Merge pull request #2527 from sparklemotion/2525-update-libxml-2_9_14-v1_13_x

dep: update libxml2 to v2.9.14 (v1.13.x branch)

Author: flavorjones

.rubocop_todo.yml

@@ -1,6 +1,6 @@
 # This configuration was generated by
 # `rubocop --auto-gen-config`
-# on 2022-01-03 20:31:05 UTC using RuboCop version 1.24.1.
+# on 2022-05-04 15:17:01 UTC using RuboCop version 1.28.2.
 # The point is for the user to remove these configuration records
 # one by one as the offenses are removed from the code base.
 # Note that changes in the inspected code, or installation of new
@@ -22,16 +22,21 @@ Lint/MissingSuper:
     - 'lib/nokogiri/xml/document_fragment.rb'
     - 'lib/nokogiri/xml/processing_instruction.rb'
 
-# Offense count: 4
-# Cop supports --auto-correct.
+# Offense count: 1
+# This cop supports safe auto-correction (--auto-correct).
+Lint/RedundantCopDisableDirective:
+  Exclude:
+    - 'lib/nokogiri/xml/processing_instruction.rb'
+
+# Offense count: 2
+# This cop supports safe auto-correction (--auto-correct).
 # Configuration parameters: ContextCreatingMethods, MethodCreatingMethods.
 Lint/UselessAccessModifier:
   Exclude:
-    - 'lib/nokogiri/css/xpath_visitor.rb'
     - 'lib/nokogiri/html5.rb'
     - 'lib/nokogiri/html5/document.rb'
 
-# Offense count: 18
+# Offense count: 17
 # Configuration parameters: CountBlocks.
 Metrics/BlockNesting:
   Max: 5
@@ -50,8 +55,8 @@ Performance/CollectionLiteralInLoop:
     - 'test/xml/test_dtd_encoding.rb'
     - 'test/xml/test_node_reparenting.rb'
 
-# Offense count: 11
-# Cop supports --auto-correct.
+# Offense count: 12
+# This cop supports safe auto-correction (--auto-correct).
 Performance/StringIdentifierArgument:
   Exclude:
     - 'lib/nokogiri/html5.rb'
@@ -60,20 +65,14 @@ Performance/StringIdentifierArgument:
     - 'test/test_memory_leak.rb'
     - 'test/xml/test_builder.rb'
 
-# Offense count: 1
-# Cop supports --auto-correct.
-Performance/StringInclude:
-  Exclude:
-    - 'lib/nokogiri/css/xpath_visitor.rb'
-
 # Offense count: 2
-# Cop supports --auto-correct.
+# This cop supports safe auto-correction (--auto-correct).
 Performance/TimesMap:
   Exclude:
     - 'test/html5/test_nokogumbo.rb'
 
 # Offense count: 6
-# Cop supports --auto-correct.
+# This cop supports safe auto-correction (--auto-correct).
 # Configuration parameters: EnforcedStyle.
 # SupportedStyles: nested, compact
 Style/ClassAndModuleChildren:
@@ -85,14 +84,26 @@ Style/ClassAndModuleChildren:
     - 'test/xml/sax/test_parser.rb'
     - 'test/xml/sax/test_push_parser.rb'
 
+# Offense count: 17
+# This cop supports safe auto-correction (--auto-correct).
+# Configuration parameters: AllowedVars.
+Style/FetchEnvVar:
+  Exclude:
+    - 'ext/nokogiri/extconf.rb'
+    - 'rakelib/debug.rake'
+    - 'rakelib/extensions.rake'
+    - 'rakelib/rdoc.rake'
+    - 'test/helper.rb'
+
 # Offense count: 2
 Style/MissingRespondToMissing:
   Exclude:
     - 'lib/nokogiri/xml/builder.rb'
 
-# Offense count: 78
-# Cop supports --auto-correct.
-# Configuration parameters: AllowHeredoc, AllowURI, URISchemes, IgnoreCopDirectives, IgnoredPatterns.
+# Offense count: 81
+# This cop supports safe auto-correction (--auto-correct).
+# Configuration parameters: AllowHeredoc, AllowURI, URISchemes, IgnoreCopDirectives, AllowedPatterns, IgnoredPatterns.
 # URISchemes: http, https
+# AllowedPatterns: \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z), \A\s*(remote_)?test(_\w+)?\s.*(do|->)(\s|\Z)
 Layout/LineLength:
   Max: 250

dependencies.yml

@@ -1,7 +1,7 @@
 libxml2:
-  version: "2.9.13"
-  sha256: "276130602d12fe484ecc03447ee5e759d0465558fbc9d6bd144e3745306ebf0e"
-  # sha-256 hash provided in https://download.gnome.org/sources/libxml2/2.9/libxml2-2.9.13.sha256sum
+  version: "2.9.14"
+  sha256: "60d74a257d1ccec0475e749cba2f21559e48139efba6ff28224357c7c798dfee"
+  # sha-256 hash provided in https://download.gnome.org/sources/libxml2/2.9/libxml2-2.9.14.sha256sum
 
 libxslt:
   version: "1.1.35"

lib/nokogiri/xml/node.rb

@@ -123,8 +123,8 @@ class Node
       # [Yields] Nokogiri::XML::Node
       # [Returns] Nokogiri::XML::Node
       #
-      def initialize(name, document) # rubocop:disable Style/RedundantInitialize
-        # This is intentionally empty.
+      def initialize(name, document)
+        # This is intentionally empty, and sets the method signature for subclasses.
       end
 
