moxml 0.1.19 → 0.1.21

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 76508cc0d2699469ade87f48e38ea73289abc4f95cb7a313f6d71548477248f3
-  data.tar.gz: 1752cb433953a869cedd1d88fc565c53f77bd08201c5977247150e1e7b75a386
+  metadata.gz: cd873f36f1ee8d7799299cedbc4bfa7da00d588e311693be8b6e718b9e09fa8d
+  data.tar.gz: 2597df5af105dfcfdc84586b98473e99af95681fbf1ef24b0c2c6698280e6dbb
 SHA512:
-  metadata.gz: a4219577f2e00a9e4f00f1ce1bbaf4f03d6c841dfe8adb9983162a6389b1e276725984dc010f451cac55fbd8049a81451d641fc073fc1cf8417f90ee6e50cdf5
-  data.tar.gz: 44f3e08b174e2689fed3a60ee3ef0b805fdc14cbeeb905ebde15e5241be3e3f168441dc95afe99dd244fc1047c581fce912ec3e061299e4e101f63329f00def9
+  metadata.gz: b39f087ced4fc9ea76722e32c8b036ee7e4819798bec248af520f51bb20039cb76459837e1b5a65540f46bf7112ee5922eece22cac763ef25b3235a0c5ca60f6
+  data.tar.gz: 75847baca549e9cb0902203f2d8b9f8e4e764c048dc0cd57260d6995f1aa233c5db6ecdb6786344cdfb097b4ec73a088d1c12b4d79b2fcb5217c936ba4605ec1

data/.rubocop_todo.yml

@@ -1,11 +1,65 @@
 # This configuration was generated by
 # `rubocop --auto-gen-config`
-# on 2026-04-23 07:48:23 UTC using RuboCop version 1.86.0.
+# on 2026-05-03 12:53:32 UTC using RuboCop version 1.86.0.
 # 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
 # versions of RuboCop, may require this file to be generated again.
 
+# Offense count: 5
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyle, IndentationWidth.
+# SupportedStyles: with_first_argument, with_fixed_indentation
+Layout/ArgumentAlignment:
+  Exclude:
+    - 'lib/moxml/adapter/base.rb'
+    - 'lib/moxml/adapter/libxml.rb'
+    - 'lib/moxml/builder.rb'
+
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyle, IndentationWidth.
+# SupportedStyles: with_first_element, with_fixed_indentation
+Layout/ArrayAlignment:
+  Exclude:
+    - 'lib/moxml/xpath/compiler.rb'
+
+# Offense count: 9
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyleAlignWith.
+# SupportedStylesAlignWith: either, start_of_block, start_of_line
+Layout/BlockAlignment:
+  Exclude:
+    - 'lib/moxml/adapter/libxml.rb'
+    - 'lib/moxml/adapter/ox.rb'
+    - 'spec/integration/shared_examples/edge_cases.rb'
+    - 'spec/integration/shared_examples/high_level/document_builder_behavior.rb'
+    - 'spec/integration/shared_examples/node_wrappers/entity_reference_behavior.rb'
+    - 'spec/moxml/xpath/functions/node_functions_spec.rb'
+
+# Offense count: 9
+# This cop supports safe autocorrection (--autocorrect).
+Layout/BlockEndNewline:
+  Exclude:
+    - 'lib/moxml/adapter/libxml.rb'
+    - 'lib/moxml/adapter/ox.rb'
+    - 'spec/integration/shared_examples/edge_cases.rb'
+    - 'spec/integration/shared_examples/high_level/document_builder_behavior.rb'
+    - 'spec/integration/shared_examples/node_wrappers/entity_reference_behavior.rb'
+    - 'spec/moxml/xpath/functions/node_functions_spec.rb'
+
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+Layout/ClosingParenthesisIndentation:
+  Exclude:
+    - 'lib/moxml/adapter/oga.rb'
+
+# Offense count: 1
+# This cop supports safe autocorrection (--autocorrect).
+Layout/ElseAlignment:
+  Exclude:
+    - 'lib/moxml/adapter/base.rb'
+
 # Offense count: 4
 # This cop supports safe autocorrection (--autocorrect).
 Layout/EmptyLineAfterGuardClause:
@@ -27,13 +81,67 @@ Layout/EmptyLines:
   Exclude:
     - 'lib/moxml/adapter/ox.rb'
 
