canon 0.2.9 → 0.2.12

data/lib/canon/formatters/html5_formatter.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "html_formatter_base"
-
 module Canon
   module Formatters
     # HTML5 formatter using Nokogiri::HTML5 parser

data/lib/canon/formatters/html_formatter.rb

@@ -1,9 +1,6 @@
 # frozen_string_literal: true
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
-require_relative "html_formatter_base"
-require_relative "../pretty_printer/html"
-require_relative "../validators/html_validator"
 
 module Canon
   module Formatters

data/lib/canon/formatters/json_formatter.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "json"
-require_relative "../validators/json_validator"
 
 module Canon
   module Formatters

data/lib/canon/formatters/xml_formatter.rb

@@ -1,10 +1,6 @@
 # frozen_string_literal: true
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
-require_relative "../xml_backend"
-require_relative "../xml/c14n"
-require_relative "../pretty_printer/xml"
-require_relative "../validators/xml_validator"
 
 module Canon
   module Formatters

data/lib/canon/formatters/yaml_formatter.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "yaml"
-require_relative "../validators/yaml_validator"
 
 module Canon
   module Formatters

data/lib/canon/formatters.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module Canon
+  # Format-specific canonicalizers and parsers for the top-level
+  # `Canon.format` / `Canon.parse` API. Distinct from {DiffFormatter},
+  # which renders comparison output.
+  module Formatters
+    autoload :Html4Formatter, "canon/formatters/html4_formatter"
+    autoload :Html5Formatter, "canon/formatters/html5_formatter"
+    autoload :HtmlFormatter, "canon/formatters/html_formatter"
+    autoload :HtmlFormatterBase, "canon/formatters/html_formatter_base"
+    autoload :JsonFormatter, "canon/formatters/json_formatter"
+    autoload :XmlFormatter, "canon/formatters/xml_formatter"
+    autoload :YamlFormatter, "canon/formatters/yaml_formatter"
+  end
+end

data/lib/canon/html/data_model.rb

@@ -1,14 +1,6 @@
 # frozen_string_literal: true
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
-require_relative "../data_model"
-require_relative "../xml/nodes/root_node"
-require_relative "../xml/nodes/element_node"
-require_relative "../xml/nodes/namespace_node"
-require_relative "../xml/nodes/attribute_node"
-require_relative "../xml/nodes/text_node"
-require_relative "../xml/nodes/comment_node"
-require_relative "../xml/nodes/processing_instruction_node"
 
 module Canon
   module Html
@@ -81,7 +73,6 @@ module Canon
       def self.serialize(node)
         # HTML nodes use the same serialization as XML
         # Delegate to XML serialization implementation
-        require_relative "../xml/data_model"
         Canon::Xml::DataModel.serialize(node)
       end
 
@@ -89,7 +80,7 @@ module Canon
       def self.build_from_nokogiri(nokogiri_doc)
         root = Canon::Xml::Nodes::RootNode.new
 
-        if nokogiri_doc.respond_to?(:root) && nokogiri_doc.root
+        if nokogiri_doc.is_a?(Nokogiri::XML::Document) && nokogiri_doc.root
           # For Documents (HTML4, HTML5): process the root element
           root.add_child(build_element_node(nokogiri_doc.root))
 
@@ -222,7 +213,6 @@ module Canon
           whitespace_sensitive_tags = %w[pre code textarea script style]
 
           # Check if whitespace is between inline siblings
-          require_relative "../comparison/whitespace_sensitivity"
           unless whitespace_sensitive_tags.include?(parent_name) ||
               Canon::Comparison::WhitespaceSensitivity.inline_whitespace_significant?(nokogiri_text)
             return nil

data/lib/canon/html.rb

@@ -1,9 +1,10 @@
 # frozen_string_literal: true
 
-require_relative "html/data_model"
-
 module Canon
-  # HTML-specific functionality for Canon
+  # HTML-specific functionality for Canon.
+  #
+  # Children are autoloaded — never `require_relative` them.
   module Html
+    autoload :DataModel, "canon/html/data_model"
   end
 end

data/lib/canon/options/cli_generator.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "registry"
-
 module Canon
   module Options
     # Generates Thor CLI options from the Options Registry

data/lib/canon/options/registry.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../comparison/match_options"
-
 module Canon
   module Options
     # Centralized registry for all Canon options

data/lib/canon/options.rb

@@ -0,0 +1,9 @@
+# frozen_string_literal: true
+
+module Canon
+  # CLI option definitions and the option registry.
+  module Options
+    autoload :CliGenerator, "canon/options/cli_generator"
+    autoload :Registry, "canon/options/registry"
+  end
+end

