canon 0.2.9 → 0.2.12

data/lib/canon/tree_diff.rb

@@ -1,33 +1,32 @@
 # frozen_string_literal: true
 
 module Canon
+  # Semantic tree-diff algorithm — distinct from the DOM positional
+  # diff in {Canon::Comparison}.
+  #
+  # This module computes signature-based tree matches and produces
+  # INSERT/DELETE/UPDATE/MOVE operations. Sub-namespaces:
+  #
+  # * Core — TreeNode, Matching, NodeSignature, NodeWeight,
+  #   AttributeComparator, XmlEntityDecoder
+  # * Matchers — HashMatcher, SimilarityMatcher, StructuralPropagator,
+  #   UniversalMatcher
+  # * Operations — Operation, OperationDetector
+  # * Adapters — format-specific tree adapters (XML, JSON, HTML, YAML)
+  # * OperationConverterHelpers — MetadataEnricher, ReasonBuilder,
+  #   PostProcessor, UpdateChangeHandler
+  #
+  # Top-level entry points: OperationConverter and TreeDiffIntegrator.
+  #
+  # All children are autoloaded — never `require_relative` them.
   module TreeDiff
-    # Tree diff module for semantic object tree diffing
+    autoload :Adapters, "canon/tree_diff/adapters"
+    autoload :Core, "canon/tree_diff/core"
+    autoload :Matchers, "canon/tree_diff/matchers"
+    autoload :OperationConverter, "canon/tree_diff/operation_converter"
+    autoload :OperationConverterHelpers,
+             "canon/tree_diff/operation_converter_helpers"
+    autoload :Operations, "canon/tree_diff/operations"
+    autoload :TreeDiffIntegrator, "canon/tree_diff/tree_diff_integrator"
   end
 end
-
-# Load core components
-require_relative "tree_diff/core/tree_node"
-require_relative "tree_diff/core/node_signature"
-require_relative "tree_diff/core/node_weight"
-require_relative "tree_diff/core/matching"
-
-# Load matchers
-require_relative "tree_diff/matchers/hash_matcher"
-require_relative "tree_diff/matchers/similarity_matcher"
-require_relative "tree_diff/matchers/structural_propagator"
-require_relative "tree_diff/matchers/universal_matcher"
-
-# Load operations
-require_relative "tree_diff/operations/operation"
-require_relative "tree_diff/operations/operation_detector"
-require_relative "tree_diff/operation_converter"
-
-# Load adapters
-require_relative "tree_diff/adapters/xml_adapter"
-require_relative "tree_diff/adapters/json_adapter"
-require_relative "tree_diff/adapters/html_adapter"
-require_relative "tree_diff/adapters/yaml_adapter"
-
-# Load integrator
-require_relative "tree_diff/tree_diff_integrator"

data/lib/canon/validators/base_validator.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../errors"
-
 module Canon
   module Validators
     # Base class for all input validators
@@ -29,17 +27,14 @@ module Canon
         line = nil
         column = nil
 
-        # Try to extract line/column from error message
-        if error.respond_to?(:line)
+        if error.is_a?(Nokogiri::XML::SyntaxError)
           line = error.line
+          column = error.column
         elsif error.message =~ /line[:\s]+(\d+)/i
           line = ::Regexp.last_match(1).to_i
-        end
-
-        if error.respond_to?(:column)
-          column = error.column
-        elsif error.message =~ /column[:\s]+(\d+)/i
-          column = ::Regexp.last_match(1).to_i
+          if error.message =~ /column[:\s]+(\d+)/i
+            column = ::Regexp.last_match(1).to_i
+          end
         end
 
         { line: line, column: column }

data/lib/canon/validators/html_validator.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
-require_relative "base_validator"
 
 module Canon
   module Validators
@@ -94,13 +93,8 @@ module Canon
       #
       # @param error [Nokogiri::XML::SyntaxError] The syntax error
       # @return [String, nil] Additional details about the error
-      def self.extract_details(error)
-        return nil unless error.respond_to?(:errors)
-
-        details = error.errors.map(&:message).reject do |msg|
-          msg == error.message
-        end
-        details.join("; ") unless details.empty?
+      def self.extract_details(_error)
+        nil
       end
 