-# Offense count: 330
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+Layout/EmptyLinesAroundMethodBody:
+  Exclude:
+    - 'lib/moxml/adapter/ox.rb'
+
+# Offense count: 1
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyleAlignWith.
+# SupportedStylesAlignWith: keyword, variable, start_of_line
+Layout/EndAlignment:
+  Exclude:
+    - 'lib/moxml/adapter/base.rb'
+
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyle, IndentationWidth.
+# SupportedStyles: consistent, consistent_relative_to_receiver, special_for_inner_method_call, special_for_inner_method_call_in_parentheses
+Layout/FirstArgumentIndentation:
+  Exclude:
+    - 'lib/moxml/adapter/oga.rb'
+
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: AllowMultipleStyles, EnforcedHashRocketStyle, EnforcedColonStyle, EnforcedLastArgumentHashStyle.
+# SupportedHashRocketStyles: key, separator, table
+# SupportedColonStyles: key, separator, table
+# SupportedLastArgumentHashStyles: always_inspect, always_ignore, ignore_implicit, ignore_explicit
+Layout/HashAlignment:
+  Exclude:
+    - 'lib/moxml/builder.rb'
+
+# Offense count: 20
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: Width, EnforcedStyleAlignWith, AllowedPatterns.
+# SupportedStylesAlignWith: start_of_line, relative_to_receiver
+Layout/IndentationWidth:
+  Exclude:
+    - 'lib/moxml/adapter/base.rb'
+    - 'lib/moxml/adapter/libxml.rb'
+    - 'lib/moxml/adapter/ox.rb'
+    - 'spec/integration/shared_examples/edge_cases.rb'
+    - 'spec/integration/shared_examples/high_level/document_builder_behavior.rb'
+    - 'spec/integration/shared_examples/node_wrappers/entity_reference_behavior.rb'
+    - 'spec/moxml/xpath/functions/node_functions_spec.rb'
+
+# Offense count: 344
 # This cop supports safe autocorrection (--autocorrect).
 # Configuration parameters: Max, AllowHeredoc, AllowURI, AllowQualifiedName, URISchemes, AllowRBSInlineAnnotation, AllowCopDirectives, AllowedPatterns, SplitStrings.
 # URISchemes: http, https
 Layout/LineLength:
   Enabled: false
 
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyle.
+# SupportedStyles: symmetrical, new_line, same_line
+Layout/MultilineMethodCallBraceLayout:
+  Exclude:
+    - 'lib/moxml/adapter/oga.rb'
+
 # Offense count: 1
 # This cop supports safe autocorrection (--autocorrect).
 # Configuration parameters: EnforcedStyle, IndentationWidth.
@@ -42,6 +150,17 @@ Layout/MultilineOperationIndentation:
   Exclude:
     - 'lib/moxml/adapter/ox.rb'
 
+# Offense count: 10
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: AllowInHeredoc.
+Layout/TrailingWhitespace:
+  Exclude:
+    - 'lib/moxml/adapter/base.rb'
+    - 'lib/moxml/adapter/libxml.rb'
+    - 'lib/moxml/adapter/ox.rb'
+    - 'lib/moxml/builder.rb'
+    - 'lib/moxml/xpath/compiler.rb'
+
 # Offense count: 7
 # Configuration parameters: AllowedMethods.
 # AllowedMethods: enums
@@ -61,11 +180,10 @@ Lint/DuplicateBranch:
     - 'lib/moxml/document.rb'
     - 'lib/moxml/entity_registry.rb'
 