data/lib/canon/pretty_printer/html.rb

@@ -2,7 +2,6 @@
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
 require "stringio"
-require_relative "html_void_elements"
 
 module Canon
   module PrettyPrinter

data/lib/canon/pretty_printer/xml_normalized.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require "nokogiri" unless RUBY_ENGINE == "opal"
-require_relative "html_void_elements"
 
 module Canon
   module PrettyPrinter
@@ -398,7 +397,6 @@ module Canon
 
       # Load the default visualization map from DiffFormatter constants.
       def default_vis_map
-        require_relative "../diff_formatter"
         Canon::DiffFormatter::DEFAULT_VISUALIZATION_MAP
       rescue LoadError, NameError
         { " " => "░", "\t" => "⇥", "\n" => "↵", "\r" => "⏎", "\u00A0" => "␣" }

data/lib/canon/pretty_printer.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+module Canon
+  # Pretty-printers that emit format-aware, fixture-ready output.
+  module PrettyPrinter
+    autoload :Html, "canon/pretty_printer/html"
+    autoload :HtmlVoidElements, "canon/pretty_printer/html_void_elements"
+    autoload :Json, "canon/pretty_printer/json"
+    autoload :Xml, "canon/pretty_printer/xml"
+    autoload :XmlNormalized, "canon/pretty_printer/xml_normalized"
+  end
+end

data/lib/canon/tree_diff/adapters/html_adapter.rb

@@ -193,7 +193,7 @@ module Canon
         # @param element [Nokogiri::XML::Element] Element to check
         # @return [Boolean] True if element is whitespace-sensitive
         def whitespace_sensitive?(element)
-          return false unless element.respond_to?(:name)
+          return false unless Canon::Comparison::NodeInspector.element_node?(element)
 
           # List of HTML elements where whitespace is semantically significant
           whitespace_sensitive_tags = %w[pre code textarea script style]

data/lib/canon/tree_diff/adapters.rb

@@ -0,0 +1,14 @@
+# frozen_string_literal: true
+
+module Canon
+  module TreeDiff
+    # Format-specific tree adapters that wrap parsed documents into
+    # Canon::TreeDiff::Core::TreeNode trees for the matcher.
+    module Adapters
+      autoload :HTMLAdapter, "canon/tree_diff/adapters/html_adapter"
+      autoload :JSONAdapter, "canon/tree_diff/adapters/json_adapter"
+      autoload :XMLAdapter, "canon/tree_diff/adapters/xml_adapter"
+      autoload :YAMLAdapter, "canon/tree_diff/adapters/yaml_adapter"
+    end
+  end
+end

data/lib/canon/tree_diff/core/attribute_comparator.rb

@@ -35,18 +35,12 @@ module Canon
         # @param attrs2 [Hash] Second attribute hash
         # @return [Boolean] True if attributes are considered equal
         def equal?(attrs1, attrs2)
-          # Handle nil/empty cases
           return true if attrs1.nil? && attrs2.nil?
           return false if attrs1.nil? || attrs2.nil?
 
-          attrs1 = attrs1.to_h if attrs1.respond_to?(:to_h)
-          attrs2 = attrs2.to_h if attrs2.respond_to?(:to_h)
-
           if attribute_order == :strict
-            # Strict mode: order matters
             attrs1 == attrs2
           else
-            # Ignore/normalize mode: sort keys for comparison
             normalize_for_comparison(attrs1) == normalize_for_comparison(attrs2)
           end
         end

data/lib/canon/tree_diff/core/node_signature.rb

@@ -142,7 +142,7 @@ module Canon
           whitespace_sensitive_tags = %w[pre code textarea script style]
 
           # Check if this node is whitespace-sensitive
-          if node.respond_to?(:label)
+          if node.is_a?(TreeNode)
             label = node.label.to_s.downcase
             return true if whitespace_sensitive_tags.include?(label)
           end

data/lib/canon/tree_diff/core/tree_node.rb

@@ -350,15 +350,22 @@ module Canon
         #
         # @return [String] XPath expression
         def xpath
-          # If we have a source node that supports xpath, use it
-          if @source_node.respond_to?(:path)
-            return @source_node.path
-          end
+          return @source_node.path if nokogiri_source?(@source_node)
 
-          # Otherwise construct path from tree structure
           construct_path
         end
 