       # Build error details from multiple errors

data/lib/canon/validators/json_validator.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "json"
-require_relative "base_validator"
 
 module Canon
   module Validators

data/lib/canon/validators/xml_validator.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
-require_relative "base_validator"
 
 module Canon
   module Validators
@@ -38,13 +37,8 @@ module Canon
       #
       # @param error [Nokogiri::XML::SyntaxError] The syntax error
       # @return [String, nil] Additional details about the error
-      def self.extract_details(error)
-        return nil unless error.respond_to?(:errors)
-
-        details = error.errors.map(&:message).reject do |msg|
-          msg == error.message
-        end
-        details.join("; ") unless details.empty?
+      def self.extract_details(_error)
+        nil
       end
 
       private_class_method :extract_details

data/lib/canon/validators/yaml_validator.rb

@@ -3,7 +3,6 @@
 require "yaml"
 require "date"
 require "time"
-require_relative "base_validator"
 
 module Canon
   module Validators

data/lib/canon/validators.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+module Canon
+  # Format-specific validators that raise {Error} on invalid input.
+  module Validators
+    autoload :BaseValidator, "canon/validators/base_validator"
+    autoload :HtmlValidator, "canon/validators/html_validator"
+    autoload :JsonValidator, "canon/validators/json_validator"
+    autoload :XmlValidator, "canon/validators/xml_validator"
+    autoload :YamlValidator, "canon/validators/yaml_validator"
+  end
+end

data/lib/canon/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Canon
-  VERSION = "0.2.9"
+  VERSION = "0.2.12"
 end

data/lib/canon/xml/c14n.rb

@@ -1,9 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "data_model"
-require_relative "processor"
-require_relative "xpath_engine"
-
 module Canon
   module Xml
     # XML Canonicalization 1.1 implementation

data/lib/canon/xml/data_model.rb

@@ -2,16 +2,6 @@
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
 require "set"
-require_relative "../data_model"
-require_relative "../xml_backend"
-require_relative "../xml_parsing"
-require_relative "nodes/root_node"
-require_relative "nodes/element_node"
-require_relative "nodes/namespace_node"
-require_relative "nodes/attribute_node"
-require_relative "nodes/text_node"
-require_relative "nodes/comment_node"
-require_relative "nodes/processing_instruction_node"
 
 module Canon
   module Xml
@@ -139,7 +129,7 @@ module Canon
       def self.build_from_nokogiri(nokogiri_doc, preserve_whitespace: false)
         root = Nodes::RootNode.new
 
-        if nokogiri_doc.respond_to?(:root) && nokogiri_doc.root
+        if nokogiri_doc.is_a?(Nokogiri::XML::Document) && nokogiri_doc.root
           root.add_child(build_element_node(nokogiri_doc.root,
                                             preserve_whitespace: preserve_whitespace))
           nokogiri_doc.children.each do |child|
@@ -275,7 +265,7 @@ preserve_whitespace: false)
       def self.build_from_moxml(moxml_doc, preserve_whitespace: false)
         root = Nodes::RootNode.new
 
-        if moxml_doc.respond_to?(:root) && moxml_doc.root
+        if moxml_doc.is_a?(Moxml::Document) && moxml_doc.root
           root.add_child(build_moxml_element_node(moxml_doc.root,
                                                   preserve_whitespace: preserve_whitespace))
         end
@@ -327,7 +317,7 @@ preserve_whitespace: false)
           element.add_namespace(ns_node)
         end
 
-        unless element.namespaces.any? do |n|
+        unless element.namespace_nodes.any? do |n|
           n.prefix == "xml"
         end
           element.add_namespace(Nodes::NamespaceNode.new(
@@ -348,7 +338,7 @@ preserve_whitespace: false)
       end
 
       def self.build_moxml_text_node(moxml_text, preserve_whitespace: false)
-        content = moxml_text.text
+        content = moxml_text.content
 
         if !preserve_whitespace && content.strip.empty? && moxml_text.parent.is_a?(Moxml::Element)
           return nil
@@ -358,7 +348,7 @@ preserve_whitespace: false)
       end
 
       def self.build_moxml_comment_node(moxml_comment)
-        Nodes::CommentNode.new(value: moxml_comment.text)
+        Nodes::CommentNode.new(value: moxml_comment.content)
       end
 
       def self.build_moxml_pi_node(moxml_pi)

data/lib/canon/xml/line_range_mapper.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../pretty_printer/xml"
-
 module Canon
   module Xml
     # Maps DOM elements to line ranges in pretty-printed XML