-# Offense count: 5
+# Offense count: 4
 Lint/DuplicateMethods:
   Exclude:
     - 'lib/moxml/config.rb'
-    - 'lib/moxml/element.rb'
     - 'lib/moxml/node.rb'
 
 # Offense count: 4
@@ -91,7 +209,7 @@ Lint/EmptyWhen:
 # Offense count: 3
 Lint/HashCompareByIdentity:
   Exclude:
-    - 'lib/moxml/native_attachment.rb'
+    - 'lib/moxml/native_attachment/native.rb'
 
 # Offense count: 1
 Lint/IneffectiveAccessModifier:
@@ -127,12 +245,12 @@ Metrics/BlockLength:
 Metrics/BlockNesting:
   Max: 4
 
-# Offense count: 76
+# Offense count: 75
 # Configuration parameters: AllowedMethods, AllowedPatterns, Max.
 Metrics/CyclomaticComplexity:
   Enabled: false
 
-# Offense count: 186
+# Offense count: 188
 # Configuration parameters: CountComments, CountAsOne, AllowedMethods, AllowedPatterns.
 Metrics/MethodLength:
   Max: 110
@@ -180,6 +298,12 @@ Naming/VariableNumber:
     - 'spec/moxml/allocation_guard_spec.rb'
     - 'spec/support/allocation_helper.rb'
 
+# Offense count: 1
+# Configuration parameters: MinSize.
+Performance/CollectionLiteralInLoop:
+  Exclude:
+    - 'lib/moxml/xpath/compiler.rb'
+
 # Offense count: 5
 RSpec/BeforeAfterAll:
   Exclude:
@@ -205,12 +329,12 @@ RSpec/ContextWording:
     - 'spec/moxml/xpath/parser_spec.rb'
     - 'spec/performance/benchmark_spec.rb'
 
-# Offense count: 24
+# Offense count: 25
 # Configuration parameters: IgnoredMetadata.
 RSpec/DescribeClass:
   Enabled: false
 
-# Offense count: 295
+# Offense count: 328
 # Configuration parameters: CountAsOne.
 RSpec/ExampleLength:
   Max: 64
@@ -240,13 +364,13 @@ RSpec/LeakyConstantDeclaration:
     - 'spec/moxml/declaration_preservation_spec.rb'
     - 'spec/moxml/sax_spec.rb'
 
-# Offense count: 2
+# Offense count: 4
 # Configuration parameters: .
 # SupportedStyles: have_received, receive
 RSpec/MessageSpies:
   EnforcedStyle: receive
 
-# Offense count: 390
+# Offense count: 414
 RSpec/MultipleExpectations:
   Max: 10
 
@@ -306,6 +430,22 @@ Security/Eval:
   Exclude:
     - 'spec/moxml/xpath/ruby/generator_spec.rb'
 
+# Offense count: 12
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyle, ProceduralMethods, FunctionalMethods, AllowedMethods, AllowedPatterns, AllowBracesOnProceduralOneLiners, BracesRequiredMethods.
+# SupportedStyles: line_count_based, semantic, braces_for_chaining, always_braces
+# ProceduralMethods: benchmark, bm, bmbm, create, each_with_object, measure, new, realtime, tap, with_object
+# FunctionalMethods: let, let!, subject, watch
+# AllowedMethods: lambda, proc, it
+Style/BlockDelimiters:
+  Exclude:
+    - 'lib/moxml/adapter/libxml.rb'
+    - 'lib/moxml/adapter/ox.rb'
+    - 'spec/integration/shared_examples/edge_cases.rb'
+    - 'spec/integration/shared_examples/high_level/document_builder_behavior.rb'
+    - 'spec/integration/shared_examples/node_wrappers/entity_reference_behavior.rb'
+    - 'spec/moxml/xpath/functions/node_functions_spec.rb'
+
 # Offense count: 1
 Style/DocumentDynamicEvalDefinition:
   Exclude:
@@ -329,6 +469,18 @@ Style/MissingRespondToMissing:
   Exclude:
     - 'lib/moxml/xpath/ruby/node.rb'
 