+        # True when the supplied source node is a Nokogiri node that
+        # exposes an XPath via +path+.  Nokogiri is an optional
+        # backend so the constant is guarded.
+        def nokogiri_source?(node)
+          return false unless Canon::XmlBackend.nokogiri?
+
+          node.is_a?(Nokogiri::XML::Node)
+        end
+
+        private :nokogiri_source?
+
         # Construct path from tree structure
         #
         # @return [String] Path expression

data/lib/canon/tree_diff/core.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+module Canon
+  module TreeDiff
+    # Core tree-diff primitives: TreeNode, matching, signatures, weights,
+    # attribute comparison, and XML entity decoding. Children are
+    # autoloaded — never `require_relative` them.
+    module Core
+      autoload :AttributeComparator, "canon/tree_diff/core/attribute_comparator"
+      autoload :Matching, "canon/tree_diff/core/matching"
+      autoload :NodeSignature, "canon/tree_diff/core/node_signature"
+      autoload :NodeWeight, "canon/tree_diff/core/node_weight"
+      autoload :TreeNode, "canon/tree_diff/core/tree_node"
+      autoload :XmlEntityDecoder, "canon/tree_diff/core/xml_entity_decoder"
+    end
+  end
+end

data/lib/canon/tree_diff/matchers/hash_matcher.rb

@@ -1,12 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../core/tree_node"
-require_relative "../core/node_signature"
-require_relative "../core/node_weight"
-require_relative "../core/matching"
-require_relative "../core/attribute_comparator"
-require_relative "../core/xml_entity_decoder"
-
 module Canon
   module TreeDiff
     module Matchers

data/lib/canon/tree_diff/matchers/similarity_matcher.rb

@@ -1,9 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../core/tree_node"
-require_relative "../core/node_signature"
-require_relative "../core/matching"
-
 module Canon
   module TreeDiff
     module Matchers
@@ -165,7 +161,7 @@ module Canon
           whitespace_sensitive_tags = %w[pre code textarea script style]
 
           # Check if this node is whitespace-sensitive
-          if node.respond_to?(:label)
+          if node.is_a?(Core::TreeNode)
             label = node.label.to_s.downcase
             return true if whitespace_sensitive_tags.include?(label)
           end

data/lib/canon/tree_diff/matchers/structural_propagator.rb

@@ -1,9 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../core/tree_node"
-require_relative "../core/node_weight"
-require_relative "../core/matching"
-
 module Canon
   module TreeDiff
     module Matchers
@@ -218,7 +214,7 @@ module Canon
           whitespace_sensitive_tags = %w[pre code textarea script style]
 
           # Check if this node is whitespace-sensitive
-          if node.respond_to?(:label)
+          if node.is_a?(Core::TreeNode)
             label = node.label.to_s.downcase
             return true if whitespace_sensitive_tags.include?(label)
           end

data/lib/canon/tree_diff/matchers.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+module Canon
+  module TreeDiff
+    # Tree-matching strategies — hash-based, similarity-based,
+    # structural propagation, and the universal fallback.
+    module Matchers
+      autoload :HashMatcher, "canon/tree_diff/matchers/hash_matcher"
+      autoload :SimilarityMatcher, "canon/tree_diff/matchers/similarity_matcher"
+      autoload :StructuralPropagator,
+               "canon/tree_diff/matchers/structural_propagator"
+      autoload :UniversalMatcher, "canon/tree_diff/matchers/universal_matcher"
+    end
+  end
+end

data/lib/canon/tree_diff/operation_converter.rb

@@ -1,13 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../diff/diff_node"
-require_relative "../comparison/match_options"
-# OperationConverter helper modules
-require_relative "operation_converter_helpers/metadata_enricher"
-require_relative "operation_converter_helpers/reason_builder"
-require_relative "operation_converter_helpers/post_processor"
-require_relative "operation_converter_helpers/update_change_handler"
-
 module Canon
   module TreeDiff
     # Converts TreeDiff Operations to DiffNodes for integration with Canon's
@@ -124,6 +116,8 @@ module Canon
         end
       end
 
+      public :convert_operation
+
       # Convert INSERT operation to DiffNode
       #
       # @param operation [Operation] Insert operation
@@ -153,7 +147,7 @@ module Canon
 
       # Determine dimension for INSERT/DELETE operations based on node type
       def dimension_for_insert_delete(tree_node)
-        label = tree_node.respond_to?(:label) ? tree_node.label : nil
+        label = tree_node.is_a?(Canon::TreeDiff::Core::TreeNode) ? tree_node.label : nil
         return :comments if label == "comment"
 
         :element_structure
@@ -359,7 +353,7 @@ module Canon
       def extract_source_node(tree_node)
         return nil if tree_node.nil?
 
