comment` and `Text`). This change was necessitated by behavioral changes in the upcoming libxml 2.13.0 release. If you wish to create an empty CDATA node, pass an empty string. [#3156] @flavorjones
+
+
 ## v1.16.3 / 2024-03-15
 
 ### Dependencies

ext/java/nokogiri/XmlCdata.java

@@ -46,6 +46,9 @@ public class XmlCdata extends XmlText
     IRubyObject rbDocument = args[0];
     content = args[1];
 
+    if (content.isNil()) {
+      throw context.runtime.newTypeError("expected second parameter to be a String, received NilClass");
+    }
     if (!(rbDocument instanceof XmlNode)) {
       String msg = "expected first parameter to be a Nokogiri::XML::Document, received " + rbDocument.getMetaClass();
       throw context.runtime.newTypeError(msg);

ext/nokogiri/xml_cdata.c

@@ -20,11 +20,10 @@ rb_xml_cdata_s_new(int argc, VALUE *argv, VALUE klass)
   VALUE rb_content;
   VALUE rb_rest;
   VALUE rb_node;
-  xmlChar *c_content = NULL;
-  int c_content_len = 0;
 
   rb_scan_args(argc, argv, "2*", &rb_document, &rb_content, &rb_rest);
 
+  Check_Type(rb_content, T_STRING);
   if (!rb_obj_is_kind_of(rb_document, cNokogiriXmlNode)) {
     rb_raise(rb_eTypeError,
              "expected first parameter to be a Nokogiri::XML::Document, received %"PRIsVALUE,
@@ -40,15 +39,8 @@ rb_xml_cdata_s_new(int argc, VALUE *argv, VALUE klass)
     c_document = noko_xml_document_unwrap(rb_document);
   }
 
-  if (!NIL_P(rb_content)) {
-    c_content = (xmlChar *)StringValuePtr(rb_content);
-    c_content_len = RSTRING_LENINT(rb_content);
-  }
-
-  c_node = xmlNewCDataBlock(c_document, c_content, c_content_len);
-
+  c_node = xmlNewCDataBlock(c_document, (xmlChar *)StringValueCStr(rb_content), RSTRING_LENINT(rb_content));
   noko_xml_document_pin_node(c_node);
-
   rb_node = noko_xml_node_wrap(klass, c_node);
   rb_obj_call_init(rb_node, argc, argv);
 

ext/nokogiri/xml_comment.c

@@ -23,25 +23,20 @@ new (int argc, VALUE *argv, VALUE klass)
 
   rb_scan_args(argc, argv, "2*", &document, &content, &rest);
 
+  Check_Type(content, T_STRING);
   if (rb_obj_is_kind_of(document, cNokogiriXmlNode)) {
     document = rb_funcall(document, document_id, 0);
   } else if (!rb_obj_is_kind_of(document, cNokogiriXmlDocument)
              && !rb_obj_is_kind_of(document, cNokogiriXmlDocumentFragment)) {
     rb_raise(rb_eArgError, "first argument must be a XML::Document or XML::Node");
   }
-
   xml_doc = noko_xml_document_unwrap(document);
 
-  node = xmlNewDocComment(
-           xml_doc,
-           (const xmlChar *)StringValueCStr(content)
-         );
-
+  node = xmlNewDocComment(xml_doc, (const xmlChar *)StringValueCStr(content));
+  noko_xml_document_pin_node(node);
   rb_node = noko_xml_node_wrap(klass, node);
   rb_obj_call_init(rb_node, argc, argv);
 
-  noko_xml_document_pin_node(node);
-
   if (rb_block_given_p()) { rb_yield(rb_node); }
 
   return rb_node;

ext/nokogiri/xml_text.c

@@ -20,6 +20,7 @@ rb_xml_text_s_new(int argc, VALUE *argv, VALUE klass)
 
   rb_scan_args(argc, argv, "2*", &rb_string, &rb_document, &rb_rest);
 
+  Check_Type(rb_string, T_STRING);
   if (!rb_obj_is_kind_of(rb_document, cNokogiriXmlNode)) {
     rb_raise(rb_eTypeError,
              "expected second parameter to be a Nokogiri::XML::Document, received %"PRIsVALUE,
@@ -36,9 +37,7 @@ rb_xml_text_s_new(int argc, VALUE *argv, VALUE klass)
   }
 
   c_node = xmlNewDocText(c_document, (xmlChar *)StringValueCStr(rb_string));
-
   noko_xml_document_pin_node(c_node);
-
   rb_node = noko_xml_node_wrap(klass, c_node) ;
   rb_obj_call_init(rb_node, argc, argv);
 

test/xml/test_cdata.rb

@@ -28,11 +28,10 @@
       assert_same(doc, node.document)
     end
 
-    it "has nil content when passed nil" do
-      node = Nokogiri::XML::CDATA.new(Nokogiri::XML::Document.new, nil)
-
-      assert_instance_of(Nokogiri::XML::CDATA, node)
-      assert_nil(node.content)
+    it "when passed nil raises TypeError" do
+      assert_raises(TypeError) do
+        Nokogiri::XML::CDATA.new(Nokogiri::XML::Document.new, nil)
+      end
     end
 
     it "does not accept anything but a string" do