+# Offense count: 1
+# This cop supports safe autocorrection (--autocorrect).
+Style/MultilineIfModifier:
+  Exclude:
+    - 'lib/moxml/builder.rb'
+
+# Offense count: 1
+# This cop supports safe autocorrection (--autocorrect).
+Style/MultilineTernaryOperator:
+  Exclude:
+    - 'lib/moxml/adapter/base.rb'
+
 # Offense count: 1
 # This cop supports safe autocorrection (--autocorrect).
 # Configuration parameters: AllowMethodComparison, ComparisonsThreshold.
@@ -365,7 +517,25 @@ Style/RedundantConstantBase:
     - 'spec/moxml/adapter/headed_ox_spec.rb'
 
 # Offense count: 1
+# This cop supports safe autocorrection (--autocorrect).
+Style/RedundantParentheses:
+  Exclude:
+    - 'lib/moxml/xpath/compiler.rb'
+
+# Offense count: 8
 # This cop supports unsafe autocorrection (--autocorrect-all).
 Style/SelectByKind:
   Exclude:
+    - 'spec/integration/shared_examples/edge_cases.rb'
+    - 'spec/integration/shared_examples/entity_reference_whitespace.rb'
     - 'spec/integration/shared_examples/node_wrappers/entity_reference_behavior.rb'
+    - 'spec/integration/shared_examples/node_wrappers/node_set_behavior.rb'
+    - 'spec/moxml/xpath/functions/node_functions_spec.rb'
+
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: EnforcedStyle, MinSize.
+# SupportedStyles: percent, brackets
+Style/SymbolArray:
+  Exclude:
+    - 'lib/moxml/xpath/compiler.rb'

data/docs/_guides/node-api-consistency.adoc

@@ -390,6 +390,10 @@ Text nodes contain character data.
 | `#text`
 | Alias for #content
 | ✅ Yes
+
+| `#to_s`
+| Returns the text content (same as `#content`)
+| ✅ Yes
 |===
 
 ==== Identity Methods

data/lib/moxml/adapter/base.rb

@@ -35,12 +35,22 @@ module Moxml
                   # not valid UTF-8, fall back to encoding as UTF-8 with
                   # replacement to avoid raising on gsub.
                   dup = xml.dup.force_encoding("UTF-8")
-                  dup.valid_encoding? ? dup : xml.dup.encode("UTF-8", "ASCII-8BIT", invalid: :replace, undef: :replace)
+                  if dup.valid_encoding?
+                    dup
+                  else
+                    xml.dup.encode("UTF-8",
+                                   "ASCII-8BIT", invalid: :replace, undef: :replace)
+                  end
                 elsif xml.encoding == Encoding::UTF_8
                   xml
                 else
                   xml.encode("UTF-8")
                 end
+          # Fast path: no `&` means no entity references to mark — skip
+          # the regex scan and string allocation entirely. The vast
+          # majority of XML payloads contain no entity references.
+          return str unless str.include?("&")
+
           str.gsub(ENTITY_NAME_RE) do |match|
             STANDARD_ENTITIES.include?(::Regexp.last_match(1)) ? match : "#{ENTITY_MARKER}#{::Regexp.last_match(1)};"
           end

data/lib/moxml/adapter/customized_libxml/node.rb

@@ -8,6 +8,9 @@ module Moxml
       # This wrapper hides LibXML's strict document ownership model,
       # allowing nodes to be moved between documents transparently.
       # Similar pattern to Ox adapter's customized classes.
+      #
+      # The Libxml adapter owns wrapper type mapping in one place so the
+      # wrapper classes do not duplicate node-type knowledge.
       class Node
         attr_reader :native
 