-        tree_node.respond_to?(:source_node) ? tree_node.source_node : tree_node
+        tree_node.is_a?(Canon::TreeDiff::Core::TreeNode) ? tree_node.source_node : tree_node
       end
 
       # Determine if a diff is normative based on match options
@@ -383,12 +377,10 @@ module Canon
         return false if node.nil?
 
         # Get element name from node
-        element_name = if node.respond_to?(:label)
-                         node.label # TreeNode
-                       elsif node.respond_to?(:name)
-                         node.name # Nokogiri node
+        element_name = if node.is_a?(Canon::TreeDiff::Core::TreeNode)
+                         node.label
                        else
-                         return false
+                         Canon::Comparison::NodeInspector.name(node)
                        end
 
         # Check if it's in our metadata elements list

data/lib/canon/tree_diff/operation_converter_helpers/metadata_enricher.rb

@@ -1,8 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../../diff/path_builder"
-require_relative "../../diff/node_serializer"
-
 module Canon
   module TreeDiff
     module OperationConverterHelpers
@@ -45,13 +42,7 @@ module Canon
         def self.serialize(tree_node)
           return nil if tree_node.nil?
 
-          # Extract source node from TreeNode
-          source = if tree_node.respond_to?(:source_node)
-                     tree_node.source_node
-                   else
-                     tree_node
-                   end
-
+          source = tree_node.is_a?(Core::TreeNode) ? tree_node.source_node : tree_node
           Canon::Diff::NodeSerializer.serialize(source)
         end
 
@@ -62,8 +53,7 @@ module Canon
         def self.extract_attributes(tree_node)
           return nil if tree_node.nil?
 
-          # Use TreeNode's attributes directly (already normalized by adapter)
-          tree_node.respond_to?(:attributes) ? (tree_node.attributes || {}) : {}
+          tree_node.is_a?(Core::TreeNode) ? (tree_node.attributes || {}) : {}
         end
       end
     end

data/lib/canon/tree_diff/operation_converter_helpers/post_processor.rb

@@ -20,7 +20,7 @@ module Canon
           # For each DELETE, try to find a matching INSERT
           deletes.each do |delete_node|
             node1 = delete_node.node1
-            next unless node1.respond_to?(:name) && node1.respond_to?(:attributes)
+            next unless backend_element?(node1)
 
             # Skip if node has no attributes (can't be attribute order diff)
             next if node1.attributes.nil? || node1.attributes.empty?
@@ -28,7 +28,7 @@ module Canon
             # Find inserts with same element name at same position
             matching_insert = inserts.find do |insert_node|
               node2 = insert_node.node2
-              next false unless node2.respond_to?(:name) && node2.respond_to?(:attributes)
+              next false unless backend_element?(node2)
               next false unless node1.name == node2.name
 
               # Must have attributes to differ in order
@@ -59,6 +59,17 @@ module Canon
           diff_nodes
         end
 
+        # True when +node+ is a backend element (Nokogiri or Moxml) that
+        # exposes its attributes via +attributes+.  Canon-native
+        # +ElementNode+ uses +attribute_nodes+ and is therefore excluded.
+        def self.backend_element?(node)
+          return false unless node
+          return false unless Canon::Comparison::NodeInspector.element_node?(node)
+
+          Canon::XmlBackend.nokogiri? ? node.is_a?(Nokogiri::XML::Node) : false
+        end
+        private_class_method :backend_element?
+
         # Check if two attribute hashes are equal ignoring order
         #
         # @param attrs1 [Hash] First attribute hash
@@ -68,11 +79,6 @@ module Canon
           return true if attrs1.nil? && attrs2.nil?
           return false if attrs1.nil? || attrs2.nil?
 