data/lib/canon/xml/nodes/attribute_node.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../node"
-
 module Canon
   module Xml
     module Nodes

data/lib/canon/xml/nodes/comment_node.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../node"
-
 module Canon
   module Xml
     module Nodes

data/lib/canon/xml/nodes/element_node.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../node"
-
 module Canon
   module Xml
     module Nodes

data/lib/canon/xml/nodes/namespace_node.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../node"
-
 module Canon
   module Xml
     module Nodes

data/lib/canon/xml/nodes/processing_instruction_node.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../node"
-
 module Canon
   module Xml
     module Nodes

data/lib/canon/xml/nodes/root_node.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../node"
-
 module Canon
   module Xml
     module Nodes

data/lib/canon/xml/nodes/text_node.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../node"
-
 module Canon
   module Xml
     module Nodes

data/lib/canon/xml/nodes.rb

@@ -0,0 +1,19 @@
+# frozen_string_literal: true
+
+module Canon
+  module Xml
+    # XPath data model node types. All nodes inherit from
+    # {Canon::Xml::Node}. Children are autoloaded — never
+    # `require_relative` them.
+    module Nodes
+      autoload :AttributeNode, "canon/xml/nodes/attribute_node"
+      autoload :CommentNode, "canon/xml/nodes/comment_node"
+      autoload :ElementNode, "canon/xml/nodes/element_node"
+      autoload :NamespaceNode, "canon/xml/nodes/namespace_node"
+      autoload :ProcessingInstructionNode,
+               "canon/xml/nodes/processing_instruction_node"
+      autoload :RootNode, "canon/xml/nodes/root_node"
+      autoload :TextNode, "canon/xml/nodes/text_node"
+    end
+  end
+end

data/lib/canon/xml/processor.rb

@@ -1,10 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "character_encoder"
-require_relative "namespace_handler"
-require_relative "attribute_handler"
-require_relative "xml_base_handler"
-
 module Canon
   module Xml
     # C14N 1.1 processor

data/lib/canon/xml/sax_builder.rb

@@ -1,13 +1,6 @@
 # frozen_string_literal: true
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
-require_relative "nodes/root_node"
-require_relative "nodes/element_node"
-require_relative "nodes/namespace_node"
-require_relative "nodes/attribute_node"
-require_relative "nodes/text_node"
-require_relative "nodes/comment_node"
-require_relative "nodes/processing_instruction_node"
 
 module Canon
   module Xml
@@ -28,7 +21,7 @@ module Canon
     # For C14N, use strip_doctype: true to avoid DTD default attribute expansion:
     #   root = SaxBuilder.parse(xml_string, strip_doctype: true)
     #
-    class SaxBuilder < Nokogiri::XML::SAX::Document
+    class SaxBuilder < (RUBY_ENGINE == "opal" ? Object : Nokogiri::XML::SAX::Document)
       # Parse XML string and return Canon::Xml::Node tree
       #
       # @param xml_string [String] XML content to parse

data/lib/canon/xml/whitespace_normalizer.rb

@@ -43,9 +43,9 @@ module Canon
       # @param node [Moxml::Node] Node to check
       # @return [Boolean] true if node is whitespace-only and should be ignored
       def inter_element_whitespace?(node)
-        return false unless node.respond_to?(:text?) && node.text?
+        return false unless node.is_a?(Nokogiri::XML::Text) || node.is_a?(Moxml::Text)
 
-        text = node.respond_to?(:content) ? node.content.to_s : node.text.to_s
+        text = node.is_a?(Moxml::Text) ? node.content.to_s : node.content.to_s
         text.strip.empty?
       end
 