data/lib/moxml/adapter/customized_libxml/text.rb

@@ -19,7 +19,12 @@ module Moxml
         # LibXML's .content already contains escaped text, but it over-escapes
         # quotes which don't need escaping in text nodes (only in attributes)
         def to_xml
-          @native.content.gsub(""", '"')
+          content = @native.content
+          # Skip the gsub allocation entirely when there's nothing to undo —
+          # the common case for parsed text without literal quotes.
+          return content unless content.include?(""")
+
+          content.gsub(""", '"')
         end
       end
     end

data/lib/moxml/adapter/customized_ox/text.rb

@@ -3,8 +3,21 @@
 module Moxml
   module Adapter
     module CustomizedOx
-      # Ox uses Strings, but a string cannot have a parent reference
-      class Text < ::Ox::Node; end
+      # Ox uses Strings for text content, but a String cannot carry a @parent
+      # back-reference. We subclass ::Ox::Node so a Text wrapper can hold one.
+      #
+      # ::Ox::Node subclasses that are neither ::Ox::Element nor ::Ox::Document
+      # are unknown to Ox.dump's standard XML emitter, so they fall through to
+      # Ox's generic object-marshalling format. The serializer in
+      # Moxml::Adapter::Ox#serialize special-cases this class to emit the value
+      # with proper XML escaping. The #to_s override ensures string
+      # interpolation (`"#{text}"`) produces the text content rather than the
+      # default Object representation.
+      class Text < ::Ox::Node
+        def to_s
+          value.to_s
+        end
+      end
     end
   end
 end

data/lib/moxml/adapter/customized_rexml/formatter.rb

@@ -68,6 +68,7 @@ module Moxml
                     next if child.is_a?(::REXML::Text) &&
                       child.to_s.strip.empty? &&
                       !(child.next_sibling.nil? && child.previous_sibling.nil?)
+
                     write(child, output)
                   end
                 when :eref

data/lib/moxml/adapter/libxml/entity_ref_registry.rb

@@ -0,0 +1,105 @@
+# frozen_string_literal: true
+
+module Moxml
+  module Adapter
+    class Libxml < Base
+      # Tracks entity-reference insertions that cannot live in LibXML's native
+      # node tree, plus the child sequence needed to serialize them in order.
+      class EntityRefRegistry
+        ENTITY_REFS_KEY = :_entity_ref_pairs
+        CHILD_SEQUENCE_KEY = :_child_seq_pairs
+        NON_WHITESPACE_RE = /\S/
+        private_constant :ENTITY_REFS_KEY, :CHILD_SEQUENCE_KEY, :NON_WHITESPACE_RE
+
+        def initialize(attachments, doc)
+          @attachments = attachments
+          @doc = doc
+        end
+
+        def active?
+          @doc ? @attachments.key?(@doc, ENTITY_REFS_KEY) : false
+        end
+
+        def register(element, ref)
+          return unless @doc && element
+
+          path = path_for(element)
+
+          refs_by_path = @attachments.get(@doc, ENTITY_REFS_KEY) || {}
+          (refs_by_path[path] ||= []) << ref
+          @attachments.set(@doc, ENTITY_REFS_KEY, refs_by_path)
+
+          seq_by_path = @attachments.get(@doc, CHILD_SEQUENCE_KEY) || {}
+          existing = seq_by_path[path]
+          if existing
+            existing << :eref
+          else
+            seq_by_path[path] = Array.new(count_native_children(element), :native)
+            seq_by_path[path] << :eref
+            @attachments.set(@doc, CHILD_SEQUENCE_KEY, seq_by_path)
+          end
+        end
+
+        def append_native(element)
+          return unless @doc && element
+
+          seq_by_path = @attachments.get(@doc, CHILD_SEQUENCE_KEY)
+          return unless seq_by_path
+
+          seq = seq_by_path[path_for(element)]
+          return unless seq
+
+          seq << :native
+        end
+
+        def refs_for(element)
+          return nil unless @doc && element
+
+          refs_by_path = @attachments.get(@doc, ENTITY_REFS_KEY)
+          refs_by_path && refs_by_path[path_for(element)]
+        end
+
+        def sequence_for(element)
+          return nil unless @doc && element
+
+          seq_by_path = @attachments.get(@doc, CHILD_SEQUENCE_KEY)
+          seq_by_path && seq_by_path[path_for(element)]
+        end
+
+        def serialization_for(element)
+          refs = refs_for(element)
+          return [nil, nil] unless refs && !refs.empty?
+
+          seq = sequence_for(element)
+          return [nil, nil] unless seq
+
+          [refs, seq]
+        end
+
+        private
+
+        def path_for(element)
+          element.path
+        end
+
+        def count_native_children(element)
+          return 0 unless element.is_a?(::LibXML::XML::Node) && element.children?
+
+          count = 0
+          element.each_child do |child|
+            count += 1 unless blank_text_node?(child)
+          end
+          count
+        end
+
+        def blank_text_node?(child)
+          child.text? && blank_content?(child.content)
+        end
+
+        def blank_content?(content)
+          content.nil? || !content.match?(NON_WHITESPACE_RE)
+        end
+      end
+    end
+  end
+end