-          # Convert to hashes if needed
-          attrs1 = attrs1.to_h if attrs1.respond_to?(:to_h)
-          attrs2 = attrs2.to_h if attrs2.respond_to?(:to_h)
-
-          # Compare as sets (order-independent)
           attrs1.sort.to_h == attrs2.sort.to_h
         end
 

data/lib/canon/tree_diff/operation_converter_helpers/reason_builder.rb

@@ -16,7 +16,7 @@ module Canon
           node = operation[:node]
           content = operation[:content]
 
-          if node.respond_to?(:label)
+          if node.is_a?(Core::TreeNode)
             # Include content preview for clarity
             "Element inserted: #{content || "<#{node.label}>"}"
           else
@@ -32,7 +32,7 @@ module Canon
           node = operation[:node]
           content = operation[:content]
 
-          if node.respond_to?(:label)
+          if node.is_a?(Core::TreeNode)
             # Include content preview for clarity
             "Element deleted: #{content || "<#{node.label}>"}"
           else

data/lib/canon/tree_diff/operation_converter_helpers/update_change_handler.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../operation_converter_helpers/reason_builder"
-
 module Canon
   module TreeDiff
     module OperationConverterHelpers
@@ -20,8 +18,8 @@ module Canon
         def self.convert(operation, metadata, is_metadata, normative_determiner)
           tree_node1 = operation[:node1] # TreeNode from adapter
           tree_node2 = operation[:node2] # TreeNode from adapter
-          node1 = tree_node1.respond_to?(:source_node) ? tree_node1.source_node : tree_node1
-          node2 = tree_node2.respond_to?(:source_node) ? tree_node2.source_node : tree_node2
+          node1 = tree_node1.is_a?(Core::TreeNode) ? tree_node1.source_node : tree_node1
+          node2 = tree_node2.is_a?(Core::TreeNode) ? tree_node2.source_node : tree_node2
           changes = operation[:changes]
 
           # Handle case where changes is a boolean or non-hash value
@@ -150,8 +148,8 @@ is_metadata, normative_determiner, tree_node1, tree_node2)
         # @param tree_node2 [Object] Second tree node
         # @return [Symbol] The dimension to use (:text_content or :comments)
         def self.dimension_for_value_change(tree_node1, tree_node2)
-          label1 = tree_node1.respond_to?(:label) ? tree_node1.label : nil
-          label2 = tree_node2.respond_to?(:label) ? tree_node2.label : nil
+          label1 = tree_node1.is_a?(Core::TreeNode) ? tree_node1.label : nil
+          label2 = tree_node2.is_a?(Core::TreeNode) ? tree_node2.label : nil
 
           # If either node is a comment, use :comments dimension
           return :comments if label1 == "comment" || label2 == "comment"

data/lib/canon/tree_diff/operation_converter_helpers.rb

@@ -0,0 +1,18 @@
+# frozen_string_literal: true
+
+module Canon
+  module TreeDiff
+    # Helper modules consumed by OperationConverter when converting
+    # tree operations into DiffNodes.
+    module OperationConverterHelpers
+      autoload :MetadataEnricher,
+               "canon/tree_diff/operation_converter_helpers/metadata_enricher"
+      autoload :PostProcessor,
+               "canon/tree_diff/operation_converter_helpers/post_processor"
+      autoload :ReasonBuilder,
+               "canon/tree_diff/operation_converter_helpers/reason_builder"
+      autoload :UpdateChangeHandler,
+               "canon/tree_diff/operation_converter_helpers/update_change_handler"
+    end
+  end
+end

data/lib/canon/tree_diff/operations/operation_detector.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require_relative "../core/xml_entity_decoder"
-
 module Canon
   module TreeDiff
     module Operations
@@ -600,6 +598,10 @@ module Canon
           depth
         end
 
+        public :normalize_text, :calculate_depth, :text_similarity,
+               :extract_text_content, :collect_all_nodes, :nodes_identical?,
+               :detect_changes
+
         # Check if a node is in a whitespace-sensitive context
         #
         # HTML elements where whitespace is significant: <pre>, <code>, <textarea>, <script>, <style>
@@ -615,13 +617,12 @@ module Canon
           # Check if this node or any ancestor is whitespace-sensitive
           current = node
           while current
-            if current.respond_to?(:label)
+            if current.is_a?(Core::TreeNode)
               label = current.label.to_s.downcase
               return true if whitespace_sensitive_tags.include?(label)
             end
 
-            # Check parent
-            current = current.parent if current.respond_to?(:parent)
+            current = current.is_a?(Core::TreeNode) ? current.parent : nil
             break unless current
           end
 

data/lib/canon/tree_diff/operations.rb

@@ -0,0 +1,13 @@
+# frozen_string_literal: true
+
+module Canon
+  module TreeDiff
+    # Tree-diff operations: Operation (INSERT/DELETE/UPDATE/MOVE) and
+    # the OperationDetector that emits them.
+    module Operations
+      autoload :Operation, "canon/tree_diff/operations/operation"
+      autoload :OperationDetector,
+               "canon/tree_diff/operations/operation_detector"
+    end
+  end
+end