fix: fragment context parsing

CHANGELOG.md

@@ -52,6 +52,7 @@ We welcome feedback on this API at [#2360](https://github.com/sparklemotion/noko
 
 * `Node#wrap` and `NodeSet#wrap` now also accept a `Node` type argument, which will be `dup`ed for each wrapper. For cases where many nodes are being wrapped, creating a `Node` once using `Document#create_element` and passing that `Node` multiple times is significantly faster than re-parsing markup on each call. [[#2657](https://github.com/sparklemotion/nokogiri/issues/2657)]
 * [CRuby] Invocation of custom XPath or CSS handler functions may now use the `nokogiri` namespace prefix. Historically, the JRuby implementation _required_ this namespace but the CRuby implementation did not support it. It's recommended that all XPath and CSS queries use the `nokogiri` namespace going forward. Invocation without the namespace is planned for deprecation in v1.15.0 and removal in a future release. [[#2147](https://github.com/sparklemotion/nokogiri/issues/2147)]
+* `HTML5::Document#quirks_mode` and `HTML5::DocumentFragment#quirks_mode` expose the quirks mode used by the parser.
 
 
 ### Improved
@@ -94,6 +95,7 @@ We welcome feedback on this API at [#2360](https://github.com/sparklemotion/noko
 * [CRuby] `Nokogiri::HTML5::Document#url` now correctly returns the URL passed to the constructor method. Previously it always returned `nil`. [[#2583](https://github.com/sparklemotion/nokogiri/issues/2583)]
 * [CRuby] `HTML5` encoding detection is now case-insensitive with respect to `meta` tag charset declaration. [[#2693](https://github.com/sparklemotion/nokogiri/issues/2693)]
 * [CRuby] `HTML5` fragment parsing in context of an annotation-xml node now works. Previously this rarely-used path invoked rb_funcall with incorrect parameters, resulting in an exception, a fatal error, or potentially a segfault. [[#2692](https://github.com/sparklemotion/nokogiri/issues/2692)]
+* [CRuby] `HTML5` quirks mode during fragment parsing more closely matches document parsing. [[#2646](https://github.com/sparklemotion/nokogiri/issues/2646)]
 * [JRuby] Fixed a bug with adding the same namespace to multiple nodes via `#add_namespace_definition`. [[#1247](https:<//github.com/sparklemotion/nokogiri/issues/1247)]
 * [JRuby] `NodeSet#[]` now raises a TypeError if passed an invalid parameter type. [[#2211](https://github.com/sparklemotion/nokogiri/issues/2211)]
 

ext/nokogiri/gumbo.c

@@ -361,6 +361,7 @@ parse_continue(VALUE parse_args)
   build_tree(doc, (xmlNodePtr)doc, output->document);
   VALUE rdoc = noko_xml_document_wrap(args->klass, doc);
   rb_iv_set(rdoc, "@url", args->url_or_frag);
+  rb_iv_set(rdoc, "@quirks_mode", INT2NUM(output->document->v.document.doc_type_quirks_mode));
   args->doc = NULL; // The Ruby runtime now owns doc so don't delete it.
   add_errors(output, rdoc, args->input, args->url_or_frag);
   return rdoc;
@@ -517,8 +518,11 @@ fragment(
   // Quirks mode.
   VALUE doc = rb_funcall(doc_fragment, rb_intern_const("document"), 0);
   VALUE dtd = rb_funcall(doc, internal_subset, 0);
-  if (NIL_P(dtd)) {
+  VALUE doc_quirks_mode = rb_iv_get(doc, "@quirks_mode");
+  if (NIL_P(ctx) || NIL_P(doc_quirks_mode)) {
     quirks_mode = GUMBO_DOCTYPE_NO_QUIRKS;
+  } else if (NIL_P(dtd)) {
+    quirks_mode = GUMBO_DOCTYPE_QUIRKS;
   } else {
     VALUE dtd_name = rb_funcall(dtd, name, 0);
     VALUE pubid = rb_funcall(dtd, rb_intern_const("external_id"), 0);
@@ -565,6 +569,7 @@ fragment_continue(VALUE parse_args)
   args->doc = NULL; // The Ruby runtime owns doc so make sure we don't delete it.
   xmlNodePtr xml_frag = extract_xml_node(doc_fragment);
   build_tree(xml_doc, xml_frag, output->root);
+  rb_iv_set(doc_fragment, "@quirks_mode", INT2NUM(output->document->v.document.doc_type_quirks_mode));
   add_errors(output, doc_fragment, args->input, rb_utf8_str_new_static("#fragment", 9));
   return Qnil;
 }

lib/nokogiri/html5/document.rb

@@ -21,6 +21,18 @@
 
 module Nokogiri
   module HTML5
+    # Enum for the HTML5 parser quirks mode values. Values returned by HTML5::Document#quirks_mode
+    #
+    # See https://dom.spec.whatwg.org/#concept-document-quirks for more information on HTML5 quirks
+    # mode.
+    #
+    # Since v1.14.0
+    module QuirksMode
+      NO_QUIRKS = 0 # The document was parsed in "no-quirks" mode
+      QUIRKS = 1 # The document was parsed in "quirks" mode
+      LIMITED_QUIRKS = 2 # The document was parsed in "limited-quirks" mode
+    end
+
     # Since v1.12.0
     #
     # 💡 HTML5 functionality is not available when running JRuby.
@@ -29,6 +41,13 @@ class Document < Nokogiri::HTML4::Document
       # Document.read_memory
       attr_reader :url
 