data/lib/moxml/adapter/libxml/entity_restorer.rb

@@ -0,0 +1,92 @@
+# frozen_string_literal: true
+
+module Moxml
+  module Adapter
+    class Libxml < Base
+      # Restores configured character entities into explicit Moxml
+      # EntityReference nodes after LibXML has parsed the native tree.
+      class EntityRestorer
+        def initialize(doc, adapter: Libxml)
+          @doc = doc
+          @ctx = doc.context
+          @registry = @ctx.entity_registry
+          @config = @ctx.config
+          @adapter = adapter
+        end
+
+        def run
+          return unless @registry && @doc.root
+
+          walk(@doc.root)
+        end
+
+        private
+
+        def walk(element)
+          # Snapshot because we may add/remove siblings during the walk.
+          element.children.to_a.each do |child|
+            if child.is_a?(::Moxml::Text)
+              restore_text_node(child)
+            elsif child.is_a?(::Moxml::Element)
+              walk(child)
+            end
+          end
+        end
+
+        # Matches DocumentBuilder's previous behavior, including the libxml
+        # limitation that adjacent native text nodes get merged.
+        def restore_text_node(text_node)
+          content = text_node.content
+          return unless content
+
+          chunks = chunk_text(content)
+          return if chunks.size == 1 && chunks.first.first == :text
+
+          parent = text_node.parent
+          return unless parent
+
+          text_node.remove
+          chunks.each { |type, payload| append_chunk(parent, type, payload) }
+        end
+
+        def chunk_text(content)
+          chunks = []
+          buffer = +""
+          restorable = @registry.restorable_codepoints
+
+          content.each_char do |char|
+            cp = char.ord
+            if restorable.include?(cp) &&
+                (name = @registry.primary_name_for_codepoint(cp)) &&
+                @registry.should_restore?(cp, config: @config)
+              unless buffer.empty?
+                chunks << [:text, buffer.dup]
+                buffer.clear
+              end
+              chunks << [:eref, name]
+            else
+              buffer << char
+            end
+          end
+
+          chunks << [:text, buffer.dup] unless buffer.empty?
+          chunks
+        end
+
+        def append_chunk(parent, type, payload)
+          case type
+          when :text
+            parent.add_child(::Moxml::Text.new(@adapter.create_native_text(payload), @ctx))
+          when :eref
+            parent.add_child(
+              ::Moxml::EntityReference.new(
+                @adapter.create_native_entity_reference(payload),
+                @ctx,
+              ),
+            )
+          end
+        end
+      end
+    end
+  end
+end