data/lib/canon/xml.rb

@@ -0,0 +1,33 @@
+# frozen_string_literal: true
+
+module Canon
+  # Native XML data model, parsing, and DOM utilities.
+  #
+  # This namespace holds Canon's own XML representation (independent of
+  # Nokogiri/Moxml), including:
+  # * the XPath data model nodes (Canon::Xml::Node and Canon::Xml::Nodes::*)
+  # * the SAX builder that produces them
+  # * element matching, line-range mapping, xpath, C14N, processors
+  # * canonicalization (C14n) and serialization helpers
+  #
+  # All children are autoloaded from this file. The nested Nodes namespace
+  # is itself a sibling and is autoloaded on first reference to
+  # Canon::Xml::Nodes.
+  module Xml
+    autoload :AttributeHandler, "canon/xml/attribute_handler"
+    autoload :C14n, "canon/xml/c14n"
+    autoload :CharacterEncoder, "canon/xml/character_encoder"
+    autoload :DataModel, "canon/xml/data_model"
+    autoload :ElementMatcher, "canon/xml/element_matcher"
+    autoload :LineRangeMapper, "canon/xml/line_range_mapper"
+    autoload :NamespaceHandler, "canon/xml/namespace_handler"
+    autoload :NamespaceHelper, "canon/xml/namespace_helper"
+    autoload :Node, "canon/xml/node"
+    autoload :Nodes, "canon/xml/nodes"
+    autoload :Processor, "canon/xml/processor"
+    autoload :SaxBuilder, "canon/xml/sax_builder"
+    autoload :WhitespaceNormalizer, "canon/xml/whitespace_normalizer"
+    autoload :XmlBaseHandler, "canon/xml/xml_base_handler"
+    autoload :XPathEngine, "canon/xml/xpath_engine"
+  end
+end

data/lib/canon/xml_backend.rb

@@ -1,20 +1,6 @@
 # frozen_string_literal: true
 
 module Canon
-  # Centralized XML backend detection for Canon.
-  #
-  # Canon supports two XML backends:
-  # - :nokogiri — MRI with Nokogiri installed (default, existing code path)
-  # - :moxml    — Opal runtime or MRI without Nokogiri (uses Oga via moxml)
-  #
-  # The active backend is determined once at load time and cached.
-  # All XML-related code should check `Canon::XmlBackend.moxml?` or
-  # `Canon::XmlBackend.nokogiri?` to select the appropriate code path.
-  #
-  # This module intentionally does NOT wrap Nokogiri through moxml.
-  # Each backend path is independent — the Nokogiri path is the existing
-  # battle-tested code; the moxml path is a parallel implementation for
-  # environments where Nokogiri is unavailable.
   module XmlBackend
     class << self
       def active
@@ -33,6 +19,56 @@ module Canon
         @active = nil
       end
 
+      # Whether the node is a document fragment (any variant).
+      def document_fragment?(node)
+        if nokogiri?
+          node.is_a?(Nokogiri::XML::DocumentFragment) ||
+            node.is_a?(Nokogiri::HTML4::DocumentFragment) ||
+            node.is_a?(Nokogiri::HTML5::DocumentFragment)
+        else
+          false
+        end
+      end
+
+      # Whether the node is an HTML document (any variant).
+      def html_document?(node)
+        if nokogiri?
+          node.is_a?(Nokogiri::HTML::Document) ||
+            node.is_a?(Nokogiri::HTML4::Document) ||
+            node.is_a?(Nokogiri::HTML5::Document)
+        else
+          false
+        end
+      end
+
+      # Detect HTML version from a Nokogiri node.
+      # Returns :html5 or :html4. Defaults to :html5 for non-Nokogiri nodes.
+      def html_version_from_node(node)
+        if nokogiri?
+          if node.is_a?(Nokogiri::HTML5::Document) ||
+              node.is_a?(Nokogiri::HTML5::DocumentFragment)
+            :html5
+          elsif node.is_a?(Nokogiri::HTML4::Document) ||
+              node.is_a?(Nokogiri::HTML4::DocumentFragment)
+            :html4
+          else
+            :html5
+          end
+        else
+          :html5
+        end
+      end
+
+      # Parse an HTML string into an XML fragment.
+      def xml_fragment(html_string)
+        if nokogiri?
+          Nokogiri::XML.fragment(html_string)
+        else
+          raise Canon::Error,
+                "HTML fragment parsing requires the Nokogiri backend"
+        end
+      end
+
       private
 
       def detect

data/lib/canon/xml_parsing.rb

@@ -14,7 +14,7 @@ module Canon
   module XmlParsing
     class << self
       def moxml_context
-        @moxml_context ||= Moxml.new(:oga)
+        @moxml_context ||= Moxml.new(RUBY_ENGINE == "opal" ? :rexml : :oga)
       end
 
       # --- Parsing ---
@@ -47,10 +47,15 @@ module Canon
       end
 
       # --- Type checks (backend-safe) ---
+      #
+      # Both Nokogiri and Moxml are loaded as dependencies. XmlBackend
+      # determines which is used for *parsing*, but nodes from either
+      # library may flow through comparison code (e.g. tests, format
+      # detection). Under Nokogiri backend, both types are checked.
 