       ###

nokogiri.gemspec

@@ -330,7 +330,7 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency("rake-compiler-dock", "~> 1.2")
   spec.add_development_dependency("rdoc", "~> 6.3")
   spec.add_development_dependency("rexical", "~> 1.0.7")
-  spec.add_development_dependency("rubocop", "~> 1.23")
+  spec.add_development_dependency("rubocop", "~> 1.28", ">= 1.28.2")
   spec.add_development_dependency("rubocop-minitest", "~> 0.17")
   spec.add_development_dependency("rubocop-performance", "~> 1.12")
   spec.add_development_dependency("rubocop-rake", "~> 0.6")

patches/libxml2/0010-Revert-Different-approach-to-fix-quadratic-behavior.patch

@@ -373,6 +373,7 @@ if java?
   end
 else
   require "rake/extensiontask"
+  require "yaml"
 
   dependencies = YAML.load_file("dependencies.yml")
 

rakelib/extensions.rake

@@ -173,7 +173,7 @@ class TestComment < Nokogiri::TestCase
         let(:body) { doc.at_css("body") }
         let(:subject) { doc.at_css("div#under-test") }
 
-        if Nokogiri.uses_libxml?
+        if Nokogiri.uses_libxml?("<=2.9.13")
           it "ignores up to the next '>'" do # NON-COMPLIANT
             assert_equal 2, body.children.length
             assert_equal body.children[0], subject
@@ -183,10 +183,33 @@ class TestComment < Nokogiri::TestCase
             assert_predicate body.children[1], :text?
             assert_equal "-->hello", body.children[1].content
           end
+        elsif Nokogiri.uses_libxml?
+          it "parses as pcdata" do # NON-COMPLIANT
+            assert_equal 1, body.children.length
+            assert_equal subject, body.children.first
+
+            assert_equal 3, subject.children.length
+            subject.children[0].tap do |child|
+              assert_predicate(child, :text?)
+              assert_equal("<! comment ", child.content)
+            end
+            subject.children[1].tap do |child|
+              assert_predicate(child, :element?)
+              assert_equal("div", child.name)
+              assert_equal("inner content", child.content)
+            end
+            subject.children[2].tap do |child|
+              assert_predicate(child, :text?)
+              assert_equal("-->hello", child.content)
+            end
+          end
         end
 
         if Nokogiri.jruby?
           it "ignores up to the next '-->'" do # NON-COMPLIANT
+            assert_equal 1, body.children.length
+            assert_equal subject, body.children.first
+
             assert_equal 1, subject.children.length
             assert_predicate subject.children[0], :text?
             assert_equal "hello", subject.children[0].content

test/html4/test_comments.rb

@@ -801,18 +801,25 @@ def test_leaking_dtd_nodes_after_internal_subset_removal
 
             it "skips to the next start tag" do
               # see https://github.com/sparklemotion/nokogiri/issues/2461 for why we're testing this edge case
-              if Nokogiri.uses_libxml?(">= 2.9.13")
-                skip_unless_libxml2_patch("0010-Revert-Different-approach-to-fix-quadratic-behavior.patch")
-              end
-
               doc = Nokogiri::HTML4.parse(input)
               body = doc.at_xpath("//body")
 
-              expected_error_snippet = Nokogiri.uses_libxml? ? "invalid element name" : "Missing start element name"
-              assert_includes(doc.errors.first.to_s, expected_error_snippet)
-
-              assert_equal("this < that", body.children.first.text, body.to_html)
-              assert_equal(["div", "div"], body.children.map(&:name), body.to_html)
+              if Nokogiri.uses_libxml?("= 2.9.13")
+                # <body><div>this <div>second element</div></div></body>
+                assert_equal(1, body.children.length)
+                body.children.first.tap do |div|
+                  assert_equal(2, div.children.length)
+                  assert_equal("this ", div.children[0].content)
+                  assert_equal("div", div.children[1].name)
+                  assert_equal("second element", div.children[1].content)
+                end
+              else
+                # <body><div>this &lt; that</div><div>second element</div></body>
+                assert_equal(2, body.children.length)
+                assert_equal(["div", "div"], body.children.map(&:name), body.to_html)
+                assert_equal("this < that", body.children[0].text, body.to_html)
+                assert_equal("second element", body.children[1].text, body.to_html)
+              end
             end
           end
 

test/html4/test_document.rb

@@ -60,13 +60,7 @@ def test_version_info_for_libxml
     major = Regexp.last_match(1).to_i
     minor = Regexp.last_match(2).to_i
     bug = Regexp.last_match(3).to_i
-    if RbConfig::CONFIG["target_os"].include?("freebsd13") && Nokogiri::VersionInfo.instance.libxml2_using_system?
-      pending("https://github.com/sparklemotion/nokogiri/issues/2506") do
-        assert_equal("#{major}.#{minor}.#{bug}", Nokogiri::VERSION_INFO["libxml"]["loaded"])
-      end
-    else
-      assert_equal("#{major}.#{minor}.#{bug}", Nokogiri::VERSION_INFO["libxml"]["loaded"])
-    end
+    assert_equal("#{major}.#{minor}.#{bug}", Nokogiri::VERSION_INFO["libxml"]["loaded"])
 
     assert(version_info["libxml"].key?("iconv_enabled"))
     assert(version_info["libxslt"].key?("datetime_enabled"))