+      # Get the parser's quirks mode value. See HTML5::QuirksMode.
+      #
+      # This method returns `nil` if the parser was not invoked (e.g., `Nokogiri::HTML5::Document.new`).
+      #
+      # Since v1.14.0
+      attr_reader :quirks_mode
+
       class << self
         # :call-seq:
         #   parse(input)
@@ -110,6 +129,7 @@ def do_parse(string_or_io, url, encoding, options)
       def initialize(*args) # :nodoc:
         super
         @url = nil
+        @quirks_mode = nil
       end
 
       # :call-seq:

lib/nokogiri/html5/document_fragment.rb

@@ -28,6 +28,13 @@ class DocumentFragment < Nokogiri::HTML4::DocumentFragment
       attr_accessor :document
       attr_accessor :errors
 
+      # Get the parser's quirks mode value. See HTML5::QuirksMode.
+      #
+      # This method returns `nil` if the parser was not invoked (e.g., `Nokogiri::HTML5::DocumentFragment.new(doc)`).
+      #
+      # Since v1.14.0
+      attr_reader :quirks_mode
+
       # Create a document fragment.
       def initialize(doc, tags = nil, ctx = nil, options = {})
         self.document = doc

test/html5/test_quirks_mode.rb

@@ -0,0 +1,119 @@
+# encoding: utf-8
+# frozen_string_literal: true
+
+require "helper"
+
+describe Nokogiri::HTML5 do
+  describe "Document#quirks_mode" do
+    let(:document) { Nokogiri::HTML5::Document.parse(html) }
+
+    describe "without parsing anything" do
+      it "returns nil" do
+        assert_nil(Nokogiri::HTML5::Document.new.quirks_mode)
+      end
+    end
+
+    describe "on a document with a doctype" do
+      let(:html) { "<!DOCTYPE html><p>hello</p>" }
+
+      it "returns NO_QUIRKS" do
+        assert_equal(Nokogiri::HTML5::QuirksMode::NO_QUIRKS, document.quirks_mode)
+      end
+    end
+
+    describe "on a document without a doctype" do
+      let(:html) { "<html><p>hello</p>" }
+
+      it "returns QUIRKS" do
+        assert_equal(Nokogiri::HTML5::QuirksMode::QUIRKS, document.quirks_mode)
+      end
+    end
+  end
+
+  describe "DocumentFragment#quirks_mode" do
+    let(:input)            { "<p><table>" }
+    let(:no_quirks_output) { "<p></p><table></table>" }
+    let(:quirks_output)    { "<p><table></table></p>" }
+
+    describe "without parsing anything" do
+      let(:fragment) { Nokogiri::HTML5::DocumentFragment.new(Nokogiri::HTML5::Document.new) }
+
+      it "returns nil" do
+        assert_nil(fragment.quirks_mode)
+      end
+    end
+
+    describe "in context" do
+      describe "document did not invoke the parser" do
+        let(:document) { Nokogiri::HTML5::Document.new }
+
+        it "parses the fragment in no-quirks mode" do
+          context_node = document.create_element("div")
+          fragment = context_node.fragment(input)
+
+          assert_equal(Nokogiri::HTML5::QuirksMode::NO_QUIRKS, fragment.quirks_mode)
+          assert_equal(no_quirks_output, fragment.to_html)
+        end
+      end
+
+      describe "document has a doctype" do
+        let(:document) { Nokogiri::HTML5::Document.parse("<!DOCTYPE html><div>") }
+
+        it "parses the fragment in no-quirks mode" do
+          context_node = document.at_css("div")
+          fragment = context_node.fragment(input)
+
+          assert_equal(Nokogiri::HTML5::QuirksMode::NO_QUIRKS, fragment.quirks_mode)
+          assert_equal(no_quirks_output, fragment.to_html)
+        end
+      end
+
+      describe "document does not have a doctype" do
+        let(:document) { Nokogiri::HTML5::Document.parse("<div>") }
+
+        it "parses the fragment in quirks mode" do
+          context_node = document.at_css("div")
+          fragment = context_node.fragment(input)
+
+          assert_equal(Nokogiri::HTML5::QuirksMode::QUIRKS, fragment.quirks_mode)
+          assert_equal(quirks_output, fragment.to_html)
+        end
+      end
+    end
+
+    describe "no context" do
+      describe "document did not invoke the parser" do
+        let(:document) { Nokogiri::HTML5::Document.new }
+
+        it "parses the fragment in no-quirks mode" do
+          fragment = document.fragment(input)
+
+          assert_equal(Nokogiri::HTML5::QuirksMode::NO_QUIRKS, fragment.quirks_mode)
+          assert_equal(no_quirks_output, fragment.to_html)
+        end
+      end
+
+      describe "document has a doctype" do
+        let(:document) { Nokogiri::HTML5::Document.parse("<!DOCTYPE html><div>") }
+
+        it "parses the fragment in no-quirks mode" do
+          fragment = document.fragment(input)
+
+          assert_equal(Nokogiri::HTML5::QuirksMode::NO_QUIRKS, fragment.quirks_mode)
+          assert_equal(no_quirks_output, fragment.to_html)
+        end
+      end
+
+      describe "document does not have a doctype" do
+        let(:document) { Nokogiri::HTML5::Document.parse("<div>") }
+
+        it "parses the fragment in no-quirks mode" do
+          fragment = document.fragment(input)
+
+          assert_equal(Nokogiri::HTML5::QuirksMode::NO_QUIRKS, fragment.quirks_mode)
+          assert_equal(no_quirks_output, fragment.to_html)
+        end
+      end
+    end
+  end
+end if Nokogiri.uses_gumbo?