       def document?(obj)
         if XmlBackend.nokogiri?
-          obj.is_a?(Nokogiri::XML::Document)
+          obj.is_a?(Nokogiri::XML::Document) || obj.is_a?(Moxml::Document)
         else
           obj.is_a?(Moxml::Document)
         end
@@ -58,7 +63,7 @@ module Canon
 
       def xml_node?(obj)
         if XmlBackend.nokogiri?
-          obj.is_a?(Nokogiri::XML::Node)
+          obj.is_a?(Nokogiri::XML::Node) || obj.is_a?(Moxml::Node)
         else
           obj.is_a?(Moxml::Node)
         end
@@ -66,7 +71,7 @@ module Canon
 
       def element?(node)
         if XmlBackend.nokogiri?
-          node.is_a?(Nokogiri::XML::Element)
+          node.is_a?(Nokogiri::XML::Element) || node.is_a?(Moxml::Element)
         else
           node.is_a?(Moxml::Element)
         end
@@ -74,7 +79,7 @@ module Canon
 
       def text_node?(node)
         if XmlBackend.nokogiri?
-          node.is_a?(Nokogiri::XML::Text)
+          node.is_a?(Nokogiri::XML::Text) || node.is_a?(Moxml::Text)
         else
           node.is_a?(Moxml::Text)
         end
@@ -82,7 +87,7 @@ module Canon
 
       def comment?(node)
         if XmlBackend.nokogiri?
-          node.is_a?(Nokogiri::XML::Comment)
+          node.is_a?(Nokogiri::XML::Comment) || node.is_a?(Moxml::Comment)
         else
           node.is_a?(Moxml::Comment)
         end
@@ -90,7 +95,7 @@ module Canon
 
       def cdata?(node)
         if XmlBackend.nokogiri?
-          node.is_a?(Nokogiri::XML::CDATA)
+          node.is_a?(Nokogiri::XML::CDATA) || node.is_a?(Moxml::Cdata)
         else
           node.is_a?(Moxml::Cdata)
         end
@@ -98,7 +103,7 @@ module Canon
 
       def processing_instruction?(node)
         if XmlBackend.nokogiri?
-          node.is_a?(Nokogiri::XML::ProcessingInstruction)
+          node.is_a?(Nokogiri::XML::ProcessingInstruction) || node.is_a?(Moxml::ProcessingInstruction)
         else
           node.is_a?(Moxml::ProcessingInstruction)
         end
@@ -108,7 +113,7 @@ module Canon
         if XmlBackend.nokogiri?
           obj.is_a?(Nokogiri::XML::DocumentFragment)
         else
-          obj.is_a?(Moxml::DocumentFragment)
+          false
         end
       end
 
@@ -142,7 +147,14 @@ module Canon
         if XmlBackend.nokogiri?
           node.is_a?(Nokogiri::XML::Node) ? node.content : node.to_s
         else
-          node.is_a?(Moxml::Node) ? node.text : node.to_s
+          case node
+          when Moxml::Text, Moxml::Cdata, Moxml::Comment
+            node.content.to_s
+          when Moxml::Node
+            node.text.to_s
+          else
+            node.to_s
+          end
         end
       end
 
@@ -252,17 +264,19 @@ module Canon
         node.to_xml
       end
 
-      def moxml_canonicalize(node, _options)
-        node.to_xml
+      def moxml_canonicalize(_node, _options)
+        raise Canon::Error,
+              "C14N canonicalization is not supported by the moxml backend. " \
+              "Use the Nokogiri backend or a different preprocessing option."
       end
 
       def moxml_node_type(node)
-        return :element if node.is_a?(Moxml::Element)
-        return :text if node.is_a?(Moxml::Text)
-        return :comment if node.is_a?(Moxml::Comment)
-        return :cdata if node.is_a?(Moxml::Cdata)
-        return :document if node.is_a?(Moxml::Document)
-        return :processing_instruction if node.is_a?(Moxml::ProcessingInstruction)
+        return :element if node.element?
+        return :text if node.text?
+        return :comment if node.comment?
+        return :cdata if node.cdata?
+        return :document if node.document?
+        return :processing_instruction if node.processing_instruction?
 
         nil
       end